CHARACTERISING THE RELATIONSHIP BETWEEN FATIGUE AND DEPRESSION · 2017-09-04 · depression and...
Transcript of CHARACTERISING THE RELATIONSHIP BETWEEN FATIGUE AND DEPRESSION · 2017-09-04 · depression and...
CHARACTERISING THE RELATIONSHIP
BETWEEN FATIGUE AND DEPRESSION
Elizabeth Corfield
BSc (Honours I)
Submitted in fulfilment of the requirements for the degree of
Doctor of Philosophy
Institute of Health and Biomedical Innovation
Faculty of Health
Queensland University of Technology
2017
Characterising the Relationship between Fatigue and Depression i
Keywords
Candidate gene, chronic fatigue syndrome, comorbidity, depression, familiality,
fatigue, gene-based analysis, genetic relationship, genetics, genetic association,
genome-wide association, heritability, major depression, minor depression, major
depressive disorder, minor depressive disorder, population genetics,
symptomatology, and twin study.
ii Characterising the Relationship between Fatigue and Depression
Abstract
Fatigue is a common symptom, which is represented by a spectrum and associated
with numerous disorders, including major depressive disorder. Fatigue associated
with major depressive disorder further reduces the quality of life and increased
functional impairment in patients. However, little is known about the etiology of
fatigue and major depressive disorder. Additionally, the underlying mechanisms that
facilitate the high prevalence of comorbid fatigue and depression are poorly
understood. The objective of this dissertation was to increase our knowledge and
understanding of the comorbidity and genetics of fatigue and depression.
The majority of this project has been conducted utilising an older, Australian
twin cohort (comprising 2,281 twin pairs) with a self-report measure of fatigue and
depression. Microarray genotyping data was available for a subset of this population
(307 fatigue cases and 744 non-fatigued controls). Phenotypic characterisation was
undertaken to determine the fatigue and depression status of individuals within the
cohort. Additionally, a small chronic fatigue syndrome cohort (comprising 47
patients and 55 healthy controls) was utilised in Chapter 7. Microarray genotyping
data was available for all individuals in the chronic fatigue syndrome cohort.
A symptomatic analysis was conducted to determine if the depression symptom
profile differed between fatigued and non-fatigued individuals. Similarly, differences
in the fatigue symptom profile were investigated in individuals with major depressive
disorder, minor depressive disorder, and non-depressed individuals. This analysis
was conducted utilising logistic regression modelling. Results from this analysis
indicated fatigued individuals experienced significantly increased depression
symptomatology and prevalence. The most significant finding from this analysis was
the identification that the overlapping symptomatology between fatigue and
depression was not driving the association between the two phenotypes.
The familiality and heritability of fatigue were then investigated to determine
the relative importance of genetic and environmental factors in the total phenotypic
variation. Relative risks and structural equation modelling were utilising within this
analysis, which was replicated for major depressive disorder and minor depressive
disorder. Additionally, the liability threshold model was fitted to determine if major
Characterising the Relationship between Fatigue and Depression iii
depressive disorder and minor depressive disorder exist on a continuum. A larger
Australian depression cohort was utilised to determine whether broadening the
depression case phenotype, to include minor depressive disorder, in genome-wide
association analyses, will facilitate the elucidation of the molecular mechanisms of
major depressive disorder. Results from these analyses indicated fatigue, minor
depressive disorder, and major depressive disorder are all familial and have
significant additive genetic contributions. The most important finding from these
analyses is that minor depressive disorder and major depressive disorder exist on a
genetic continuum and that utilisation of a broad depression phenotype (which
includes both minor depressive disorder and major depressive disorder cases) should
facilitate further elucidation of the underlying genetic architecture of major
depressive disorder.
Expansion of the familiality and heritability analyses was utilised to assess the
magnitude of shared heritability between depression and fatigue. Furthermore, the
co-twin control method was utilised to determine whether the association between
depression and fatigue is explained by a causal, non-causal, non-causal shared
environment, or non-causal genetic model. Results from these analyses indicated a
significant additive genetic correlation of 0.71 (95% confidence interval = 0.51-0.92)
and bivariate heritability of 21% (95% confidence interval = 10-35%) exist between
depression and fatigue. Additionally, the association between depression and fatigue
is likely explained by a non-causal genetic relationship. The most important finding
from this analysis was that the contribution of shared genetic factors remained
significant independently of the overlapping symptomatology of the traits.
A genome-wide association analysis and gene-based investigation of fatigue
were then conducted, utilising linear mixed modelling, including a genetic
relationship matrix, to account for the relatedness within the data. While, an
evaluation of previous genetic findings associated with CFS was conducted, utilising
a chi-squared allelic test and gene-based analysis was conducted in a chronic fatigue
syndrome cohort. Results from this analysis indicated previously implicated genes
and risk loci are likely false positives and are unlikely to be associated with fatigue
or chronic fatigue syndrome. The most important finding from this analysis was the
identification of six genomic locations of interest, which are potentially associated
with fatigue.
iv Characterising the Relationship between Fatigue and Depression
Overall, these results provided evidence supporting a substantial additive
genetic overlap between fatigue and depression, which is independent of their
overlapping symptoms.
Characterising the Relationship between Fatigue and Depression v
Table of Contents
Keywords .................................................................................................................................. i
Abstract .................................................................................................................................... ii
Table of Contents ......................................................................................................................v
List of Figures ....................................................................................................................... viii
List of Tables .......................................................................................................................... ix
List of Abbreviations ............................................................................................................. xii
List of Publications ............................................................................................................... xiv
List of Presentations ................................................................................................................xv
Statement of Original Authorship ......................................................................................... xvi
Acknowledgements .............................................................................................................. xvii
Chapter 1: Introduction ...................................................................................... 1
1.1 Background and Significance .........................................................................................1
1.2 Purpose ...........................................................................................................................2 1.2.1 Aims .....................................................................................................................2 1.2.2 Hypotheses ...........................................................................................................3
1.3 Thesis Outline .................................................................................................................4
Chapter 2: Literature Review ............................................................................. 5
2.1 Fatigue Classifications ....................................................................................................5
2.2 Classification of CFS, ME/CFS, ME, and SEID ............................................................6 2.2.1 CFS Definition .....................................................................................................6 2.2.2 ME/CFS Definition ..............................................................................................7 2.2.3 ME Definition .......................................................................................................8 2.2.4 SEID Definition ....................................................................................................9 2.2.5 Differences between CFS, ME/CFS, and ME ......................................................9
2.3 Epidemiology of Fatigue ..............................................................................................11
2.4 Pathophysiology of CFS, ME/CFS and ME .................................................................13 2.4.1 Infection and Immune Dysfunction ....................................................................13 2.4.2 Endocrine and Metabolic Dysfunction ...............................................................14 2.4.3 Cardiovascular and Neurologic Dysfunction .....................................................15 2.4.4 Psychiatric Disorders ..........................................................................................15 2.4.5 Genetics ..............................................................................................................16
2.5 Heritability of Fatigue...................................................................................................16
2.6 Molecular Genetics of Fatigue ......................................................................................18
2.7 Classification of MDD and MiDD................................................................................25
2.8 Epidemiology of MDD and MiDD ...............................................................................26
2.9 Pathophysiology of MDD .............................................................................................27 2.9.1 Endocrine and Neurologic Dysfunction .............................................................27 2.9.2 Genetics ..............................................................................................................27 2.9.3 Environmental Factors .......................................................................................28
vi Characterising the Relationship between Fatigue and Depression
2.10 Heritability of MDD and MiDD .................................................................................. 28
2.11 Molecular Genetics of MDD ........................................................................................ 32
2.12 Comorbidity between MDD and Fatigue ..................................................................... 36 2.12.1 Heritability Links between Fatigue and Depression .......................................... 37
Chapter 3: Co-occurrence and Symptomatology of Fatigue and Depression49
3.1 Abstract ........................................................................................................................ 51
3.2 Introduction .................................................................................................................. 52
3.3 Methods ........................................................................................................................ 55 3.3.1 Sample and Questionnaires ................................................................................ 55 3.3.2 Statistical Analysis ............................................................................................. 59
3.4 Results .......................................................................................................................... 61 3.4.1 Study Population ................................................................................................ 61 3.4.2 Fatigued Individuals Report a Higher Proportion of Depression
Symptoms .......................................................................................................... 62 3.4.3 Depressed Individuals Report Higher Proportions of Fatigue Symptoms ......... 64
3.5 Discussion .................................................................................................................... 67
Chapter 4: Familiality and Heritability of Fatigue in an Australian Twin
Sample 72
4.1 Abstract ........................................................................................................................ 74
4.2 Introduction .................................................................................................................. 75
4.3 Methods ........................................................................................................................ 79 4.3.1 Study Cohort and Fatigue Classification ........................................................... 79 4.3.2 Statistical Analysis ............................................................................................. 80
4.4 Results .......................................................................................................................... 81
4.5 Discussion .................................................................................................................... 84
Chapter 5: A Continuum of Genetic Liability for Minor and Major
Depression 87
5.1 Abstract ........................................................................................................................ 89
5.2 Introduction .................................................................................................................. 90
5.3 Materials and Methods ................................................................................................. 92 5.3.1 Study Cohorts .................................................................................................... 92 5.3.2 Statistical Analysis ............................................................................................. 93
5.4 Results .......................................................................................................................... 95
5.5 Discussion .................................................................................................................. 102
Chapter 6: Shared Genetic Factors in the Co-Occurrence of Depression and
Fatigue 107
6.1 Abstract ...................................................................................................................... 109
6.2 Introduction ................................................................................................................ 110
6.3 Materials and Methods ............................................................................................... 111 6.3.1 Study Cohort .................................................................................................... 111 6.3.2 Diagnosis of Depression and Fatigue .............................................................. 111 6.3.3 Familial Clustering .......................................................................................... 112 6.3.4 Genetic Analysis .............................................................................................. 113
Characterising the Relationship between Fatigue and Depression vii
6.3.5 Relationship Analysis .......................................................................................114
6.4 Results ........................................................................................................................116 6.4.1 Relative Risks ...................................................................................................116 6.4.2 Polychoric Correlations ....................................................................................119 6.4.3 Bivariate Heritability Estimates .......................................................................119 6.4.4 Co-twin Control ................................................................................................120
6.5 Discussion ...................................................................................................................121
Chapter 7: Systematic Evaluation of Risk Loci from Candidate Gene and
Genome-wide Association Studies of Fatigue ...................................................... 126
7.1 Abstract .......................................................................................................................126
7.2 Introduction ................................................................................................................127
7.3 Methods ......................................................................................................................129 7.3.1 Previously Implicated Genes ............................................................................129 7.3.2 Study Cohorts, Genotyping Data and Quality-control .....................................145 7.3.3 Statistical Analysis ...........................................................................................147
7.4 Results ........................................................................................................................148 7.4.1 Previously Implicated SNPs and Genes ...........................................................148 7.4.2 Genome-wide association results .....................................................................150
7.5 Discussion ...................................................................................................................155
Chapter 8: General Discussion ....................................................................... 159
8.1 Summary of Findings .................................................................................................159
8.2 Limitations ..................................................................................................................161
8.3 Future Directions ........................................................................................................162
8.4 Conclusions ................................................................................................................163
Bibliography ........................................................................................................... 164
Appendices .............................................................................................................. 191
viii Characterising the Relationship between Fatigue and Depression
List of Figures
Figure 6.1. Expected outcomes of the co-twin control method under the causal,
non-causal, non-causal shared environment, and non-causal genetic
models within the general population (light grey), discordant DZ twin
pairs (grey) who share 50% of their genetics and 100% of their
common environment, and discordant MZ twin pairs (dark grey) who
share 100% of their genetics and common environment. Under a
causal model an association is expected within all three groups. Under
a non-causal model, an association is expected within the general
population, discordant DZ cohort will have a small association, and
discordant MZ cohort will have no association. Similarly, under the
non-causal shared environmental model, discordant DZ and MZ twin
pairs have a small, equal association. Finally, under the non-causal
genetic model, discordant DZ twin pairs have an association, whereas
discordant MZ twin pairs have a smaller association. ............................... 115
Figure 6.2. Path diagram of the bivariate Cholesky model variance estimates
(with their 95% confidence intervals) for two-category depression and
fatigue. The observed traits are shown in the rectangles. Similarly, the
latent variables (additive genetic factors: A, and unique environmental
factors: E) are depicted by circles. The arrows depict the relationship
between the variables. ................................................................................ 120
Figure 6.3. Left: The observed odds ratios (OR) for a current diagnosis of
fatigue given a current diagnosis of depression in the general
population (1,266 unrelated twin singles), 99 discordant DZ twin
pairs, and 96 discordant MZ twin pairs. Right: The observed OR for a
current diagnosis of depression given a current diagnosis of fatigue in
the general population (1,266 unrelated twin singles), 200 discordant
DZ twin pairs, and 215 discordant MZ twin pairs. In both situations,
the observed OR patterns are consistent with a non-causal genetic
model. ......................................................................................................... 121
Figure 7.1. Manhattan plot of the chronic fatigue syndrome (CFS) cohort
genome-wide association raw p-values. The horizontal dashed line
corresponds to the genome-wide significance threshold (p < 5 × 10-8).
The three genes suggestively associated (p < 1 × 10-4) with CFS in
gene-based analyses are indicated in green (CDCP2), pink (EMCN),
and blue (CAPRIN1). ................................................................................. 152
Figure 7.2. Manhattan plot of the fatigue cohort genome-wide association raw
p-values. The horizontal dashed line corresponds to the genome-wide
significance threshold (p < 5 × 10-8). The two genes suggestively
associated (p < 1 × 10-4) with fatigue in gene-based analyses are
indicated in pink (TBCA) and blue (PLXDC2). ......................................... 153
Characterising the Relationship between Fatigue and Depression ix
List of Tables
Table 2.1. Comparison of patients meeting the Canadian consensus criteria and
International consensus criteria compared to those diagnosed using the
Centres for Disease Control criteria for differences in demographics,
heart rate, cognitive measures, and responses to the 36-item Short-
form health survey and World Health Organisation disability
adjustment schedule 2.0 items. .................................................................... 10
Table 2.2. Population prevalence estimates for prolonged fatigue, chronic
fatigue, idiopathic chronic fatigue, chronic fatigue syndrome, and
myalgic encephalomyelitis/chronic fatigue syndrome. ................................ 12
Table 2.3. List of the pathogens investigated as potential triggering agents in
chronic fatigue syndrome onset. .................................................................. 14
Table 2.4. Heritability estimates (and their 95% confidence intervals) of the
unique additive genetic factors (A), common environmental factors
(C), and unique environmental factors (E) contributing to fatigue
severity, interfering fatigue, short-duration fatigue, abnormal
tiredness, abnormal fatigue, prolonged fatigue, chronic fatigue,
idiopathic chronic fatigue, and chronic fatigue syndrome. .......................... 18
Table 2.5. Candidate genes and implicated single nucleotide polymorphisms
associated with chronic fatigue syndrome. .................................................. 20
Table 2.6. List of reported genome-wide significant (7.5 × 10-8) risk loci
associated with chronic fatigue syndrome from a genome-wide
association study of 42 cases and 38 controls by Schlauch and
colleagues (2016). ........................................................................................ 23
Table 2.7. List of genes associated (p < 2.77 × 10-6) or suggestively associated
(p < 1.00 × 10-4) with self-reported tiredness. ............................................. 25
Table 2.8. Heritability estimates (and their 95% confidence intervals) of the
unique additive genetic factors (A), common environmental factors
(C), and unique environmental factors (E) contributing to major
depressive disorder (MDD).......................................................................... 30
Table 2.9. Candidate genes associated with major depressive disorder in meta-
analysis studies............................................................................................. 33
Table 2.10. Summary of the genome-wide association studies conducted for
major depressive disorder (MDD). .............................................................. 34
Table 2.11. List of genome-wide significant (5 × 10-8) risk loci associated with
depression from a genome-wide association studies by the
CONVERGE consortium (2015), Hek and colleagues (2013), and
Direk and colleagues (2016). ....................................................................... 36
Table 2.12. Heritability estimates (and their 95% confidence intervals) of the
unique additive genetic factors (A), common environmental factors
(C), and unique environmental factors (E) from previous trivariate and
x Characterising the Relationship between Fatigue and Depression
multivariate common factor twin models, which include a fatigue and
depression phenotype. .................................................................................. 45
Table 2.13. Heritability estimates (and their 95% confidence intervals) of the
unique additive genetic factors (A), common environmental factors
(C), and unique environmental factors (E) from previous bivariate,
trivariate, and multivariate Cholesky twin models, which include a
fatigue and depression phenotype. ............................................................... 46
Table 2.13. Continued Heritability estimates (and their 95% confidence
intervals) of the unique additive genetic factors (A), common
environmental factors (C), and unique environmental factors (E) from
previous bivariate, trivariate, and multivariate Cholesky twin models,
which include a fatigue and depression phenotype. ..................................... 47
Table 3.1. Questionnaire items used to assess fatigue. .............................................. 56
Table 3.2. Questionnaire items used to assess the criteria of a major depressive
episode. ........................................................................................................ 58
Table 3.3. Prevalence ratios of fatigue and depression. ............................................. 63
Table 3.4. Logistic regression, unadjusted and and adjusted for, relatedness,
comparing the depression symptoms exhibited by fatigued individuals
(N = 766) to non-fatigued (N = 1,849) individuals. ..................................... 63
Table 3.5. Logistic regression, both unadjusted and adjusted for relatedness, of
fatigue symptoms exhibited by depressed (N = 275) and non-
depressed (N = 2,340) individuals................................................................ 65
Table 3.6. Logistic regression of fatigue symptoms exhibited by individuals
with major depressive disorder (N = 50), minor depressive disorder (N
= 225), and are non-depressed (N = 2,340). ................................................. 66
Table 4.1. Previously published variance estimates (with their 95% confidence
intervals) for varying fatigue classifications, in adults, from univariate
structural equation modelling. ...................................................................... 78
Table 4.2. Relative riska of fatigue within complete monozygotic (MZ), same-
sex dizygotic (DZss), and opposite-sex dizygotic (DZos) twin pairs. ........... 82
Table 4.3. Tetrachoric correlations (r) with their 95% confidence intervals (CI)
for fatigue according to zygosity. ................................................................ 83
Table 4.4. Fit statistics and variance estimates (with their 95% confidence
intervals) from univariate structural equation modelling. ............................ 84
Table 5.1. Relative riska of depression and fatigue within monozygotic (MZ),
same-sex dizygotic (DZss), and opposite-sex dizygotic (DZos) twin
pairs. ............................................................................................................. 96
Table 5.2. Liability threshold model fit p-values. ...................................................... 97
Table 5.3. Polychoric correlations with their 95% confidence intervals for
depression according to zygosity. ................................................................ 98
Table 5.4. Fit statistics and variance estimates (with their 95% confidence
intervals) from univariate structural equation modelling. .......................... 100
Characterising the Relationship between Fatigue and Depression xi
Table 6.1. Cross-tabulationa of two-category depression and fatigue status
within twin pairs. ....................................................................................... 116
Table 6.2. Relative riska of two-category depression and fatigue within
monozygotic (MZ), same-sex dizygotic (DZss), and opposite-sex
dizygotic (DZos) twin pairs. ....................................................................... 118
Table 6.3. Polychoric correlations with their 95% confidence intervals for two-
category depression and fatigue according to zygosity. ............................ 119
Table 6.4. Bivariate heritability model fits. ............................................................. 120
Table 7.1. Summary of genes from candidate gene association studies for
fatigue traits. .............................................................................................. 131
Table 7.2. Summary of SNPs from genome-wide association studies for
chronic fatigue syndrome. .......................................................................... 136
Table 7.3. Summary of SNPs from genome-wide association study for self-
reported tiredness.a ..................................................................................... 144
Table 7.4. Summary of genes from gene-based association analysis of self-
reported tiredness. ...................................................................................... 144
Table 7.5. Summary of SNPs from genome-wide association studies of
depression phenotypes, in Europeans. ....................................................... 145
Table 7.6. Summary of SNPs reaching suggestive significance thresholds for
chronic fatigue syndrome and fatigue. ....................................................... 154
Table 7.7. Summary of genes reaching suggestive significance thresholds from
gene-based association analysis for chronic fatigue syndrome and
fatigue. ....................................................................................................... 154
xii Characterising the Relationship between Fatigue and Depression
List of Abbreviations
χ2 Chi-squared test
Δ df Difference in degrees of freedom
-2LL Minus two log-likelihood
A Additive genetic factors
AIC Akaike information criterion
bp Base pair
C Common environmental factors
CCC Canadian consensus criteria
CDC Centres for Disease Control
CF Chronic fatigue
CFS Chronic fatigue syndrome
CGA Candidate gene association
Chr Chromosome
CI Confidence interval
D Non-additive (dominance) genetic factors
DSM Diagnostic and statistical manual of mental disorders
DSSI/sAD Delusions-symptoms states inventory, states of anxiety and depression
DZ Dizygotic
E Unique environmental factors
F Female
Freq Frequency
GC Genotyping call
GHQ General Health Questionnaire
GWA Genome-wide association
h2 Narrow-sense heritability
HR Heart rate
HRC Haplotype reference consortium
HREC Human Research Ethics Committee
HWE Hardy-Weinberg equilibrium
ICC International consensus criteria
ICF Idiopathic chronic fatigue
IOM Institute of Medicine
ins/del Insertion/deletion
M Male
MAF Minor allele frequency
matSpD Matrix spectral decomposition
ME Myalgic encephalomyelitis
ME/CFS Myalgic encephalomyelitis/chronic fatigue syndrome
MDD Major depressive disorder
MD Major depression
Characterising the Relationship between Fatigue and Depression xiii
MiDD Minor depressive disorder
MiD Minor depression
MZ Monozygotic
N Number
NC Not calculable
NS Non-significant
OA Other allele
OR Odds ratio
os Opposite sex
p p-value
PR Prevalence ratio
RR Relative risk
QIMRB QIMR Berghofer Medical Research Institute
r Correlation
RA Risk allele
rc Common environmental correlation
re Unique environmental correlation
rg Genetic correlation
SE Standard error
SEID Systematic exertion intolerance disease
SNP Single nucleotide polymorphism
SOFA Schedule of Fatigue and Anergia
ss Same-sex
UTR Untranslated region
xiv Characterising the Relationship between Fatigue and Depression
List of Publications
Corfield, E. C., Martin, N. G., & Nyholt, D. R. (2016). Co-occurrence and
symptomatology of fatigue and depression. Comprehensive Psychiatry, 71, 1-10.
doi:10.1016/j.comppsych.2016.08.004
The content outlined in this paper relates to Chapter 3.
Corfield, E. C., Martin, N. G., & Nyholt, D. R. (In Press, accepted 20March 2017).
Familiality and heritability of fatigue in an Australian twin sample. Twin Research
and Human Genetics
The content outlined in this paper relates to Chapter 4.
Corfield, E. C., Yang Y.. Martin, N. G., & Nyholt, D. R. (In Press, accepted 4 April
2017). A continuum of genetic liability for minor and major depressive disorder.
Translational Psychiatry
The content outlined in this paper relates to Chapter 5.
Corfield, E. C., Martin, N. G., & Nyholt, D. R. (2016). Shared Genetic Factors in
the Co-Occurrence of Depression and Fatigue. Twin Research and Human Genetics,
1-9. doi:10.1017/thg.2016.79
The content outlined in this paper relates to Chapter 6.
Characterising the Relationship between Fatigue and Depression xv
List of Presentations
ORAL
Corfield, E. C., Martin, N. G., & Nyholt, D. R. (2016) Shared genetics of depression
and fatigue. Behavior Genetics Association forty-sixth Annual Meeting. Brisbane,
Australia.
POSTER
Corfield, E. C., Marshall-Gradisnik, S.M., Martin, N. G., & Nyholt, D. R. (2016)
Systematic evaluation of risk loci from candidate gene and genome-wide association
studies of fatigue. American Society of Human Genetics sixty-sixth Annual Meeting.
Vancouver, Canada.
Corfield, E. C., Martin, N. G., & Nyholt, D. R. (2015) Genetic heritability of minor
and major depressive disorder. GeneMappers eleventh sesquiennial Conference.
Perth, Australia.
This poster won best student poster presentation.
xvi Characterising the Relationship between Fatigue and Depression
Statement of Original Authorship
The work contained in this thesis has not been previously submitted to meet
requirements for an award at this or any other higher education institution. To the
best of my knowledge and belief, the thesis contains no material previously
published or written by another person except where due reference is made.
Signature:
Date: 28/08/2017
QUT Verified Signature
Characterising the Relationship between Fatigue and Depression xvii
Acknowledgements
Firstly, to my supervisor, Dale, thank you for your continuous support, patience, and
guidance throughout the last three years. My PhD experience has been incomparable
with my honours year and I doubt this dissertation would be complete without your
help. To Nick, thank you for providing feedback and comments on my manuscripts
and to everyone from level 7 of the Bancroft building at QIMR Berghofer and the
members of SGEL, both past and present, thank you for answering my plethora of
questions. Finally, to my family, thank you for your support throughout this journey.
Chapter 1: Introduction 1
Chapter 1: Introduction
1.1 BACKGROUND AND SIGNIFICANCE
Abnormal tiredness or fatigue are multidimensional symptoms, which exist on a
continuum (Wessely et al., 1997). The fatigue symptom spectrum ranges from non-
specific to essential; with disorders such as chronic fatigue syndrome (CFS)
requiring specified durations and severities before a diagnosis can be made (Griffith
& Zarrouf, 2008; Hadzi-Pavlovic et al., 2000). Little is known about the underlying
mechanisms of fatigue or CFS. Although, throughout the entire fatigue continuum
high levels of comorbidity with depression are observed. Furthermore, individuals
with medically unexplained fatigue have an increased risk of a lifetime major
depressive disorder (MDD) diagnosis compared to individuals who have never been
fatigued (Addington et al., 2001). The comorbidity between fatigue and depression
could be accountable to overlapping symptoms. With fatigue or loss of energy
representing the second highest depression symptom reported, in a community-based
MDD outpatient population (Zimmerman et al., 2015). However, a number of
prescribed antidepressants do not treat fatigue symptoms or result in fatigue and
drowsiness. After antidepressant treatment, unresolved or residual fatigue is highly
prevalent in partial responders and remitted MDD patients (Fava et al., 2014).
Additionally, the severity of a major depressive episode has been identified as an
independent predictor of residual fatigue in remitted or partially remitted MDD
patients (Chung et al., 2015). However, currently available antidepressant therapies
inadequately treat residual fatigue, which leads to higher functional impairment and
MDD relapse.
Considerable economic burden and reduction in quality of life are associated
with fatigue and depression. As a depression symptom, fatigue is associated with
higher health care utilisation, increased medication use, 10-20% greater annual
health care cost, and lower quality of life (Robinson et al., 2015). CFS has an
estimated annual cost, in the United States, of approximately 18.7 to 24.0 billion US
dollars (Jason et al., 2008). In 2010, MDD had an estimated annual cost, in the
United States, of approximately 210.5 billion US dollars (Greenberg et al., 2015);
meanwhile, minor depressive disorder (MiDD), a less severe form of MDD (Ayuso-
2 Chapter 1: Introduction
Mateos et al., 2010; Fils et al., 2010), has approximately two-thirds the annual cost
(Cuijpers et al., 2007). Therefore, fatigue, CFS, MDD, and MiDD represent
significant health burdens.
Heritability estimates indicate both fatigue and MDD have moderate genetic
contributions. Fatigue experienced for at least one month has an estimated
heritability of 18-51% (Schur et al., 2007; Sullivan et al., 2005). Similarly, CFS, in
females, has an estimated heritability of 51% (Schur et al., 2007). Meanwhile, MDD
has an estimated heritability of 37% (Sullivan et al., 2000). However, to date, the
heritability of MiDD has not been investigated. An underlying genetic relationship
has been implicated between fatigue and MDD by the identification of genetic
factors which contribute to the heritability of both traits. Numerous genetic
association analyses have been conducted for MDD while few have been conducted
for CFS. To date, no genes have been consistently associated with CFS. However, in
August 2016, the first risk loci robustly associated with MDD, in Europeans, were
reported.
1.2 PURPOSE
The overall objective of this project is to increase our knowledge and understanding
of the comorbidity and genetics of fatigue and depression. In silico phenotypic and
genotypic analysis will be conducted within this project, which will be divided into
four sections. Initially, a symptomatic analysis of fatigue and depression will be
conducted. Next, the familiality and heritability of fatigue and depression will be
characterised. The genetic relationship between fatigue and depression will then be
investigated. Finally, the molecular genetics of fatigue and CFS will be investigated.
The overall hypothesis of this project was that both fatigue and depression will have
a significant genetic contribution, with shared genetic effects driving the comorbidity
between the traits.
1.2.1 Aims
The symptomatic analysis aims to:
1. Determine if the depression symptom profiles of fatigued and non-fatigued
individuals are quantitatively different.
2. Determine if the fatigue symptom profiles of MDD, MiDD, and non-
depressed individuals are quantitatively different.
Chapter 1: Introduction 3
The familiality and heritability analysis aims to:
1. Assess the familiality of fatigue experienced over the past few weeks.
2. Identify the heritability of fatigue.
3. Assess the familiality of minor and major depression.
4. Investigate whether minor and major depression lie on a single genetic
continuum.
The genetic relationship analysis aims to:
1. Determine the genetic correlation between fatigue and depression (MDD
plus MiDD).
2. Determine if the heritability of fatigue is different in MDD compared to
MiDD and non-depressed controls.
3. Characterise the type of relationship between fatigue and depression.
The molecular genetic analysis aims to:
1. Characterise the genetic risk associated with fatigue and CFS.
1.2.2 Hypotheses
The hypotheses of the symptomatic analysis were that:
1. Fatigued individuals report an increased preponderance of depression
symptoms compared to non-fatigued individuals.
2. Depressed (MDD plus MiDD) individuals will report an increased
preponderance of fatigue symptoms compared to non-depressed
individuals; with MDD cases reporting more fatigue symptoms that MiDD
cases.
The hypotheses of the familiality and heritability analysis were that:
1. Fatigue experienced over the past few weeks has a familialial contribution,
with a significant additive genetic contribution.
2. Minor and major depression are familialial and lie on a single genetic
continuum.
The hypotheses of the genetic relationship analysis were that:
4 Chapter 1: Introduction
1. A non-causal genetic relationship, driven by a significant genetic
correlation, explains the association between fatigue and depression (MDD
plus MiDD); with fatigue exhibiting a similar genetic correlation with
MDD and MiDD cases.
The hypotheses of the molecular genetic analysis were that:
1. Genetic risk variants will be associated with fatigue and CFS, with similar
biological pathways contributing to the etiology of both traits.
1.3 THESIS OUTLINE
In order to address the project objectives and aims a literature review was initially
conducted, which is detailed in Chapter 2. The following five chapters comprise
articles which are in preparation, currently under review or have been published after
peer-review. To reduce repetition within the dissertation a separate methods chapter
has not been included, rather the methodology utilised and results obtained are
detailed within the appropriate chapter. For ease of reading the formatting of
individual manuscripts has been standardised within this thesis. As such, the
abstracts have been included as a single paragraph and the numbering of figures and
tables has been updated to contain the chapter number followed by the individual
figure or table number. Additionally, to reduce repetition, a single reference list,
containing all references utilised throughout this thesis, is provided at the end of the
dissertation rather than providing a reference list after each manuscript.
A prologue has been included before each of the results chapters to maintain
the logical flow of the dissertation. Chapter 3 details the findings of the symptomatic
analysis of fatigue and depression. The familiality and heritability analyses are
detailed in two separate chapters. Chapter 4 details the findings of the first two
aims—which relate to fatigue, while Chapter 5 details the findings of the last two
aims—which relate to MDD and MiDD. Chapter 6 details the findings of the
analysis investigating the genetic relationship between fatigue and depression.
Chapter 7 details the findings of the molecular genetics analysis of fatigue and CFS.
Finally, Chapter 8 contains a general discussion which briefly describes the main
findings of each study, illustrates the relationship between the results of the
individual chapters, and explains the implications and future directions of this
project.
Chapter 2: Literature Review 5
Chapter 2: Literature Review
Abnormal tiredness or fatigue is a highly prevalent symptom, which is difficult to
quantify, and associated with numerous medical diagnoses. As a diagnostic
symptom, fatigue has been significantly associated with a number of illnesses,
including mental health, gut/stomach problems, arthritis/rheumatism, heart
conditions, breathing difficulties, cancer, skin disease, back problems, diabetes, and
blood pressure (Williamson et al., 2005). Fatigue has an estimated prevalence of
25.6% in general practice, representing the main complaint in a quarter of all
consultations (Cullen et al., 2002). Symptoms of fatigue exist on a continuum and
can be classified into physical, mental, and emotional dimensions (Lewis & Wessely,
1992; Wessely et al., 1997). The fatigue symptom spectrum ranges from non-specific
to essential; with disorders such as CFS requiring specified severities and durations
before a diagnosis can be made (Griffith & Zarrouf, 2008; Hadzi-Pavlovic et al.,
2000).
2.1 FATIGUE CLASSIFICATIONS
Quantifiable difficulties associated with fatigue has resulted in classification based
on arbitrary durations and severities by either self-report or clinical evaluation
(David et al., 1990). Prolonged fatigue and chronic fatigue (CF) are classified as self-
reported fatigue experienced for at least one month and at least six months,
respectively (Fukuda et al., 1994). Classification of prolonged fatigue requires
persistent fatigue; however, relapsing fatigue is acceptable for classification of CF.
After clinical evaluation, individuals exhibiting unexplainable CF can be classified
with idiopathic chronic fatigue (ICF), CFS, or myalgic encephalomyelitis/chronic
fatigue syndrome (ME/CFS). Furthermore, individuals can be diagnosed with
myalgic encephalomyelitis (ME), which has a highly similar classification to both
CFS and ME/CFS; however, presentation with unexplained CF is not required.
Finally, systemic exertion intolerance disease (SEID) has been defined as a result of
the recent investigation into chronic fatigue syndrome or myalgic encephalomyelitis
classifications (Institute of Medicine (IOM). 2015).
6 Chapter 2: Literature Review
2.2 CLASSIFICATION OF CFS, ME/CFS, ME, AND SEID
CFS is a complex, multisystem disorder, of unknown aetiology. Attempts to describe
the underlying pathology of CFS have resulted in numerous nomenclature changes,
based on speculation of the disorders aetiology. Generally, the terms CFS and ME
are used interchangeably to describe the complex disorder. Furthermore, numerous
case definitions have been published based on aetiological research; with each
definition representing increased knowledge and understanding about the disorder. A
common feature of all the definitions is the requirement for diagnosis to occur based
on exclusion of any medical condition which explains the symptoms experienced by
the individual.
2.2.1 CFS Definition
In 1988, the first definition for CFS was published by the Centres for Disease
Control (CDC) (Holmes et al., 1988). The original case definition requires the
presence of new onset unexplained CF and either at least six symptom criteria and
two physical criteria, or at least eight of the symptom criteria. The eleven symptom
criteria are: mild fever or chills, sore throat, painful lymph nodes (cervical or
axillary), muscle weakness, muscle discomfort or myalgia, post-exertional fatigue (at
least 24 hours), headaches, migratory arthralgia, neuropsychologic complaints, sleep
disturbance, and acute onset (Holmes et al., 1988). The three physical symptom
criteria of the original CFS definition are: low grade fever, nonexudative pharyngitis,
and palpable or tender lymph nodes (cervical or axillary).
In 1994, the CDC released a revised version of the CFS definition, primarily to
provide a research framework and standardise the definition used worldwide (Fukuda
et al., 1994). The CDC definition requires the presence of unexplained CF and
concurrent occurrence of at least four of the eight diagnostic symptoms, which have
not predated the fatigue but have been persistent or relapsing for at least six months.
The diagnostic symptoms are: sore throat, tender lymph nodes (cervical or axillary),
headaches, cognitive difficulties, unrefreshing sleep, multijoint pain, muscle pain,
and post-exertional malaise (for at least 24 hours). ICF is diagnosed when an
individual exhibits unexplained CF but does not meet the diagnostic criteria for CFS.
Chapter 2: Literature Review 7
2.2.2 ME/CFS Definition
In 2003, the Canadian consensus criteria (CCC) for ME/CFS was published to aid
clinicians with the diagnosis and treatment of CFS (Carruthers et al., 2003). The key
distinguishing features of the CCC definition compared to the CDC definition are the
addition of extra compulsory symptoms and the grouping of symptoms based on
their region of pathogenesis. The CCC criteria has six symptom groups of which the
first 4 are compulsory for diagnosis (Carruthers et al., 2003). Furthermore, at least
two of the symptoms from category five and one symptom from two of the
subgroups of category six are required. The categorical definition of ME/CFS is:
1. Fatigue: CF which is unexplained and of a new onset, presenting with
both physical and mental fatigue.
2. Post-exertional malaise and/or fatigue: loss of physical and mental
stamina, post-exertional malaise, fatigue or pain, with a recovery period
exceeding 24 hours.
3. Sleep dysfunction: unrefreshing sleep or sleep quantity.
4. Pain: muscle or joint pain and headaches
5. Neurological/cognitive manifestations: confusion, impaired
concentration and short-term memory, disorientation, and perceptual and
sensory disturbances.
6. a) Autonomic manifestations: orthostatic intolerance, light-headedness,
extreme pallor, nausea and irritable bowel syndrome, bladder dysfunction,
palpitations with/without cardiac arrhythmias, and exertional dyspnea.
b) Neuroendocrine manifestations: loss of thermostatic stability,
intolerance to temperature change, weight change, loss of adaptability and
worsening of symptoms with stress.
c) Immune manifestation: Tender lymph nodes, recurrent sore throat,
recurrent flu-like symptoms, general malaise, and new sensitivities to food,
medication and/or chemicals.
Diagnosis with ME/CFS requires symptom persistence for at least 6 months.
The CCC definition also endorses the classification of ICF when individuals do not
meet the criteria for ME/CFS but have unexplained CF.
8 Chapter 2: Literature Review
2.2.3 ME Definition
In 2011, the international consensus criteria (ICC) for myalgic encephalomyelitis
(ME) was published (Carruthers et al., 2011). The CCC was used as the basis for
construction of the ICC. However, there are a number of differences between the
definitions. The ICC has increased stringency and enables diagnosis without
requiring 6 months of symptom persistence. Diagnosis with ME can be further
classified based on symptom severity, with mild cases exhibiting a 50% reduction in
activity, moderate cases mostly housebound, severe cases mostly bedridden, and very
severe cases totally bedridden, requiring help with basic functions (Carruthers et al.,
2011). The ICC definition separates the symptom criteria into four impairment types.
The first category is compulsory and at least one symptom must be exhibited from
each of the other three impairment types for diagnosis with ME (Carruthers et al.,
2011). The four categories used for diagnosis of ME are:
1. Post-exertional neuroimmune exhaustion: post-exertional physical
and/or cognitive fatigability, post-exertional symptom exacerbation, post-
exertional exhaustion, prolonged post-exertional recovery period, and a
low threshold of physical and mental fatigability.
2. Neurological impairments: neurocognitive impairments (cognitive
difficulties and short-term memory loss), pain (headaches, and non-
inflammatory pain), sleep disturbance, and neurosensory, perceptual and
motor disturbances.
3. Immune, gastro-intestinal, and genitourinary impairments: flu-like
symptoms, susceptibility to viral infections, gastro-intestinal tract,
genitourinary, and sensitivities to food medicine, or chemicals.
4. Energy production/transportation impairments: cardiovascular,
respiratory, loss of thermostatic stability, and intolerance of extremes in
temperatures.
Individuals can be diagnosed with typical ME if they do not meet the full
criteria for ME, but satisfy the post-exertional neuroimmune exhaustion criterion and
exhibit at least one symptom from the three impairment types (Carruthers et al.,
2011).
Chapter 2: Literature Review 9
2.2.4 SEID Definition
In 2015, the Institute of Medicine (IOM) published a definition for SEID (Institute of
Medicine (IOM). 2015). Evidence-based consensus of the investigating committee
was used to construct the report (Clayton, 2015). The definition for SEID requires
individuals to exhibit: CF accompanied by impairment, post-exertional malaise, and
unrefreshing sleep (Institute of Medicine (IOM). 2015). Furthermore, cognitive
impairment or orthostatic intolerance must also be exhibited.
2.2.5 Differences between CFS, ME/CFS, and ME
Comparison of patients diagnosed using the CCC and ICC definitions compared to
the CDC criteria has revealed a number of demographic and symptomatic differences
(Table 2.1) (Brown et al., 2013; Jason et al., 2012; Jason et al., 2013; Johnston et al.,
2014). However, the set of symptoms examined in each study varies, making
comparison difficult, particularly when some symptoms are split into varying
components in one study but not others. Differences in fatigue, sleep, pain,
neurological/neurocognitive, autonomic, neuroendocrine, and immune symptoms
have been identified between ME/CFS and CFS patients in all four independent
samples studied (Jason et al., 2012; Jason et al., 2013). Similarly, differences in post-
exertional malaise, neurological impairments, immune, gastrointestinal, and
genitourinary impairments, and energy production/transportation impairments have
been identified between ME and CFS patients (Brown et al., 2013). Indicating,
patients diagnosed with ME or ME/CFS represent a subset of CFS patients, which
experience more severe clinical symptoms.
10 Chapter 2: Literature Review
Table 2.1. Comparison of patients meeting the Canadian consensus criteria and International consensus criteria compared to those diagnosed using the Centres for Disease
Control criteria for differences in demographics, heart rate, cognitive measures, and responses to the 36-item Short-form health survey and World Health Organisation
disability adjustment schedule 2.0 items.
ME/CFS vs. CFS ME vs. CFS
Cohort Chicago DePaul Solve CFS Biobank Newcastle Chicago South East Queensland
Study Jason et al. (2012) Jason et al. (2013) Brown et al. (2013) Johnston et al. (2014)
Demographic differences
Sex NS NS NS NS NS Increased number of females
Current psychiatric diagnosis Higher NS - NS Higher - Disability payment NS NS More on disability NS NS -
Higher education NS NS NS Higher NS -
Heart rate (HR) and cognition tests
HR lying down Higher - - - Higher -
HR standing 2 minutes Higher - - - Higher -
HR standing 10 minutes Higher - - - Higher - Cognition test trail making A Higher completion time - - - NS -
Cognition test trail making B Higher completion time - - - NS -
Short-form health survey
Physical functioning Worse Worse Worse Worse Worse Worse
Role physical NS NS NS NS NS Worse
Bodily pain Worse Worse Worse Worse Worse Worse General health Worse NS Worse Worse NS NS
Social functioning Worse NS Worse Worse Worse Worse
Mental health Worse NS Worse NS NS NS Role emotional Worse NS NS NS NS NS
Vitality Worse NS Worse NS Worse NS
World Health Organisation disability adjustment schedule 2.0
Cognition - - - - - Worse
Mobility - - - - - Worse
Self-care - - - - - Worse Getting along - - - - - Worse
Life activities - - - - - Worse
Participation - - - - - Worse
CFS: chronic fatigue syndrome: ME/CFS: myalgic encephalomyelitis/chronic fatigue syndrome; ME: myalgic encephalomyelitis; NS: not significant
Chapter 2: Literature Review 11
2.3 EPIDEMIOLOGY OF FATIGUE
Epidemiological studies investigating prolonged fatigue, CF, ICF, CFS, ME/CFS,
and ME have been conducted in varying populations, resulting in a range of
population prevalence estimates and overall incidence rates due to differing
ethnicities and study designs (Table 2.2). The population prevalence estimations
range from: 6.16-28.00% for prolonged fatigue (EvengÅRd et al., 2005; Hamaguchi
et al., 2011; Jason et al., 1999; Kim et al., 2005; Njoku et al., 2007), 2.00-12.20% for
CF (Bierl et al., 2004; Cho et al., 2009; EvengÅRd et al., 2005; Friedberg et al.,
2015; Hamaguchi et al., 2011; Jason et al., 1995; Jason et al., 1999; Kim et al., 2005;
Loge et al., 1998; Njoku et al., 2007; Patel et al., 2005; Steele et al., 1998; Wessely et
al., 1995; Wessely et al., 1997; Wong & Fielding, 2010), 1.00-9.00% for ICF
(Hamaguchi et al., 2011; Kim et al., 2005; Wessely et al., 1997), 0.07-2.60% for CFS
(Cho et al., 2009; Hamaguchi et al., 2011; Jason et al., 1999; Kawakami et al., 1998;
Kim et al., 2005; Lindal et al., 2002; Nacul et al., 2011; Njoku et al., 2007; Reyes et
al., 2003; Vincent et al., 2012; Wessely et al., 1995; Wessely et al., 1997), and 0.11
for ME/CFS (Nacul et al., 2011). Population prevalence estimates for ME and SEID
have not been established due to the recent publication of the diagnostic definitions.
Investigations into the variation observed in population prevalence estimations for
CFS have established the differences may be partially explained by differences in
study design (Johnston et al., 2013). Higher population prevalence estimations were
obtained for CFS patients diagnosed by self-report (3.28%) compared to clinical
assessment (0.76%). Similarly, studies using community based cohorts (0.87%)
reported lower estimations than investigations conducted within primary care settings
(1.72%) (Johnston et al., 2013).
12 Chapter 2: Literature Review
Table 2.2. Population prevalence estimates for prolonged fatigue, chronic fatigue, idiopathic chronic
fatigue, chronic fatigue syndrome, and myalgic encephalomyelitis/chronic fatigue syndrome.
Prevalence (%) Population Age range Study
Prolonged fatigue
7.70 Chicago, USA ≥ 18 Jason et al. (1999) 21.10 South Korea ≥ 18 Kim et al. (2005)
6.16 Nigeria ≥ 18 Njoku et al. (2007)
28.00 Japan 20-78 Hamaguchi et al. (2011) 12.33 Sweden 42-64 EvengÅRd et al. (2005)
Chronic fatigue
5.00 Chicago, USA ≥ 18 Jason et al. (1995) 10.78 Southern England, UK 18-45 Wessely et al. (1995)
11.30 Southern England, UK 18-45 Wessely et al. (1997)
11.40 Norway 19-80 Loge et al. (1998) 2.00 San Francisco, USA ≥ 18 Steele et al. (1998)
4.20 Chicago, USA ≥ 18 Jason et al. (1999)
12.19 USA 18-69 Bierl et al. (2004)
8.40 South Korea ≥ 18 Kim et al. (2005)
12.10 India 18-50 Patel et al. (2005)
9.48 Nigeria ≥ 18 Njoku et al. (2007) 12.20 São Paulo, Brazil 18-45 Cho et al. (2009)
10.30 London, England, UK 18-45 Cho et al. (2009)
10.70 Hong Kong ≥ 18 Wong and Fielding (2010) 7.30 Japan 20-78 Hamaguchi et al. (2011)
8.26 Sweden 42-64 EvengÅRd et al. (2005) 3.70 USA ≥ 18 Friedberg et al. (2015)
5.20 Ukraine ≥ 18 Friedberg et al. (2015)
Idiopathic chronic fatigue
9.00 Southern England, UK 18-45 Wessely et al. (1997)
1.00 South Korea ≥ 18 Kim et al. (2005)
1.30 Japan 20-78 Hamaguchi et al. (2011) Chronic fatigue syndrome
1.81 Southern England, UK 18-45 Wessely et al. (1995)
2.60 Southern England, UK 18-45 Wessely et al. (1997) 1.50 Japan ≥ 18 Kawakami et al. (1998)
0.42 Chicago, USA ≥ 18 Jason et al. (1999)
2.20 Iceland 19-75 Lindal et al. (2002) 2.35 Wichita, USA 18-69 Reyes et al. (2003)
0.60 South Korea ≥ 18 Kim et al. (2005)
0.68 Nigeria ≥ 18 Njoku et al. (2007) 1.60 São Paulo, Brazil 18-45 Cho et al. (2009)
2.10 England, UK 18-45 Cho et al. (2009)
1.00 Japan 20-78 Hamaguchi et al. (2011) 0.19 England, UK 18-64 Nacul et al. (2011)
0.07 Minnesota, USA ≥ 18 Vincent et al. (2012)
Myalgic encephalomyelitis/chronic fatigue syndrome
0.11 England, UK 18-64 Nacul et al. (2011)
The combined overall incidence rate of CFS and ME/CFS has been estimated
at 0.03-0.09% within primary care, hospitals, and outpatient clinics (Bakken et al.,
2014; Nacul et al., 2011). Overall incidence rate is higher in females, at 0.04-0.14%,
compared to males, at 0.01-0.04%, which concurs with the predominance of females
diagnosed with CFS, who represent ≥ 75% of all cases (Bakken et al., 2014; Capelli
et al., 2015; Prins et al., 2006). Onset of CFS predominantly occurs between twenty
and fifty years of age (Collin et al., 2017). The highest incidence of CFS occurs in
the age ranges of 30-34 and 35-39 at 0.04% (Bakken et al., 2014), with no significant
difference in age of onset between males and females (Capelli et al., 2015).
Understanding the change in incidence with age could be insightful in terms of
Chapter 2: Literature Review 13
hormonal changes occurring at particular life stages, which may contribute to the
development of CFS.
2.4 PATHOPHYSIOLOGY OF CFS, ME/CFS AND ME
Aetiological investigations into CFS have thus far been unsuccessful, despite the
substantial number of proposed causes. Dysfunction of the immune, endocrine,
nervous, cardiovascular, and digestive systems have been implicated in the
pathophysiology of CFS.
2.4.1 Infection and Immune Dysfunction
Numerous viral and bacterial pathogens have been investigated in relation to CFS
onset. Seasonality investigations have shown onset of CFS and ICF is higher in
winter (in the northern hemisphere), with a 5-fold increase observed in January
(Jason et al., 2001; Zhang et al., 2000). The implication of increased case onset in
winter provides some evidence for a pathogenic cause. However, despite persistent
attempts and numerous links to CFS onset a causal pathogen has not consistently
been identified (Table 2.3) (Armstrong et al., 2014; Devanur & Kerr, 2006; Ortega-
Hernandez & Shoenfeld, 2009).
The immune dysfunction often exhibited by CFS patients has been frequently
studied, with numerous investigations of B cells, T cells, natural killer cells,
interferons, interleukins, and immunoglobulins. However, the conflicting results
between studies raise questions about the clinical significance of the findings (Lyall
et al., 2003; Natelson et al., 2002). Particularly considering none of the abnormalities
have been identified as a biomarker. Although, the multitude of abnormalities
identified implies the immune system has a pivotal role. The subsets of immune
dysfunctions identified could imply that different subgroups of CFS cases are
associated with specific abnormalities. However, the specific role of the immune
system will be indeterminable without larger sample sizes and reduced heterogeneity
within study cohorts.
14 Chapter 2: Literature Review
Table 2.3. List of the pathogens investigated as potential triggering agents in chronic fatigue
syndrome onset.
Pathogen Associated illness Studies
RNA virus
Enteroviruses (Coxsackie A & B,
echoviruses, polioviruses)
Acute respiratory and gastrointestinal infections,
aseptic meningitis, and polio
Chia and Chia (2003); Chia (2005); Clements et al. (1995); Dalakas (2003); Gow et al. (1991); Lane et al. (2003); Ortega-
Hernandez and Shoenfeld (2009)
Dengue virus Dengue fever Seet et al. (2007) Hepatitis C virus Hepatitis C Chia and Chia (2003); Ortega-Hernandez and Shoenfeld (2009)
Ross river virus Ross river fever Hickie et al. (2006)
Murine Leukaemia virus-related virus
No evidence that MLRV can infect humans
Erlwein et al. (2010); Knox et al. (2011); Lombardi et al. (2009); Mikovits et al. (2010)
DNA virus
Cytomegalovirus Infectious mononucleosis Beqaj et al. (2008); Chia and Chia (2003)
Epstein-Barr virus Infectious mononucleosis
Glaser et al. (2005); Holmes et al. (1987); Ikuta et al. (2003);
Jones et al. (1991); Katz (2002); Kawai and Kawai (1992); Koo
(1989); Lerner et al. (2004); Ortega-Hernandez and Shoenfeld
(2009); Vernon et al. (2006); White et al. (1998); White et al.
(2001)
Herpes simplex virus Herpes simplex Bond (1993)
Hepatitis B virus Hepatitis B Canadian Laboratory Center for Disease Control (LCDC).
(1993); Nancy and Shoenfeld (2008)
Human herpesvirus-6 Roseola Ablashi et al. (2000); Chapenko et al. (2006); Chia and Chia (2003); Di Luca et al. (1995); Komaroff (2006); Lum et al.
(2014); Nicolson et al. (2003) Human herpesvirus-7 Roseola Chapenko et al. (2006); Di Luca et al. (1995)
Parvovirus B19 Fifth disease
Jacobson et al. (1997); Kerr et al. (2001); Kerr et al. (2002); Kerr
and Mattey (2008); Matano et al. (2003); Ortega-Hernandez and Shoenfeld (2009); Seishima et al. (2008)
Varicella zoster virus Chickenpox Shapiro (2009)
Bacteria
Chlamydia pneumonia Pneumonia Chia and Chia (1999); Chia and Chia (2003); Nicolson et al.
(2003)
Coxiella burnetii Q fever Arashima et al. (2004); Hickie et al. (2006); Ikuta et al. (2003); Ledina et al. (2007); Strauss et al. (2012); Wildman et al. (2002)
Mycoplasma spp. Pneumonia, fever Choppa et al. (1998); Endresen (2003); Nicolson et al. (2003);
Nijs et al. (2002); Vernon et al. (2003); Vojdani et al. (1998)
2.4.2 Endocrine and Metabolic Dysfunction
Endocrine-metabolic dysfunction is one of the most consistent findings in CFS
aetiological research. Particularly, the presence of alterations in the hypothalamic-
pituitary-adrenal (HPA) axis (Cleare, 2004; Sorenson & Jason, 2013). Dysfunctions
identified in the HPA axis of CFS patients includes: mild hypocortisolism, attenuated
diural variation and enhanced corticosteroid-induced negative feedback (Tomas et
al., 2013). The abnormalities in the HPA axis also affect the serotonergic and
noradrenergic pathways (Armstrong et al., 2014). However, the findings suggest the
alterations in the HPA axis are not a predisposing factor to CFS and occur after onset
(Cleare, 2004).
Energy metabolism has also been implicated in the pathophysiology of CFS,
with evidence suggesting mitochondrial dysfunction plays a key role (Booth et al.,
2012; Myhill et al., 2009; Vernon et al., 2006). Although recent investigations have
established mitochondrial DNA mutations are unlikely to be associated with CFS
aetiology (Billing-Ross et al., 2016; Schoeman et al., 2017). However, investigation
Chapter 2: Literature Review 15
of mitochondrial dysfunction in CFS patients has identified decreased levels of
ubiquinone and increased levels of lipid peroxidation in peripheral blood
mononuclear cells compared to controls (Castro-Marrero et al., 2013). Furthermore,
CFS cases exhibit lower mitochondrial ATP production than healthy controls. The
level of effect on mitochondrial function induced by the differing levels of
ubiquinones and lipid peroxidation is unknown. However, mitochondrial function
and oxidative stress are obviously affected which can be observed by the lower ATP
production and increased levels of isoprostanes in CFS cases compared to controls
(Armstrong et al., 2014; Castro-Marrero et al., 2013; Kennedy et al., 2005). It is
unknown if these dysfunctions are predisposing factors or occur after CFS onset.
2.4.3 Cardiovascular and Neurologic Dysfunction
Abnormalities associated with the nervous and cardiovascular system have been
identified in CFS cases. Symptoms of cardiac dysfunction have been frequently
identified in CFS cases (Miwa & Fujita, 2009b). A small left ventricle with a low
cardiac output (small heart) has been identified in a subset of CFS cases (Miwa &
Fujita, 2008, 2009a, 2009b). Furthermore, the small hearts have been associated with
orthostatic intolerance in CFS cases (Miwa & Fujita, 2011). Investigations into the
type of the orthostatic intolerance exhibited by CFS cases revealed increased rates of
neurally mediated hypotension, postural orthostatic tachycardia syndrome, and
orthostatic hypocapnia compared to controls (Bou-Holaigah et al., 1995; Hoad et al.,
2008; Natelson et al., 2007).
Brain volume abnormalities have been identified in CFS patients, with
decreased grey matter, increased white matter, and brainstem dysfunctions observed
using magnetic resonance imaging scans (Barnden et al., 2011; Barnden et al., 2015;
de Lange et al., 2005; Natelson et al., 1993; Okada et al., 2004). However, the role
these abnormalities play in CFS pathophysiology or symptomatology is unknown.
2.4.4 Psychiatric Disorders
The possibility that CFS is a psychiatric disorder or a subtype of MDD is
controversial leading to confusion and unwillingness to diagnose patients with CFS.
The high levels of MDD observed in CFS cohorts fuels the debate for classification
as a psychiatric disorder. However, there are key differences between MDD and
CFS. In particular, the diagnostic criteria for CFS which are not associated with
16 Chapter 2: Literature Review
depression and the differences in depressive symptoms experienced (Afari &
Buchwald, 2003).
2.4.5 Genetics
Recently, studies have started to investigate the genetic contribution of CFS (Narita
et al., 2003; Ortega-Hernandez et al., 2009; Sommerfeldt et al., 2011). However,
conflicting results have been obtained from gene expression analyses (Grans et al.,
2005; Kaushik et al., 2005; Powell et al., 2003). In 2006, the Critical Assessment of
Microarray Data Analysis conference challenge dataset (Wichita clinical study) was
a cohort of CFS (N = 68) and ICF (N = 81) cases and non-fatigued (N = 73) controls
(Nisenbaum et al., 2003; Reeves et al., 2005). Multiple biostatistical approaches,
ranging from gene expression and association analyses to gene-gene interactions,
have been utilised to analyse the data available within the Wichita clinical study.
Analyses using the Wichita clinical study have identified genes of potential
functional significance for CFS (Chung et al., 2007; Goertzel et al., 2006; Lin &
Huang, 2008; Lin & Hsu, 2009; Smith et al., 2006; Smith et al., 2009).
2.5 HERITABILITY OF FATIGUE
Limited studies have been conducted on the genetic basis of fatigue, prolonged
fatigue, CF, ICF, or CFS. However, a number of familial studies have indicated that
genetic and/or environmental factors are associated with predisposition to CF and
CFS within first and second degree relatives (Albright et al., 2011; Bell et al., 1991;
Buchwald et al., 2001; Levine et al., 1998; van de Putte et al., 2006; Walsh et al.,
2001). The importance of both genetic and common environmental factors was
revealed by the shared symptom complex identified between adolescents with CFS
and their mothers, which was not exhibited by the fathers (van de Putte et al., 2006).
Convincing evidence for the contribution of genetic factors to the development or
onset of CFS was provided in a population based study investigating familial
clustering in first, second, and third degree relatives (Albright et al., 2011).
Heritability predictions have been conducted in a number of twin studies,
which utilised structural equation modelling to investigate the additive genetic
factors (A), common environmental factors (C), and unique environmental factors
(E) involved in the heritability of varying levels of fatigue (Table 2.4) (Ball et al.,
2010b; Buchwald et al., 2001; Farmer et al., 1999; Schur et al., 2007; Sullivan et al.,
Chapter 2: Literature Review 17
2003; Sullivan et al., 2005). Interestingly, males are predicted to have a higher
percentage of genetic components associated with abnormal tiredness, prolonged
fatigue, and CF than females (Schur et al., 2007; Sullivan et al., 2005). Fatigue
severity (a continuous measure of the 11 core fatigue and 2 muscle pain items of the
Chalder Fatigue Questionnaire (Chalder et al., 1993)) has an estimated heritability of
30% (Ball et al., 2010b). The heritability of interfering fatigue (tiredness or fatigue
experienced for at least five days) is estimated at 6% in males and 26% in females
(Sullivan et al., 2003). Short-duration fatigue (fatigue experienced for at least one
week) has an estimated heritability of 42% (Farmer et al., 1999). The heritability of
abnormal tiredness is estimated at 30% in males and 26% in females (Sullivan et al.,
2005). Abnormal fatigue (assessed by the 11 core fatigue items of the Chalder
Fatigue Questionnaire) has an estimated heritability of 39% (Ball et al., 2010b). The
heritability of prolonged fatigue is estimated to range from 34-51% in males and 18-
27% in females (Buchwald et al., 2001; Schur et al., 2007; Sullivan et al., 2005).
Additionally, the heritability of prolonged fatigue, in a cohort of males and females,
has been estimated at 54% (Farmer et al., 1999). Furthermore, the heritability of CF
is estimated to range from 34-51% and 30-47% in males and 12-32% in females
(Buchwald et al., 2001; Schur et al., 2007; Sullivan et al., 2005). Finally, within
females ICF and CFS have both been estimated to have 51% heritability (Buchwald
et al., 2001; Schur et al., 2007).
18 Chapter 2: Literature Review
Table 2.4. Heritability estimates (and their 95% confidence intervals) of the unique additive genetic
factors (A), common environmental factors (C), and unique environmental factors (E) contributing to
fatigue severity, interfering fatigue, short-duration fatigue, abnormal tiredness, abnormal fatigue,
prolonged fatigue, chronic fatigue, idiopathic chronic fatigue, and chronic fatigue syndrome.
Population Sex
Number
of twin
pairs
(MZ / DZ)
Mean age ±
standard
deviation
(Age range)
A (%) C (%) E (%) Study
Fatigue severity
Sri Lanka F & M 816 / 1,080 (≥ 15) 30 (24-35) 0 (0-0) 70 (65-76) Ball et al. (2010b)
Interfering fatigue USA M 1,299 /
1,964
Case: 34.9 ± 9.3
Control: 35.1 ± 9.2
6 (0-46) 21 (0-25) 73 (54-90) Sullivan et al. (2003)
F 26 (0-44) 1 (0-30) 73 (56-92)
Short-duration fatigue South Wales M & F 278 / 378 (5-17) 42 38 20 Farmer et al. (1999)
Abnormal tiredness
Sweden M 3,229 / 8,824
(42-64) 30 (11-40) 0 (0-14) 70 (60-80) Sullivan et al. (2005)
F 26 (8-33) 0 (0-14) 74 (67-82)
Abnormal fatigue
Sri Lanka M & F 816 / 1,080 (≥ 15) 39 (29-49) 0 (0-0) 61 (51-71) Ball et al. (2010b) Prolonged fatigue
South Wales M & F 278 / 378 (5-17) 54 19 26 Farmer et al. (1999)
Sweden M 3,229 / 8,824
(42-64) 34 (3-45) 0 (0-25) 66 (55-79) Sullivan et al. (2005)
F 27 (6-35) 0 (0-16) 73 (65-82)
USA M 1,042 / 828
32.4 ± 14.7
(18-90)
51 (13-69) 0 (0-33) 49 (31-71) Schur et al. (2007)
F 18 (0-54) 23 (0-48) 59 (46-74) Chronic fatigue
Sweden M 3,229 /
8,824 46
30 (2-44) 0 (0-23) 70 (56-86) Sullivan et al. (2005)
F 32 (11-41) 0 (0-16) 68 (59-78) USA M
1,042 / 828 (42-64) 47 (0-68) 0 (0-39) 52 (32-79) Schur et al. (2007)
F 12 (0-48) 26 (0-48) 62 (47-78)
USA F 106 / 40 32.4 ± 14.7
(18-90) 19 (0-56) 69 (32-89) 12 (7-19) Buchwald et al. (2001)
Idiopathic chronic fatigue USA F 77 / 22 46 51 (7-96) 42 (0-85) 8 (4-13) Buchwald et al. (2001)
Chronic fatigue syndrome
USA F 648 / 258 32.4 ± 14.7
(18-90) 51 (0-82) 12 (0-72) 36 (18-65) Schur et al. (2007)
M: male; F: female; M & F: male and female.
2.6 MOLECULAR GENETICS OF FATIGUE
Few studies have investigated the molecular genetics of fatigue associated with
cancer, hepatitis C, multiple sclerosis, and hypothyroidism (Landmark-Hoyvik et al.,
2010). However, the majority of research conducted has investigated the molecular
genetics of CFS (and occasionally ICF). Candidate gene association (CGA) studies,
genome-wide association (GWA) analyses, various bioinformatic data-mining
approaches, and genetic interaction investigations have been conducted to identify
the molecular genetic contribution of CFS.
CFS CGA studies have implicated genotypic associations between CFS and
ADRB2, CHRM3, CHRNA2, CHRNA3, CHRNA4, CHRNB1, CHRNB4, CHRNE,
COMT, DCP1, HLA-DQA1, SLC6A4, TRPC2, TRPC4, TRPC6, TRPM3, TRPM3,
and TRPM8 (Table 2.5) (Marshall-Gradisnik et al., 2016a; Marshall-Gradisnik et al.,
Chapter 2: Literature Review 19
2016b; Narita et al., 2003; Smith et al., 2005; Sommerfeldt et al., 2011; Vladutiu &
Natelson, 2004). Similarly, CFS CGA haplotypic analyses have identified an
association between CFS and four DRB1 and RAGE haplotypes (Table 2.5) (Carlo-
Stella et al., 2009). Finally, CFS CGA have implicated allelic associations between
CFS and BMP2K, CHRM1, CHRM2, CHRM3, CHRM5, CHRNA2, CHRNA3,
CHRNA4, CHRNA5, CHRNA9, CHRNA10, CHRNB1, CHRNB4, CHRND, CHRNE,
CHRNG, DISC1, EIF3A, FAM126B, HTR2A, IL6ST, IL-17F, INFG., METTL3,
NR3C1, SORL1, TCF3, TNF, TRPA1, TRPC4, TRPC2, TRPC4, TRPC6, TRPM3,
TRPM4, TRPM8, TRPV2, TRPV3, UBTF, and PEX16 (Table 2.5) (Carlo-Stella et al.,
2006; Fukuda et al., 2010; Marshall-Gradisnik et al., 2015a; Marshall-Gradisnik et
al., 2016a; Marshall-Gradisnik et al., 2016b; Marshall-Gradisnik et al., 2015b;
Metzger et al., 2008; Rajeevan et al., 2007; Shimosako & Kerr, 2014; Smith et al.,
2008).
To date, three CFS GWA studies have been conducted—containing very small
sample sizes. No genome-wide significant SNP loci were identified from the first
two CFS GWA analyses (Rajeevan et al., 2015; Smith et al., 2011). Meanwhile, the
latest CFS GWA study identified 92 SNP loci reaching the studies Bonferroni
adjusted genome-wide threshold of 7.5 × 10-8 (Table 2.6) (Schlauch et al., 2016).
However, methodological concerns associated with this study raise questions about
the accuracy of these associations.
20 Chapter 2: Literature Review
Table 2.5. Candidate genes and implicated single nucleotide polymorphisms associated with chronic
fatigue syndrome.
Gene
symbol Variant
Cases /
Controls RA OR (95% CI) p-value Study
Genotypic associations
SLC6A4 5-HTTLPR 78 / 50
long and
extra-
long variants
3.04 (1.36-4.65) 0.0310 Narita et al. (2003)
DCP1 ACE 59 / 44 I allele 5.55 (1.75-17.52) 0.0200 Vladutiu and Natelson
(2004) HLA-DQ A1 49 / 102 01 allele 1.93 (1.20-3.30) 0.0080 Smith et al. (2005)
ADRB2 rs1042714 53 / 33 GG 2.48 (1.01-6.14) 0.0440 Sommerfeldt et al. (2011)
COMT rs4680 GG &
AG 2.00 (1.01-3.96) 0.0460
CHRNA2 rs891398 39 / 30 CC 11.39 (1.38-94.16) 0.0069 Marshall-Gradisnik et al.
(2016a) rs2741343 CC 11.39 (1.38-94.16) 0.0069
CHRNA3 rs12914385 TT 6.22 (1.27-30.44) 0.0136
CHRNB4 rs12441088 TT 3.57 (1.31-9.73) 0.0113
CHRNE rs33970119 GG 4.36 (1.05-18.22) 0.0328 TRPC2 rs7108612 GT 4.06 (1.18-13.96) 0.0205
TRPC4 rs655207 GG 6.22 (1.27-30.44) 0.0136
rs1570612 GG 3.81 (1.35-10.71) 0.0095 rs2985167 AA 4.21 (1.41-12.56) 0.0079
TRPM3 rs1106948 TT 4.06 (1.18-13.96) 0.0205 rs1891301 TT 3.64 (1.05-12.57) 0.0343
rs6560200 CC 5.63 (1.45-21.83) 0.0076
rs11142822 GG 5.14 (1.25-21.13) 0.0154 rs12350232 TT 3.13 (0.98-9.94) 0.0479
TRPM8 rs11563204 GA 7.19 (2.27-22.76) 0.0004
rs17865678 AG 3.56 (1.27-9.94) 0.0135 CHRM3 rs1867264 11 / 11 TA 7.11 (1.09-46.44) 0.0330 Marshall-Gradisnik et al.
(2016b) rs6688537 CA 7.11 (1.09-46.44) 0.0330
CHRNA4 rs11698563 CC 12.00 (1.12-128.84) 0.0221 CHRNB1 rs2302767 TT 17.50 (1.60-191.90) 0.0078
rs3829603 CC 26.67 (2.31-308.01) 0.0024
rs4151134 TT 17.50 (1.60-191.90) 0.0078 rs7210231 CA 7.88 (1.10-56.12) 0.0301
TRPC6 rs10791504 GG 7.88 (1.10-56.12) 0.0301
TRPM3 rs7038646 AG 7.88 (1.10-56.12) 0.0301 Haplotypic associations
DRB1
RAGE
374 75 / 141
04
T 2.70 (1.02-7.18) 0.0400 Carlo-Stella et al. (2009)
DRB1
RAGE
374
09
T NC 0.0040
DRB1 RAGE
374
11 T
2.27 (1.05-4.89) 0.0390
DRB1
RAGE
374
13
A 8.41 (1.52-46.65) 0.0150
Allelic associations
IFNG rs2430561 47 / 140 T 1.43 (0.89-2.28) 0.0440 Carlo-Stella et al. (2006)
TNF rs1799724 80 / 224 T 1.80 (1.20-2.72) 0.0040 NR3C1 rs6188 40 / 55 C 1.76 (0.83-3.73) 0.0383 Rajeevan et al. (2007)
rs852977 A 1.76 (0.83-3.73) 0.0365
rs860458 G 2.10 (0.87-5.07) 0.0180 rs1866388 A 1.85 (0.87-3.93) 0.0335
rs2918419 T 2.10 (0.87-5.07) 0.0164
HTR2A rs6311 40 / 42 A 2.52 (1.00-6.30) 0.0065 Smith et al. (2008) rs6313 T 2.32 (0.93-5.78) 0.0150
rs1923884 C 2.51 (0.95-6.67) 0.0100
IL-17F rs763780 89 / 56 T 4.07 (1.70-9.71) 0.0018 Metzger et al. (2008) DISC1 rs821616 155 / 502 T 1.50 (1.02-2.19) 0.0370 Fukuda et al. (2010)
BMP2K rs1426139 108 / 68 A 1.10 (0.28-4.29) 0.0091 Shimosako and Kerr (2014)
rs3775516 G 1.11 (0.28-4.35) 0.0025 EIF3A rs10787901 A 1.22 (0.67-2.25) < 0.0001
FAM126B rs11895568 G NC 0.0110
IL6ST rs1373998 T 1.55 (0.54-4.41) 0.0130 METTL3 rs3752411 A 2.40 (0.76-7.62) 0.0310
PEX16 rs3802758 C 3.50 (1.51-8.11) < 0.0001
SORL1 rs3737529 T 7.78 (0.40-152.39) 0.0280 TCF3 rs1860661 G 6.16 (1.80-21.13) < 0.0001
UBTF rs2071167 A 1.89 (0.92-2.25) 0.0240
See Table 2.5 footnotes on page 22.
Chapter 2: Literature Review 21
Table 2.5. Continued Candidate genes and implicated single nucleotide polymorphisms associated
with chronic fatigue syndrome.
Gene
symbol Variant
Cases /
Controls RA OR (95% CI) p-value Study
TRPA1 rs2383844 115 / 90 G 1.54 (1.04-2.29) 0.0400 Marshall-Gradisnik et al. (2015b) rs4738202 A 1.73 (1.12-2.65) 0.0180
TRPC4 rs655207 G 1.66 (1.11-2.46) 0.0180
rs6650469 T 1.66 (1.12-2.48) 0.0160 TRPM3 rs1160742 A 1.78 (1.19-2.66) 0.0080
rs1328153 C 1.99 (1.18-3.35) 0.0130
rs1504401 C 1.88 (1.06-3.36) 0.0410 rs3763619 A 1.70 (1.13-2.56) 0.0140
rs4454352 C 1.99 (1.18-3.35) 0.0130
rs7865858 A 1.65 (1.10-2.48) 0.0210 rs10115622 C 1.53 (1.02-2.29) 0.0500
rs11142508 C 1.89 (1.25-2.85) 0.0040
rs12682832 A 1.93 (1.27-2.91) 0.0030
CHRM3 rs589962 115 / 90 T 2.02 (1.32-3.09) 0.0035 Marshall-Gradisnik et al.
(2015a) rs726169 A 1.70 (1.12-2.56) 0.0235
rs1072320 G 2.12 (1.33-3.39) 0.0037 rs4463655 C 1.97 (1.32-2.96) 0.0028
rs6429157 G 1.59 (1.07-2.35) 0.0375
rs6661621 C 2.10 (1.30-3.39) 0.0054 rs6669810 C 1.66 (1.12-2.45) 0.0236
rs7520974 A 1.71 (1.15-2.53) 0.0167 rs7543259 A 2.07 (1.30-3.31) 0.0051
CHRNA10 rs2672211 C 1.85 (1.20-2.86) 0.0107
rs2672214 C 1.87 (1.21-2.88) 0.0108 rs2741862 C 1.77 (1.10-2.84) 0.0304
rs2741868 T 1.85 (1.20-2.86) 0.0119
rs2741870 G 1.83 (1.19-2.82) 0.0128 CHRNA2 rs2565048 T 2.18 (1.24-3.86) 0.0140
CHRNA5 rs951266 T 1.82 (1.19-2.79) 0.0115
rs7180002 T 1.64 (1.07-2.49) 0.0368 CHRM1 rs2075748 39 / 30 A 2.79 (1.05-7.43) 0.0369 Marshall-Gradisnik et al.
(2016a) rs11823728 C 3.45 (1.03-11.55) 0.0394
CHRM3 rs4620530 T 2.11 (1.04-4.28) 0.0381 CHRNA2 rs891398 C 2.35 (1.17-4.70) 0.0168
rs2741343 C 2.29 (1.15-4.59) 0.0186
CHRNA3 rs2869546 T 2.29 (1.13-4.66) 0.0217 rs3743074 T 2.08 (1.03-4.23) 0.0410
rs3743075 G 2.08 (1.03-4.23) 0.0410
rs4243084 G 2.12 (1.03-4.39) 0.0403 rs12914385 T 2.40 (1.17-4.92) 0.0153
CHRNA5 rs951266 T 2.22 (1.07-4.60) 0.0332
rs7180002 T 2.11 (1.02-4.36) 0.0433 CHRNB4 rs12441088 T 2.79 (1.28-6.09) 0.0090
CHRNE rs33970119 G 3.85 (0.97-15.18) 0.0414
TRPC2 rs6578398 A 2.36 (1.06-5.27) 0.0338 TRPM3 rs1106948 T 2.44 (1.22-4.87) 0.0107
rs1891301 T 2.19 (1.10-4.36) 0.0241
rs6560200 C 2.48 (1.24-4.95) 0.0100 rs11142822 G 4.41 (1.14-17.09) 0.0212
rs12350232 T 2.27 (1.14-4.52) 0.0183
TRPM8 rs6758653 G 2.54 (1.23-5.25) 0.0108 rs11563204 A 4.17 (1.67-10.41) 0.0014
rs17865678 A 4.25 (1.89-9.57) 0.0003
CHRM2 rs1424569 11 / 11 A 3.00 (0.88-10.27) 0.0300 Marshall-Gradisnik et al. (2016b) CHRM3 rs1134 C 3.15 (0.92-10.78) 0.0200
rs576386 C 3.24 (0.90-11.62) 0.0400
rs619214 T 3.67 (1.05-12.82) 0.0300 rs685550 C 6.57 (0.65-66.86) 0.0500
rs1019882 A 2.62 (0.78-8.84) 0.0500
rs1155611 C 2.62 (0.78-8.84) 0.0500 rs1155612 A 3.00 (0.88-10.27) 0.0300
rs1416789 A 2.62 (0.78-8.84) 0.0500
rs1544170 G 2.83 (0.83-9.62) 0.0400 rs1867263 G 2.71 (0.79-9.34) 0.0400
rs1867264 T 4.50 (1.26-16.08) 0.0000
rs1867265 G 2.96 (0.86-10.14) 0.0300 rs1899616 G 4.12 (1.17-14.47) 0.0100
rs2083817 T 3.05 (0.89-10.49) 0.0300
See Table 2.5 footnotes on page 22.
22 Chapter 2: Literature Review
Table 2.5. Continued Candidate genes and implicated single nucleotide polymorphisms associated
with chronic fatigue syndrome.
Gene
symbol Variant
Cases /
Controls RA OR (95% CI) p-value Study
rs2163546 G 3.11 (0.88-10.95) 0.0400
rs2165872 C 3.05 (0.89-10.49) 0.0300
rs3738436 C 2.62 (0.78-8.84) 0.0500
rs6429147 G 2.81 (0.81-9.71) 0.0400 rs6684622 G 3.50 (1.01-12.12) 0.0200
rs6688537 C 2.91 (0.85-9.94) 0.0300
rs6694220 A 2.73 (0.80-9.29) 0.0500 rs6700643 T 2.81 (0.81-9.71) 0.0400
rs7511970 G 2.62 (0.78-8.84) 0.0500
rs7513746 A 2.62 (0.78-8.84) 0.0500 rs7551001 A 2.96 (0.86-10.14) 0.0300
rs10754677 A 2.67 (0.79-9.01) 0.0500
rs10802795 T 2.62 (0.78-8.84) 0.0500
rs10802802 G 3.33 (0.97-11.49) 0.0200
rs10925941 G 2.81 (0.81-9.71) 0.0400
rs10925964 T 2.62 (0.78-8.84) 0.0500 rs11585281 C 3.14 (0.92-10.79) 0.0200
rs12029701 T 3.14 (0.92-10.79) 0.0200
rs12093821 G 3.32 (0.96-11.57) 0.0200 rs12743042 T 2.98 (0.87-10.14) 0.0300
rs16838637 A 2.96 (0.86-10.14) 0.0300 CHRM5 rs511422 C 3.06 (0.81-11.57) 0.0400
rs603152 A 3.29 (0.87-12.42) 0.0300
rs646950 T 3.26 (0.86-12.28) 0.0300 CHRNA2 rs2741341 C 4.08 (1.08-15.38) 0.0100
CHRNA4 rs11698563 C 5.99 (1.63-22.02) 0.0000
CHRNA9 rs4861065 C 4.53 (0.98-20.84) 0.0200 rs4861323 A 3.89 (0.98-15.41) 0.0100
rs7669882 A 4.53 (0.98-20.84) 0.0200
rs10009228 G 4.60 (1.16-18.18) 0.0000 rs10015231 C 2.92 (0.76-11.28) 0.0400
CHRNB1 rs2302767 T 2.6 (0.74-9.10) 0.0500
rs3829603 C 2.86 (0.80-10.15) 0.0500 rs4151134 T 3.52 (1.02-12.20) 0.0100
CHRNB4 rs12440298 T 11.00 (0.39-313.06) 0.0100
CHRND rs2767 T 3.60 (1.04-12.49) 0.0100 rs2853457 A 3.00 (0.84-10.75) 0.0400
rs3762529 T 3.15 (0.92-10.78) 0.0200
rs3791729 C 3.00 (0.88-10.24) 0.0300 rs3828246 C 3.41 (0.85-13.73) 0.0200
rs4973537 A 3.00 (0.88-10.24) 0.0300
rs11674608 C 6.79 (1.78-25.88) 0.0000 rs12463989 T 3.60 (1.04-12.49) 0.0100
rs12466358 T 3.41 (0.85-13.73) 0.0200
rs13026409 C 3.29 (0.83-13.08) 0.0200 rs67583510 G 3.67 (0.94-14.34) 0.0200
rs112001880 I 3.60 (1.04-12.49) 0.0100
CHRNE rs33970119 G 5.40 (0.44-66.84) 0.0400 CHRNG rs13018423 C 3.29 (0.83-13.08) 0.0200
TRPC6 rs11224816 T 4.10 (1.09-15.46) 0.0100
TRPM3 rs1317103 C 5.97 (1.04-34.27) 0.0100 rs3812532 C 2.83 (0.83-9.62) 0.0400
rs4620343 T 3.48 (0.85-14.22) 0.0300
rs10780950 T 4.46 (0.76-26.21) 0.0500 TRPM4 rs11083963 A 3.00 (0.86-10.48) 0.0300
TRPV2 rs3514 G 4.17 (0.65-26.90) 0.0300
rs2075763 C 5.20 (0.44-62.13) 0.0500 rs7222754 T 3.01 (0.80-11.39) 0.0500
rs12602006 A 2.71 (0.77-9.56) 0.0500
rs12942540 G 4.17 (0.65-26.90) 0.0300 rs35400274 G 4.33 (0.66-28.62) 0.0300
TRPV3 rs4790519 C 3.75 (1.04-13.49) 0.0100
RA: risk allele; OR: odds ratio; CI: confidence interval; NC: not calculable due to an allele frequency of 0 in controls.
Chapter 2: Literature Review 23
Table 2.6. List of reported genome-wide significant (7.5 × 10-8) risk loci associated with chronic
fatigue syndrome from a genome-wide association study of 42 cases and 38 controls by Schlauch and
colleagues (2016).
SNP Risk allele OR (95% CI) p-value
rs12235235 T 10.61 (4.15-27.08) 5.76 × 10-16 rs10144138 T 27.75 (6.38-120.63) 6.99 × 10-14
rs17120254 A NC 5.20 × 10-13
rs41493945 A 48.53 (6.43-366.21) 6.25 × 10-13 rs3788079 C NC 3.42 × 10-12
rs41378447 T 10.84 (4.46-26.34) 1.06 × 10-11
rs3913434 T 46.15 (6.11-348.47) 1.26 × 10-11 rs5967529 A 18.75 (8.46-41.56) 1.69 × 10-11
rs254577 C 11.04 (5.28-23.07) 2.35 × 10-11
rs270838 C 7.73 (3.31-18.05) 3.61 × 10-11 rs1523773 T NC 4.73 × 10-11
rs16827966 T 41.67 (5.51-315.00) 5.32 × 10-11
rs2249954 G 10.11 (3.96-25.82) 5.47 × 10-11 rs8029503 T 8.10 (3.47-18.93) 5.66 × 10-11
rs3095598 C 18.25 (5.32-62.62) 1.02 × 10-10
rs7010471 G 12.24 (4.09-36.66) 2.49 × 10-10 rs6757577 A 11.18 (4.10-30.51) 2.77 × 10-10
rs11157573 G 5.09 (2.40-10.77) 2.97 × 10-10
rs16987633 A 6.21 (3.10-12.43) 3.46 × 10-10 rs12312259 C 6.46 (3.05-13.69) 3.60 × 10-10
rs948440 C 6.92 (3.14-15.26) 3.92 × 10-10
rs6445832 G 8.75 (3.42-22.38) 4.36 × 10-10 rs9585049 T 15.75 (4.58-54.13) 5.25 × 10-10
rs7220341 G 4.97 (2.53-9.75) 5.41 × 10-10
rs2816751 C 5.11 (2.48-10.51) 5.43 × 10-10 rs2200706 T 5.91 (2.98-11.74) 5.48 × 10-10
rs17255510 C 10.23 (4.82-21.71) 6.61 × 10-10
rs6892217 T 8.19 (4.01-16.73) 6.61 × 10-10 rs17112444 A 21.64 (4.96-94.39) 8.02 × 10-10
rs7849492 C 5.91 (2.74-12.75) 9.95 × 10-10
rs686190 G 11.65 (3.88-34.92) 1.11 × 10-9 rs16826918 G 10.11 (3.96-25.82) 1.13 × 10-9
rs12317807 T 4.71 (2.13-10.43) 1.47 × 10-9 rs5974598 T 4.71 (2.13-10.43) 1.55 × 10-9
rs1932556 T 54.13 (7.15-409.98) 1.63 × 10-9
rs6797416 G NC 1.71 × 10-9 rs2733416 G NC 1.71 × 10-9
rs17035358 A 9.66 (3.53-26.41) 1.72 × 10-9
rs17368935 G 9.66 (3.53-26.41) 1.72 × 10-9 rs6679280 T 9.66 (3.53-26.41) 1.72 × 10-9
rs3867246 T 5.20 (2.35-11.48) 1.88 × 10-9
rs689462 C 8.54 (3.52-20.76) 2.08 × 10-9 rs9285128 A 4.00 (2.04-7.86) 2.15 × 10-9
rs822027 A 35.53 (4.69-269.28) 2.52 × 10-9
rs11168709 T 35.53 (4.69-269.28) 2.52 × 10-9 rs12055682 G 4.79 (2.43-9.41) 2.99 × 10-9
rs17047694 T 3.07 (1.60-5.90) 3.66 × 10-9
rs10978470 G 5.59 (2.64-11.85) 4.32 × 10-9
rs890527 T 3.12 (1.62-6.01) 4.60 × 10-9
rs11062852 C 4.07 (2.00-8.26) 4.84 × 10-9
rs16992281 A NC 5.22 × 10-9 rs6675622 T 9.20 (4.28-19.77) 5.94 × 10-9
rs6863118 G 7.37 (3.15-17.22) 6.22 × 10-9
rs10121299 C 4.42 (2.27-8.61) 6.62 × 10-9 rs12014391 A 9.38 (4.53-19.39) 6.66 × 10-9
rs12391243 C 8.35 (4.02-17.32) 6.68 × 10-9
rs1041296 G 4.76 (2.37-9.59) 6.89 × 10-9 rs11027583 T 7.39 (3.04-17.99) 7.03 × 10-9
rs12305678 G 4.41 (2.21-8.79) 7.87 × 10-9
rs9581771 T 7.76 (3.19-18.87) 7.96 × 10-9 rs4022211 G 7.41 (3.67-14.98) 9.09 × 10-9
rs16883408 C 5.51 (2.76-10.98) 1.06 × 10-8
rs41456945 C 18.50 (4.23-80.94) 1.07 × 10-8 rs361236 A 3.33 (1.72-6.45) 1.07 × 10-8
rs1007540 G 3.25 (1.70-6.21) 1.17 × 10-8
rs7143222 T NC 1.54 × 10-8 rs17092382 A NC 1.54 × 10-8
See Table 2.6 footnotes on page 24.
24 Chapter 2: Literature Review
Table 2.6. Continued List of reported genome-wide significant (7.5 × 10-8) risk loci associated with
chronic fatigue syndrome from a genome-wide association study of 42 cases and 38 controls by
Schlauch and colleagues (2016).
SNP Risk allele OR (95% CI) p-value
rs7549528 C NC 1.54 × 10-8 rs6854376 T 10.53 (3.50-31.63) 1.71 × 10-8
rs16902672 C 5.25 (2.60-10.59) 1.77 × 10-8
rs4473594 A 8.54 (3.52-20.76) 1.81 × 10-8 rs10737169 A 5.36 (2.68-10.75) 2.51 × 10-8
rs7883119 G 7.15 (3.52-14.55) 2.57 × 10-8
rs4623336 T 4.85 (2.29-10.27) 2.68 × 10-8 rs2748997 C 9.18 (3.59-23.48) 2.76 × 10-8
rs584569 A 8.74 (3.19-23.95) 2.84 × 10-8
rs13339179 T 12.83 (3.72-44.3) 2.89 × 10-8 rs1222400 T 12.83 (3.72-44.3) 2.89 × 10-8
rs2882361 G 10.44 (4.26-25.54) 3.02 × 10-8
rs41464146 C 18.50 (4.23-80.94) 3.22 × 10-8 rs9446695 T 17.53 (4.00-76.78) 3.46 × 10-8
rs7290437 G 2.73 (1.39-5.35) 3.52 × 10-8
rs12607783 A 4.71 (2.13-10.43) 4.31 × 10-8 rs606324 A 6.48 (2.52-16.69) 4.39 × 10-8
rs2869820 T NC 4.39 × 10-8
rs6643261 A 30.57 (7.04-132.77) 4.55 × 10-8 rs17133553 A 5.53 (2.80-10.91) 4.74 × 10-8
rs2816936 A 14.99 (5.46-41.14) 4.91 × 10-8
rs1915603 G 9.73 (2.79-33.90) 5.15 × 10-8 rs17052315 A 10.00 (3.32-30.08) 5.97 × 10-8
rs6502875 G 3.90 (2.02-7.53) 5.97 × 10-8
rs10047684 A 4.51 (2.28-8.94) 7.31 × 10-8
OR: odds ratio; CI: confidence interval; NC: not calculable due to an
allele frequency of 0 in either cases or controls
In 2017, the largest GWA of a fatigue phenotype was conducted within the UK
Biobank sample. The study by Deary and colleagues (2017) investigated self-
reported tiredness experienced over the past two weeks, which had a SNP-based
heritability (i.e., the proportion of a traits variation which is explained by all SNPs
investigated within a GWA analysis dataset) of 8.4% (standard error [SE] = 0.6%).
One genome-wide significant SNP (Affymetrix ID 1:64178756_C_T in an intergenic
region on chromosome 1, p = 1.36 × 10-11) and two suggestive peaks on chromosome
1 (rs142592148 in the intron of SLC44A5, p = 5.88 × 10-8) and 17 (rs2555592 in the
intron of PAFAH1B1, p = 6.86 × 10-8) were associated with self-reported tiredness.
Within this study, a gene-based association analysis identified five genes which
reached genome-wide significance (p < 2.77 × 10-6) and 44 genes which were
suggestively associated (p < 1.00 × 10-4) with self-reported tiredness (Table 2.7)
(Deary et al., 2017).
Chapter 2: Literature Review 25
Table 2.7. List of genes associated (p < 2.77 × 10-6) or suggestively associated (p < 1.00 × 10-4) with
self-reported tiredness.
Gene Chromosome p-value
DRD2 11 2.94 × 10-7
PRRC2C 1 1.43 × 10-6 C3orf84 3 1.45 × 10-6
ANO10 3 1.52 × 10-6
ASXL3 18 2.67 × 10-6 RHOA 3 4.07 × 10-6
CTNND1 11 4.09 × 10-6
THEM4 1 5.44 × 10-6 FBXO21 12 5.66 × 10-6
ADARB1 21 6.01 × 10-6
NAPA 19 6.06 × 10-6 KANSL1L 2 6.24 × 10-6
RHCG 15 6.90 × 10-6
PLAC8 4 6.95 × 10-6
KLF7 2 7.40 × 10-6
RPE 2 1.00 × 10-5
TMX2 11 1.36 × 10-5 SNF8 17 1.38 × 10-5
CCDC36 3 1.38 × 10-5
SSBP4 19 1.87 × 10-5 ISYNA1 19 1.93 × 10-5
RELT 11 2.37 × 10-5 CSMD3 8 2.49 × 10-5
ZDHHC5 11 2.66 × 10-5
METTL16 17 2.67 × 10-5 SRRM4 12 3.03 × 10-5
BSN 3 3.20 × 10-5
NRXN1 2 3.25 × 10-5 ZNF780A 19 3.30 × 10-5
SMC1B 22 3.33 × 10-5
TCTA 3 3.36 × 10-5 GIP 17 3.45 × 10-5
CKMT1A 15 4.09 × 10-5
NICN1 3 4.18 × 10-5 UBE2Z 17 5.11 × 10-5
DAG1 3 5.26 × 10-5
ATP11B 3 5.28 × 10-5 PSMC4 19 5.44 × 10-5
FAM168A 11 5.86 × 10-5
CCNT2 2 6.25 × 10-5 OPA1 3 6.42 × 10-5
CATSPER2 15 6.52 × 10-5
ZBTB37 1 6.67 × 10-5 ELL 19 6.91 × 10-5
SERPING1 11 7.49 × 10-5
PLGRKT 9 7.89 × 10-5 PRR12 19 8.37 × 10-5
UBA7 3 8.48 × 10-5
CAMK1D 10 9.36 × 10-5
The study by Deary and colleagues (2017)
was conducted in a UK population.
Tiredness was assessed by the question: “Over the past two weeks, how often have
you felt tired or had little energy?”; 6,948
individuals responded “nearly every day”, 6,404 individuals responded “more than half
the days”, 44,208 individuals responded
“several days”, and 51,416 individuals responded “not at all”.
2.7 CLASSIFICATION OF MDD AND MIDD
The standard definition used for diagnosis with MDD and MiDD is the Diagnostic
and Statistical Manual of Mental Disorders (DSM) (American Psychiatric
Association, 2000, 2013). To be diagnosed with MDD, an individual must have had
26 Chapter 2: Literature Review
at least one major depressive episode. A major depressive episode is assessed over a
two-week period and contains nine symptom criteria. To meet the DSM-V criteria of
a depressive episode, symptoms from at least one of the first two criteria must be
exhibited. The criteria of a depressive episode are:
1. Depressed mood
2. Loss of interest or pleasure (anhedonia)
3. Change in weight (5% of weight within a month) or appetite
4. Insomnia or hypersomnia
5. Psychomotor agitation or retardation
6. Fatigue or loss of energy
7. Feelings of worthlessness or excessive guilt
8. Inability to concentrate or indecisiveness
9. Recurrent thoughts of death and suicidal thoughts, plans, or attempts.
Diagnosis of MDD requires endorsement of at least five of the criteria
(American Psychiatric Association, 2013). Similarly, diagnosis of MiDD is assessed
using the symptoms of a major depressive episode. However, diagnosis only requires
endorsement of two (to four) of the criteria (American Psychiatric Association,
2000).
2.8 EPIDEMIOLOGY OF MDD AND MIDD
Epidemiological studies have consistently shown females have an increased risk of
diagnosis with MDD compared to males. Mean age at first onset of MDD occurs at
age 31.7 ± 12.3 (Fernandez-Pujals et al., 2015). The prevalence of a lifetime, twelve
month, and current diagnosis of MDD is estimated at 16.2%, 6.6%, and 4.1%,
respectively (Centers for Disease Control and Prevention, 2010; Kessler et al., 2003).
Meanwhile, prevalence of a current diagnosis of MiDD is 5.1% (Centers for Disease
Control and Prevention, 2010). MDD prevalence increases from puberty, with
variation in prevalence observed at different ages (Centers for Disease Control and
Prevention, 2010; Kessler et al., 2003). In adults (aged ≥ 18 years), lifetime
prevalence of MDD gradually increases before declining at ~ 60 years (Kessler et al.,
2003). Interestingly, the 12-month prevalence of MDD is highest in the 18-29 year
Chapter 2: Literature Review 27
age range after which it progressively decreases. Meanwhile, the prevalence of
current diagnosis of MDD exhibits a similar pattern to that of the lifetime prevalence,
gradually increasing from 18-24 years, peaking at 45-64 years after which it
decreases (Centers for Disease Control and Prevention, 2010). The pattern of current
diagnosis of MiDD is particularly interesting. The 18-24 age range has the highest
prevalence which gradually decreases until 35-44 years after which the prevalence
increases to the 65+ age group.
2.9 PATHOPHYSIOLOGY OF MDD
To date, aetiological investigations have indicated MDD is a complex, multifactorial
trait with a multitude of biological functions and environmental contributions
implicated in the pathophysiology of depression. To date, neuroendocrine
functioning, the central nervous system, genetics, and environmental factors have
been implicated in the pathophysiology of depression. However, in most instances, it
is impossible to determine if the altered biological function observed in MDD cases
is a causal factor rather than a consequence of depression.
2.9.1 Endocrine and Neurologic Dysfunction
Neurobiological changes associated with a sustained stress response, particularly
those involving the neuroendocrine, noradrenergic, and serotonergic pathways, have
been emphasized in MDD aetiological research (National Research Council (US) &
Institute of Medicine (US) Committee on Depression, 2009; Thase et al., 2014). In
particular hypercortisolism, elevated HPA-axis activity, and deficiency of the
serotonin neurotransmitters (National Research Council (US) & Institute of Medicine
(US) Committee on Depression, 2009; Thase et al., 2014; Verduijn et al., 2015).
Additionally, neuroimaging techniques have enabled identification of brain regions
associated with depression symptomatology, namely, the amydgala, anterior
cingulate cortex, dorsolateral prefrontal cortex, medial prefrontal cortex,
orbitofrontal cortex, and striatum (Treadway & Pizzagalli, 2014).
2.9.2 Genetics
A multitude of linkage and candidate gene associations studies have been conducted
for MDD (Flint & Kendler, 2014). However, conflicting results have been obtained
between investigations and the findings have not been replicated in GWA studies.
Although robustly associated risk loci for MDD have recently been identified, to
28 Chapter 2: Literature Review
date, the proportion of the variance explained by all SNPs included in the GWA
investigations is relatively small compared to heritability estimates identified from
twin studies—indicating our understanding of the genetics underlying MDD is
limited. Additionally, gene-gene and gene-environment interactions could contribute
to the development and recurrence of MDD, however, studies which enable
identification of these mechanisms are currently vastly underpowered.
2.9.3 Environmental Factors
Stressful life events and exposure to early adversity have consistently been
associated with MDD. An increase in depression symptoms and MDD onset has
consistently been associated with triggering stressful life events (either positive or
negative), such as death, illness, or injury of a loved one (i.e., spouse, close family
member, or friend), outstanding personal achievement, or change in work, residence,
recreation, social activities, sleeping habits, and eating habits (Kendler et al., 1999;
National Research Council (US) & Institute of Medicine (US) Committee on
Depression, 2009; Shapero et al., 2014). Similarly, MDD in adolescents or adults has
been significantly associated with childhood physical, sexual, or emotional abuse
with some evidence indicating childhood adversity is predictive of chronic or
recurrent depression (Bifulco et al., 2002; Kendler et al., 2000; National Research
Council (US) & Institute of Medicine (US) Committee on Depression, 2009).
However, the mechanisms underlying the association between environmental factors
and MDD are unknown.
2.10 HERITABILITY OF MDD AND MIDD
Numerous studies have been conducted investigating the familiality of MDD and
MiDD. Familial clustering has consistently been observed with first-degree relatives
of MDD probands having increased odds of experiencing MDD (Mantel-Haenszel
odds ratio (OR) = 2.84, 95% confidence interval (CI) = 2.31-3.49) (Gershon et al.,
1982; Maier et al., 1993; Sullivan et al., 2000; Tsuang et al., 1980; Weissman et al.,
1984; Weissman et al., 1993). Furthermore, first-degree relatives and patients with
MiDD have an increased risk of experiencing MDD (Chen et al., 2000; Cuijpers et
al., 2004; Cuijpers & Smit, 2004; Judd et al., 1997; Lewinsohn et al., 2003; Rapaport
et al., 2002). These results indicate MDD has a significant additive genetic and/or
common environmental contribution, and suggests a proportion of these genetic and
Chapter 2: Literature Review 29
environmental factors are shared with MiDD. A recent large family-based population
study in Scotland has utilised 20,198 adults aged ≥ 18 (median age of 49) to
investigate the epidemiology and heritability (modelled using the full pedigree
structure of the cohort and the phenotypic variability among family members) of
MDD (classified according to the DSM-IV criteria) (Fernandez-Pujals et al., 2015).
Within the cohort males had a lower estimated heritability for MDD at 35% (95% CI
= 8-63%; C = 8%, 95% CI = 0-20%) compared to females at 44% (95% CI = 25-
61%; C = 4%, 95% CI = 0-13%). The authors also determined recurrent MDD had a
higher heritability at 41% (95% CI = 20-60; C = 7%, 95% CI = 0-16%) compared to
single episode MDD at 28% (95% CI = 14-41; C = 3%, 95% CI = 0-8).
To date, no studies have investigated the heritability of MiDD, while numerous
twin studies have been conducted to estimate the heritability of MDD (Table 2.8)
(Bierut et al., 1999; Kendler et al., 1995; Kendler & Prescott, 1999; Kendler et al.,
2001; Kendler et al., 2006; Lyons et al., 1998; McGuffin et al., 1996; Sullivan et al.,
2000). Within clinical samples the contribution of A, C, and E have been estimated
to range from 49-58%, 0-21% and 30-42%, respectively, in males, and 17-38%, 0%,
and 62-83%, respectively, in females (Kendler et al., 1995; McGuffin et al., 1996).
Similarly, within community cohorts, the contribution of A, C, and E have been
estimated to range from 18-57%, 0-2%, and 41-81%, respectively, in males, and 36-
78%, 0%, and 22-64%, respectively, in females (Bierut et al., 1999; Kendler et al.,
1995; Kendler & Prescott, 1999; Kendler et al., 2001; Kendler et al., 2006; Lyons et
al., 1998). Finally, within a meta-analysis, the contribution of A, C, and E has been
estimated at 37%, 0%, and 63%, respectively (Sullivan et al., 2000).
30 Chapter 2: Literature Review
Table 2.8. Heritability estimates (and their 95% confidence intervals) of the unique additive genetic factors (A), common environmental factors (C), and unique
environmental factors (E) contributing to major depressive disorder (MDD).
Diagnosis Population Sex Number of twin pairs
(MZ / DZ)
Mean age ±
standard deviation
(Age range)
A (%) C (%) E (%) Notes Study
Individual depression symptoms in a community cohort.
DSM-IV United States M & F 1,206 / 1,878 and
1,325 singleton twins
(≥ 14) 29 0 71 Depressed mood Kendler et al. (2013)
29 1 70 Loss of interest/pleasure 12 5 83 Weight or appetite change
19 4 77 Insomnia or hypersomnia
20 2 78 Psychomotor changes 38 1 61 Fatigue
29 2 69 Worthlessness/guilt
25 4 71 Problems with concentration 26 5 69 Suicidal ideation
≥ 1major depressive episode reported within the specified timeframe in a community cohort.
DSM-III-R United States F 508 / 350 (≥ 14) 41 (27-54) - 59 (46-73) 1 year Kendler and Aggen (2001) 41 (26-55) - 59 (45-74) 6 month
35 (16-52) - 65 (48-84) 3 month
34 (11-55) - 66 (45-89) 1 month Lifetime MDD in a community cohort.
DSM-III United States F 1,033 twin pairs and
97 singleton twins
30.1 ± 7.6 (17-55) 39 (34-45) - 60 Remaining variance is
explained by age correction
Kendler et al. (1992)
DSM-III-R 42 (37-47) - 58
DSM-III-R Sweden M 251 / 495 NR 57 (0-73) 2 (0-73) 41 (4-100) Included in meta-analysis Kendler et al. (1995)
F 78 (18-94) 0 (0-42) 22 (6-55)
DSM-III-R United States M 1,874 / 1,498 46.6 ± 2.8 (36-55) 36 (11-47) 0 (0-20) 64 (53-75) Included in meta-analysis Lyons et al. (1998)
DSM-III-R United States M 75 / 145 (18-60) 31 (5-41) 0 (0-22) 69 (59-79) Included in meta-analysis Kendler and Prescott (1999) F 38 (1-50) 0 (0-31) 62 (50-75)
M & F 39 (30-47) - 61
DSM-III-R Australian M 1,323 / 1,339 42 ± 11.23 (28-84) 24 (0-39) 0 (0-26) 76 (61-91) Included in meta-analysis Bierut et al. (1999)
F 44 ± 12.35 (28-89) 44 (29-53) 0 (0-12) 56 (47-65)
DSM-IV M 42 ± 11.23 (28-84) 18 (0-36) 0 (0-27) 81 (64-97)
F 44 ± 12.35 (28-89) 36 (15-46) 0 (0-16) 64 (54-75)
DSM-III-R United States M 6,763 twin pairs and
895 singleton twins
NR 44 (38-50) - 56 (50-62) Kendler et al. (2001)
F 57 (51-63) - 43 (37-49)
DSM-IV United States F 855 / 661 15.5 (13-19) 40 (24-55) - 60 (45-76) Glowinski et al. (2003) DSM-IV Sweden M 4,091 / 11,402 NR 29 (19-38) - 71 (62-81) Kendler et al. (2006)
F 42 (36-47) - 58 (53-64)
M & F 38 - 62
See Table 2.8 footnotes on page 31.
Chapter 2: Literature Review 31
Table 2.8. Continued Heritability estimates (and their 95% confidence intervals) of the unique additive genetic factors (A), common environmental factors (C), and unique
environmental factors (E) contributing to major depressive disorder (MDD).
Diagnosis Population Sex Number of twin pairs
(MZ / DZ)
Mean age ±
standard deviation
(Age range)
A (%) C (%) E (%) Notes Study
DSM-IV Sri Lankan F 465 / 307 34 (≥ 15) 59 (43-72) - 41 (28-57) Ball et al. (2009)
DSM-IV United States
African-American F
254 twin pairs and 296
singleton twins (18-28) 56 (29-78) - 44 (22-72) Duncan et al. (2014)
United States
European-American
1,514 twin pairs and
1,712 singleton twins 41 (29-52) - 59 (48-71)
United States
African and European American
1,768 twin pairs and 2,008 singleton twins
43 (33-53) - 57 (47-67)
Lifetime MDD in a clinical cohort.
DSM-III-R Sweden M 23 / 64 NR 49 (0-99) 21 (0-89) 30 (1-93) Included in meta-analysis Kendler et al. (1995) F 17 (0-55) 0 (0-38) 83 (45-83)
DSM-IV United Kingdom M 68 / 109 NR 58 (4-81) 0 (0-40) 42 (19-72) Included in meta-analysis McGuffin et al. (1996)
F 38 (14-61) 0 (0-24) 62 (39-86) M & F 48 0 52
Lifetime MDD in a clinical or community cohort.
DSM-III-R Sweden M & F 274 / 559 NR 60 - 40 Included in meta-analysis Kendler et al. (1995) DSM European ancestry M & F 4,736 / 5,567 NR 37 (31-42) 0 (0-5) 63 (58-67) Meta-analysis Sullivan et al. (2000)
Note: 95% confidence intervals are reported as described in the original studies therefore, if they are absent they were not reported in the original publication. Populations are of European decent unless otherwise
specified. M: male; F: female; M & F: male and female; NR: not reported.
32 Chapter 2: Literature Review
2.11 MOLECULAR GENETICS OF MDD
Considerable effort has gone into investigating the molecular genetics of MDD, with
numerous linkage, candidate gene, and GWA studies conducted (Flint & Kendler,
2014). Linkage studies conducted for MDD have yielded some significant findings.
However, the results of comparable studies are inconstant, leading one to question
the contribution of rare large effect genetic loci to MDD (Flint & Kendler, 2014;
Lohoff, 2010). Similarly, conflicting results have been obtained from the multitude
of CGA analyses conducted. Meta-analyses have been conducted for a minority (N =
26) of the investigated genes, which resulted in the identification of six significant
genes—based on allelic tests for individual variants (Table 2.9) (Flint & Kendler,
2014; Lopez-Leon et al., 2008). Although conflicting results have been obtained
from different meta-analyses of the same variant—which could be due to differing
sample sizes. Furthermore, to date the genes implicated by CGA studies have not
been replicated in GWA studies (Bosker et al., 2011; Flint & Kendler, 2014).
Twenty-two GWA studies have been conducted which attempted to identify
genetic risk loci associated with replication of MDD CGA study results (N = 1),
MDD (N = 3), recurrent MDD (N = 4), MDD or recurrent MDD (N = 5), MDD or
recurrent MDD under a recessive model (N = 1), MDD or recurrent MDD and age at
onset (N = 2), self-reported major depression (N = 1), depressive symptoms (N = 4),
and depressive symptoms and MDD (N = 1) (Table 2.10).
Chapter 2: Literature Review 33
Table 2.9. Candidate genes associated with major depressive disorder in meta-analysis studies.
Gene
symbol Variant
Number of
Studies
Cases /
Controls RA OR (95% CI) p-value Study
APOE ε2/ ε3/ ε4 7 827 / 1616 ε3 1.96 (1.03-3.72) ≤ 0.0010 Lopez-Leon et
al. (2008)
DRD4 48 bp
repeat 5 318 / 814
2
allele 1.73 (1.29-2.32) 0.00030
Lopez Leon et
al. (2005)
GNB3 rs5443 3 375 / 492 T 1.38 (1.13-1.69) ≤ 0.0500 Lopez-Leon et al. (2008)
HTR1A rs6295 13 3,199 / 4,380 G 1.15 (1.04-1.28) 0.0060 Kishi et al.
(2013)
7 1,658 / 2,046 G 1.22 (1.03-1.44) 0.0330 Kishi et al.
(2009)
4 NA G 1.16 (0.98-1.38) NS Lopez-Leon et
al. (2008)
MTHFR rs1801133 17 3,341 / 13,840 T 1.02 (0.96-1.08) 0.5790 Peerbooms et al. (2011)
10 1,280 / 10,429 T 1.14 (1.04-1.26) < 0.0500 Gilbody et al.
(2007)
6 875 / 3,859 T 1.20 (1.07-1.34) < 0.0500 Lopez-Leon et
al. (2008)
5 291 / 897 T 1.15 (0.97-1.36) > 0.1000 Zintzaras (2006)
4 1,222 / 835 T 0.96 (0.84-1.09) 0.3900 Gaysina et al.
(2008)
SLC6A4 44 bp
ins/del 39 6,836/14,903 S 1.09 (1.02-1.16) 0.0070
Clarke et al.
(2010)
22 3,752 / 5,707 S 1.11 (1.04-1.19) ≤ 0.0500 Lopez-Leon et al. (2008)
14 1,961 / 3,402 S 1.05 (0.96-1.14) 0.2800 Lasky-Su et al.
(2005)
11 941 / 2,110 S 1.08 (0.96-1.22) 0.1980 Anguelova et al.
(2003)
4 275 / 739 S 1.23 (1.01-1.52) 0.0420 Furlong et al. (1998)
RA: risk allele; OR: odds ratio; CI: confidence interval; bp: base pair; ins/del: insertion/deletion.
34 Chapter 2: Literature Review
Table 2.10. Summary of the genome-wide association studies conducted for major depressive
disorder (MDD).
Population Number of
SNP loci Sample Cases / Controls Study
Replication of MDD candidate gene studies
Western European ancestry 2,467,430 Discovery 1,738 / 1,802 Bosker et al. (2011)
MDD
Netherlands 435,291 Discovery 1,738 / 1,802 Sullivan et al. (2009) Replication 6,079 / 5,893
Germany 491,238 Discovery 597 / 1,295 Rietschel et al. (2010)
Meta-analysis 1,006 / 1,836 Netherlands 433,556 Discovery 1,726 / 1,630 Aragam et al. (2011)
Recurrent MDD
United Kingdom 471,747 Discovery 1,636 / 1,594 Lewis et al. (2010) Meta-analysis 3,054 / 3,512
Germany and Switzerland 522,008 Discovery 926 / 866 Muglia et al. (2010)
370,697 Discovery 492 / 1,052
494,678 Meta-analysis 1,359 / 1,782
European decent 457,670 Discovery 805 / 805 Power et al. (2013)
Han Chinese Women 6,242,619 Discovery 5,303 / 5,337 CONVERGE Consortium (2015) Meta-analysis 8,534 / 8,523
MDD or recurrent MDD
United States 382,598 Discovery 1,221 / 1,636 Shyn et al. (2011) 2,391,203 Meta-analysis 3,957 / 3,428
United States 671,424 Discovery 1,020 / 1,636 Shi et al. (2011) Europe / United States 365,676 Discovery 356 / 366 Kohli et al. (2011)
Meta-analysis 4,088 / 11,001
Australia / Europe / United States
1,079,979 Discovery 2,431 / 3,673 Wray et al. (2012) 427,362 Meta-analysis 5.763 / 6.901
Australia / Europe /
United States
1,235,109 Discovery 9,240 / 9,519 Major Depressive Disorder
Working Group of the Psychiatric GWAS Consortium et al. (2013)
Replication 6,783 / 50,695
MDD or recurrent MDD under a recessive model
Australia / Europe / United States
929,138 Discovery 9,238 / 9,521 Power et al. (2014)
MDD or recurrent MDD and age of onset
European decent 471,581 Discovery 2,746 / 1,594 Power et al. (2012) Germany Replication 1,480 / 1,584
European decent 1,235,109 Discovery 8,920 / 9,519 Power et al. (2017)
Meta-analysis 22,158 / 133,749 Self-reported major depression
European 13,519,496 Discovery 75,607 / 231,747 Hyde et al. (2016)
~ 1,220,000 Meta-analysis 121,380 / 338,101 Depressive symptoms
Sardinian ~ 2,500,000 Discovery 3,972 Terracciano et al. (2010)
United States ~ 2,500,000 Discovery 839 Meta-analysis 4,811
European ~ 300,000 Discovery 4,525 Luciano et al. (2012)
Replication 527 Replication 1,383
European decent ~ 2,500,000 Discovery 34,549 Hek et al. (2013)
Meta-analysis 51,258 European decent 6,544,862 Discovery 105,739 Okbay et al. (2016)
4,661,873 Discovery 7,231 / 49,316
998,753 Discovery 9,240 / 95,19 Meta-analysis 180,866
MDD or recurrent MDD and depressive symptoms
European decent Discovery 70,017 Direk et al. (2016) 918,921 Meta-analysis 98,345
In 2011, a genome-wide significant SNP (rs1545843, p = 2.30 × 10-8) within a
gene desert on chromosome 12) was reported for a recessive model of inheritance
(Kohli et al., 2011). However, the consensus within the research community is that
this was a false positive finding (Cohen-Woods et al., 2013). In 2015, two genome-
wide significant SNP loci (rs35936514 in the intron of LHPP and rs12415800 near
the SIRT1 gene) were associated with MDD (Table 2.11) in Han Chinese Women
Chapter 2: Literature Review 35
(CONVERGE Consortium, 2015). However, these genes have not been associated
with MDD in Europeans. In 2016, 17 SNP loci within 15 independent genomic
locations (rs10514299 in an intron of TMEM161B-AS1; rs1518395 in an intron of
VRK2, rs2179744 in an intron of L3MBTL2, rs11209948 downstream of NEGR1,
rs454214 upstream of MEF2C, rs301806 in an intron of RERE, rs1475120 in an
intron of LIN28B, rs10786831 in an intron of SORCS3, rs12552 in the 3′ untranslated
region (UTR) of OLFM4, rs6476606 in an intron of PAX5, rs8025231 in an
intergenic region between MEIS2 and TMCO5A, rs12065553 in an intergenic region
on chromosome 1, rs1656369 in the intergenic region between RSRC1 and MLF1,
rs4543289 in an intergenic region on chromosome 5, rs2125716 upstream of
SLC6A15, rs2422321 downstream of NEGR1, and rs7044150 in the intergenic region
between KIAA0020 and RFX3) were associated with self-reported major depression
(Table 2.11) , which had a SNP-based heritability of 5.9%, in Europeans (Hyde et al.,
2016). Further insights into the molecular genetics of depression, in Europeans, were
identified in 2016 through the investigation of depressive symptoms and a broad
depression phenotype (that included MDD, recurrent MDD, and depressive
symptoms). Depressive symptoms, which had a SNP-based heritability of 4.7% (SE
= 0.004), was associated with two SNPs (rs7973260 in an intron of KSR2 and
rs62100776 in an intron of DCC) (Table 2.11) (Okbay et al., 2016). Similarly, one
SNP (rs9825823 located in the intron of FHIT) was associated with a broad
depression phenotype that included MDD or recurrent MDD and depressive
symptoms, which had a SNP-based heritability of 21% (SE = 0.020), in Europeans
(Table 2.11) (Direk et al., 2016). Finally, in 2017, one SNP (rs7647854 located in an
intergenic region on chromosome 3) was associated with MDD and recurrent MDD
in adults aged over 27 years (Table 2.11) (Power et al., 2017).
Additional SNP-based heritability estimates have been calculated for MDD.
The SNP-based heritability was estimated at 32% (SE = 0.086) in the Netherlands
MDD cohort (1,620 cases and 1,625 controls) (Lubke et al., 2012). Similarly, the
SNP-based heritability was estimated at 21% (SE = 0.021) in the Psychiatric
Genomics Consortium MDD cohort (9,041 cases and 9,381 controls) (Cross-
Disorder Group of the Psychiatric Genomics Consortium, 2013). Finally, the SNP-
based heritability of MDD age of onset was estimated at 17.5% (SE = 0.104) in 3,468
unrelated MDD of European decent (Ferentinos et al., 2015).
36 Chapter 2: Literature Review
Table 2.11. List of genome-wide significant (5 × 10-8) risk loci associated with depression from a
genome-wide association studies by the CONVERGE consortium (2015), Hek and colleagues (2013),
and Direk and colleagues (2016).
SNP Risk allele OR (95% CI) p-value
Recurrent MDD in Han Chinese women (CONVERGE Consortium, 2015) rs35936514 C 1.19 (1.13-1.25) 6.45 × 10-12
rs12415800 A 1.15 (1.10-1.20) 2.53 × 10-10
MDD and recurrent MDD in Europeans (Hyde et al., 2016)1 rs10514299 C 1.05 (1.04-1.07) 9.99 × 10-16
rs1518395 A 1.03 (1.02-1.05) 4.32 × 10-12
rs2179744 G 1.04 (1.02-1.05) 6.03 × 10-11 rs11209948 G 1.04 (1.02-1.05) 8.38 × 10-11
rs454214 T 1.03 (1.02-1.05) 1.09 × 10-9
rs301806 T 1.03 (1.02-1.04) 1.90 × 10-9 rs1475120 G 1.03 (1.02-1.04) 4.17 × 10-9
rs10786831 G 1.03 (1.02-1.04) 8.11 × 10-9
rs12552 G 1.05 (1.03-1.06) 8.16 × 10-9 rs6476606 G 1.03 (1.02-1.04) 1.20 × 10-8
rs8025231 A 1.04 (1.02-1.05) 1.23 × 10-8
rs12065553 A 1.03 (1.02-1.05) 1.32 × 10-8 rs1656369 A 1.04 (1.02-1.05) 1.34 × 10-8
rs4543289 T 1.03 (1.02-1.04) 1.36 × 10-8
rs2125716 G 1.04 (1.02-1.05) 3.05 × 10-8 rs2422321 A 1.03 (1.02-1.04) 3.18 × 10-8
rs7044150 T 1.03 (1.02-1.05) 4.31 × 10-8
MDD and age of onset > 27 years (Power et al., 2017) rs7647854 G 1.16 (1.11-1.21) 5.20 × 10-11
Depressive symptoms in Europeans (Okbay et al., 2016)
rs7973260 A 1.03 (1.02-1.04) 1.80 × 10-9 rs62100776 T 1.03 (1.02-1.03) 8.50 × 10-9
MDD and depressive symptoms in Europeans (Direk et al., 2016)
rs9825823 T NC 8.20 × 10-9
OR: odds ratio; CI: confidence interval; NC: not calculable. ap-values
reported are from the meta-analyses of the discovery and replication analyses,
while the OR and CI reported are with respect to the discovery cohort as the effect was not reported for the meta-analysis
2.12 COMORBIDITY BETWEEN MDD AND FATIGUE
Depressive disorders are present in 50-75% of patients presenting with medically
unexplainable symptoms (Kroenke et al., 1994; Kroenke, 2003). Furthermore, the
risk of diagnosis with a psychiatric disorder has been linked to presentation with
increased numbers of unexplained physical symptoms. Although the presence of
physical symptoms does not confirm a psychiatric explanation of the symptoms
(Kroenke et al., 1994). The prevalence of a current or lifetime diagnosis of MDD in
individuals with CF is 15.3% and 76.5%, respectively, and 10.2% and 67.3%,
respectively, when fatigue is excluded as a diagnostic criterion (Katon et al., 1991).
Similarly, the prevalence of current MDD is 10.7% and current MiDD or MDD is
14.5% in individuals with CFS (Cella et al., 2013; Janssens et al., 2015). The high
prevalence of comorbid fatigue and depression often results in fatigue being
perceived as a purely psychological symptom. However, numerous studies have
shown a subgroup of fatigued individuals exists which are not depressed (Harvey et
al., 2009; Hickie et al., 1999c; Van Der Linden et al., 1999).
Chapter 2: Literature Review 37
The CDC, CCC, and ICC diagnostic criteria for CFS, ME/CFS, and ME,
respectively, list pre-existing depression as an exclusionary condition (Carruthers et
al., 2003; Carruthers et al., 2011; Fukuda et al., 1994). However, there are a few
differential symptoms between pure cases of depression, CFS, ME/CFS, and ME
compared to CFS, ME/CFS or ME with comorbid depression, which could indicate a
concomitant presentation (Brown, 2014; Griffith & Zarrouf, 2008). Notably, cases of
CFS, ME/CFS, and ME do not exhibit anhedonia and in general have a sudden onset,
while depression onset is gradual (Brown et al., 2013; Jason et al., 2005). Similarly,
exercise helps relieve depression while exacerbating CFS symptoms (Griffith &
Zarrouf, 2008). Furthermore, the fatigue experienced within a depressive episode is
not as severe and depressed cases do not exhibit a number of the diagnostic
symptoms of CFS (i.e., sore throat and tender lymph nodes). Investigation of CFS
onset has revealed a seasonal aspect to both ICF and CFS onset, which could
implicate either an infectious agent or depression (Jason et al., 2001; Zhang et al.,
2000). However, CFS patients have reduced levels of seasonal changes in symptoms
associated with seasonal major depressive disorder (Garcia-Borreguero et al., 1998).
In 2001, Addington and colleagues (2001) showed individuals with medically
unexplained fatigue are 10.9 times more likely to have a lifetime diagnosis of
depression than non-fatigued individuals, within the community. Additionally,
individuals with remitted, incident, and recurrent medically unexplained fatigue have
significantly increased risk of new onset depression (over a 13-year period)
compared to individuals who have never been fatigued (Addington et al., 2001). The
increased risk of depression in fatigued individuals indicates shared genetic factors
may contribute to the high levels of comorbidity observed between fatigue and
depression. Therefore, investigation of the genetic underpinnings of fatigue and
depression could provide insight into the observed comorbidity between the
disorders.
2.12.1 Heritability Links between Fatigue and Depression
The overlapping heritability of fatigue (or fatigue symptoms), psychological
disorders (including MDD and anxiety), disability pension due to mood and neurotic
diagnoses, chronic widespread pain, headaches, irritable bowel syndrome, and the
immune system, has been investigated within twin studies, using structural equation
modelling (Ball et al., 2010b; Ball et al., 2011; Burri et al., 2015; Fowler et al., 2006;
38 Chapter 2: Literature Review
Hickie et al., 1999a; Hickie et al., 1999b; Hickie et al., 2001; Hur et al., 2012; Kato
et al., 2009; Narusyte et al., 2016).
Bivariate Twin Modelling of Depression and Fatigue
In 2006, the overlapping heritability of depression (assessed by the mood and
feelings questionnaire (Costello & Angold, 1988)) with short-duration fatigue
(fatigue experienced for at least one week) was investigated in English and Welsh
children (Fowler et al., 2006). The authors only investigated the full ACE model,
which included two independent A components, two independent C, and two
independent E components (Table 2.13). The first A component explained 59% (95%
CI = 37-66%) and 8% (95% CI = 0-21%) of the variance in depression and short-
duration fatigue, respectively. Similarly, the first C component explained of the
variance in 0% (95% CI = 0-17%) and 13% (0-63%) of the variance in depression
and short-duration fatigue, respectively. Similarly, the first E component explained
41% (95% CI = 34-50%) and 1% (95% CI = 0-7%) of the variance in depression and
short-duration fatigue, respectively. The remaining 52% (95% CI = 0-80%), 0%
(95% CI = 0-63%), and 26% (95% CI = 11-49%) of the variance in short-duration
fatigue was explained by the second A, C, and E component, respectively.
Bivariate twin modelling was also utilised to investigate the overlapping
heritability of depression and prolonged fatigue, in English and Welsh children
(Fowler et al., 2006). The authors only investigated the full ACE model, which
included two independent A components, two independent C, and two independent E
components (Table 2.13). The first A component explained 59% (95% CI = 37-63%)
and 11% (95% CI = 0-28%) of the variance in depression and prolonged fatigue,
respectively. Similarly, the first C component explained 0% (95% CI = 0-17%) and
34% (95% CI = 0-75%) of the variance in depression and prolonged fatigue,
respectively. Meanwhile, the first E component explained 41% (95% CI = 34-49%)
and 1% (95% CI = 0-9%) of the variance in depression and prolonged fatigue,
respectively. The remaining 30% (95% CI = 0-81%), 0% (95% CI = 0-7%), and 24%
(95% CI = 9-51%) of the variance of prolonged fatigue was explained by the second
A, C, and E component, respectively.
In 2010, the overlapping heritability of fatigue (assessed by the Chalder fatigue
questionnaire (Chalder et al., 1993)) and a lifetime indicator of MDD (assessed by
the two screening questions of the CIDI Composite International Diagnostic
Chapter 2: Literature Review 39
Interview (World Health Organization, 1990) assessing the core symptoms of a
major depressive episode [i.e. depressed mood and anhedonia]) was investigated, in
twin pairs aged 15 years or older from Sri Lanka (Ball et al., 2010b). Although the
authors indicated the overlap in heritability was largely explained by unique
environmental factors, estimates for the specific variance components were not
provided so the bivariate genetic and environmental contribution within the study
cohort is unknown.
Trivariate twin modelling of depression, fatigue, and insomnia
In 2012, the overlap in heritability of depression, fatigue and insomnia was
investigated, in adult females from the United Kingdom (Hur et al., 2012). Common
and symptom-specific A and E components were estimated to explain the variance
within the traits (Table 2.12Table 2.12). Common A components explained 16%
(95% CI = 13-21%), 26% (95% CI = 21-32%), and 19% (95% CI = 15-24%) of the
variance in depression, fatigue, and insomnia, respectively. Similarly, common E
components explained 17% (95% CI = 13-21%), 27% (95% CI = 21-33%), and 19%
(95% CI = 15-24%) of the variance in depression, fatigue, and insomnia,
respectively. The remaining 18% (95% CI = 14-23%) and 49% (95% CI = 44-53%)
of the variation in depression was explained by symptom-specific A and E factors,
respectively. Similarly, the remaining 11% (95% CI = 6-16%) and 36% (95% CI =
30-40%) of the variation in fatigue was explained by symptom-specific A and E
factors, respectively. Finally, the remaining 11% (95% CI = 6-16%) and 51% (95%
CI = 45-50%) of the variation in insomnia was explained by symptom-specific A and
E factors, respectively.
Multivariate modelling of psychological distress, anxiety, depression, and
fatigue
In 1999, multivariate twin modelling was utilised to determine the overlap in
heritability of psychological distress, anxiety, depression, and fatigue, in adults from
Australia (Hickie et al., 1999b). Three independent A components and four
independent E components were identified which explained the variation in the traits
(Table 2.13). The first A component explained 36%, 23%, 25%, and 20% of the
variation in psychological distress, anxiety, depression, and fatigue, respectively.
Similarly, the second A component explained 11%, 9%, and 5% of the variation in
anxiety, depression, and fatigue, respectively. Meanwhile, the third A component
40 Chapter 2: Literature Review
explained 20% of the variation in fatigue. The first E component explained 64%,
24%, and 30% of the variation in psychological distress, anxiety, and depression,
respectively. Similarly, the second E component explained 42% and 8% of the
variation in anxiety and depression, respectively. Furthermore, the third E component
explained the remaining 28% of the variation in depression. Finally, the fourth E
component explained the remaining 55% of the variation in fatigue.
Trivariate modelling of psychological symptoms, fatigue symptoms, and
somatic symptoms
In 2011, the overlap in heritability of psychological symptoms, fatigue symptoms,
and somatic symptoms was investigated in males and females separately, from a Sri
Lankan cohort aged 15-85 (Ball et al., 2011). Common and symptom-specific A and
E components were estimated to explain the variance within the traits (Table 2.12).
In males, common A components explained 5%, 14%, and 11% of the
variation in psychological symptoms, fatigue symptoms, and somatic symptoms,
respectively. Similarly, common C components explained 4%, 13%, and 10% of the
variation in psychological symptoms, fatigue symptoms, and somatic symptoms,
respectively. While, common E components explained 9%, 30%, and 24% of the
variation in psychological symptoms, fatigue symptoms, and somatic symptoms,
respectively. The remaining 4%, 19%, and 59% of the variation in psychological
symptoms was explained by specific A, C, and E components, respectively. The
remaining 4% and 39% of the variation in fatigue symptoms was explained by
specific A and E factors, respectively. Finally, remaining 10% and 45% of the
variation of somatic symptoms was explained by specific A and E factors,
respectively.
In females, common A components explained 3%, 7%, and 6% of the variation
in psychological symptoms, fatigue symptoms, and somatic symptoms, respectively.
Common C factors explained 8%, 21% and 19% of the variation in psychological
symptoms, fatigue symptoms, and somatic symptoms, respectively. Common E
components explained 8%, 21%, and 18% of the variation in psychological
symptoms, fatigue symptoms, and somatic symptoms, respectively. The remaining
14%, 3%, and 64% of the variation in psychological symptoms was explained by
specific A, C, and E factors. Specific E components explained the remaining 51% of
the variation in fatigue symptoms. Finally, the remaining 22% and 35% of the
Chapter 2: Literature Review 41
variation in somatic symptoms was explained by specific A and E factors,
respectively.
Trivariate modelling of MDD/generalised anxiety disorder, CF and disability
pension due to neurotic diagnoses
In 2016, the overlap in heritability of MDD or generalised anxiety disorder, CF, and
disability pension due to mental diagnoses was investigated, in adult females from
Sweden (Narusyte et al., 2016). Three independent A components and three
independent E components were identified which explained the variation in the traits
(Table 2.13). The first A component explained 42% (95% CI = 41-48), 15% (95% CI
= 12-16), and 14% (95% CI = 13-27%) of the variance in MDD or generalised
anxiety disorder, CF, and disability pension due to mental diagnoses, respectively.
Similarly, the second A component explained 27% (95% CI = 14-38%) and 0%
(95% CI = 0-9%) of the variation in CF and disability pension due to mental
diagnoses, respectively. The last A component explained 31% (95% CI = 30-48) of
the variation in disability pension due to mental diagnoses. Meanwhile, the first E
component explained 58% (95% CI = 52-61%), 1% (95% CI = 0-4%), and 3% (95%
CI = 2-4%) of the variation in MDD or generalised anxiety disorder, CF, and
disability pension due to mental diagnoses, respectively. Similarly, the second E
component explained 57% (95% CI = 48-69%) and 6% (95% CI = 3-15%) of the
variation in CF and disability pension due to mental diagnoses, respectively. Finally,
the third E component explained the remaining 46% (95% CI = 30-50%) of the
variation in disability pension due to mental diagnoses.
Multivariate modelling of dehydroepiandrosterone sulfate, fatigue,
depression, and chronic widespread musculoskeletal pain
In 2015, multivariate twin modelling was utilised to determine the overlap in
heritability of dehydroepiandrosterone sulfate (an endogenous androstane steroid
hormone that confers a reduced risk of developing depressive symptoms over a four-
year period, when high levels are observed at baseline in blood samples (Souza-
Teodoro et al., 2016)), fatigue, depression, and chronic widespread musculoskeletal
pain, was assessed in adult females from the UK (Burri et al., 2015). Three
independent A components and four independent E components were identified
which explained the variation in the traits (Table 2.13). The first A component
explained 80%, 7%, and 6% of the variation in dehydroepiandrosterone sulfate,
depression, and chronic widespread musculoskeletal pain, respectively. Similarly, the
42 Chapter 2: Literature Review
second A component explained 40% and 40% of the variation in fatigue and chronic
widespread musculoskeletal pain, respectively. Meanwhile, the third A component
explained 31% and 20% of the variation in depression and chronic widespread
musculoskeletal pain, respectively. The first E component explained 20% and 15%
of the variation in dehydroepiandrosterone sulfate and depression, respectively.
Similarly, the second E component explained 60% and 9% of the variation in fatigue
and chronic widespread musculoskeletal pain, respectively. Furthermore, the third E
component explained the remaining 38% of the variation in depression. Finally, the
fourth E component explained the remaining 34% of the variation in chronic
widespread musculoskeletal pain.
Multivariate modelling of MDD, generalised anxiety disorder, headaches,
irritable bowel syndrome, CF, and chronic widespread pain
In 2009, multivariate twin modelling was utilised to determine the overlap in
heritability of MDD, generalised anxiety disorder, headaches, irritable bowel
syndrome, CF, and chronic widespread pain, was assessed in adults from Sweden
(Kato et al., 2009). Within this study, it was determined that two latent
(unobservable) phenotypes were shared by the traits (Table 2.12). It was estimated
that 73% and 27% of the variation in the first latent phenotype was explained by A
and E factors, respectively. Similarly, 44% and 56% of the variation in the second
latent phenotype was explained by A and E factors, respectively. The first latent
phenotype was shared by all six traits investigated. The common A component from
the first latent phenotype explained 42%, 34%, 5%, 7%, 12%, and 5% of the
variation in MDD, generalised anxiety disorder, headaches, irritable bowel
syndrome, CF, and chronic widespread pain, respectively. Furthermore, the common
E component from the first latent phenotype explained 16%, 13%, 2%, 2%, 4%, and
2% of the variation in MDD, generalised anxiety disorder, headaches, irritable bowel
syndrome, CF, and chronic widespread pain, respectively. Meanwhile the second
latent phenotype was only shared by four of the traits. The common A component
from the second latent phenotype explained 12%, 15%, 20%, and 31% of the
variation in headaches, irritable bowel syndrome, CF, and chronic widespread pain,
respectively. Similarly, the common E component from the second latent phenotype
explained 15%, 20%, 26%, and 39% of the variation in headaches, irritable bowel
syndrome, CF, and chronic widespread pain, respectively. Additionally, symptom-
specific A and E factors explained the remaining variation within the six traits. The
Chapter 2: Literature Review 43
remaining 42% of the variation in MDD was explained by specific E factors.
Similarly, the remaining 53% of the variation in generalised anxiety disorder was
explained by specific E factors. Headache specific A and E factors explained 24%
and 42% of the variation, respectively. Similarly, A and E factors specific to irritable
bowel syndrome specific explained 8% and 48% of the variation, respectively.
Furthermore, A and E factors specific to CF explained 9% and 29% of the variation,
respectively. Finally, A and E factors specific to chronic widespread pain explained
16% and 7% of the variation, respectively.
Trivariate modelling of psychological distress, fatigue, and immune
responsiveness
In 1999, the overlap in heritability between psychological distress, fatigue, and
immune responsiveness, was assessed in Australian adults (Hickie et al., 1999a).
Two independent A components, one C component, and three independent E
components were estimated to explain the variation between the phenotypes (Table
2.13). The first A component explained 61%, 30%, and 9% of the variation in
psychological distress, fatigue, and immune responsiveness, respectively.
Meanwhile, the second A component explained 22% of the variation in fatigue. The
C component explained 21% of the variation in immune responsiveness. The first E
component explained 39% and 8% of the variation in psychological distress, and
immune responsiveness, respectively. Similarly, the second E component explained
48% and 2% of the variation in fatigue and immune responsiveness. Finally, the
remaining 60% of the variation in immune responsiveness was explained by the third
E component.
Multivariate modelling of fatigue and immunological factors
In 2001, the overlap in heritability between fatigue and the immunological factors
IL-4, IFN-ɣ, and sCD23, was assessed in adults from Australia (Hickie et al., 2001).
Four A components, one C component, and four E components were identified
which explained the variation in the phenotypes (Table 2.13). The first A component
explained 47%, 4%, 7%, and 9% of the variation in fatigue, IL-4, IFN-ɣ, and sCD23,
respectively. Notably, the remaining three A components only explained variation
within a single phenotype. The second A component explained 28% of the variation
in IL-4. The third A component explained 1% of the variation in IFN-ɣ. The fourth A
component explained 10% of the variation in sCD23. Meanwhile, the C component
44 Chapter 2: Literature Review
explained 35%, 39%, and 28% of the variation in IL-4, IFN-ɣ, and sCD23,
respectively. The first E component explained 53%, 2%, and 2% of the variation in
fatigue, IFN-ɣ, and sCD23, respectively. Similarly, the second E component
explained 33%, 7%, and 3% of the variation in IL-4, IFN-ɣ, and sCD23, respectively.
Furthermore, the third E component explained 44% and 20% of the variation in IFN-
ɣ and sCD23, respectively. Finally, the fourth E component explained the remaining
28% of the variation in sCD23.
In summary, results from previously published bivariate, trivariate, and
multivariate twin modelling studies indicate a genetic association likely exists
between varying fatigue classifications and depression; as well as other phenotypes,
such as immunological and psychological traits. However, differences in cohort age,
sex, ethnicity, phenotype classifications, and reported results means further
investigation is warranted. In particular, further work is required to investigate the
type of twin model that explains the comorbidity between fatigue and depression. Of
the eight previous bivariate, trivariate, and multivariate twin modelling studies, the
best-fitting model or the only model investigated was the common factor model
(Table 2.12) or the Cholesky decomposition (Table 2.13). Furthermore, the majority
of studies identified shared genetic factors between fatigue and depression, however,
some studies did not identify any genetic overlap between the traits and the
magnitude of the overlap varies between studies. One advantage of the Cholesky
decomposition is that it enables the genetic correlation between the investigated
phenotypes to be calculated. Within the previous studies which reported a Cholesky
decomposition and included a fatigue and depression phenotype the genetic
correlation range was 0-0.74 (depression and short-duration fatigue = 0.36 (Fowler et
al., 2006); depression and prolonged fatigue 0.53 (Fowler et al., 2006); depression
and fatigue = 0.74 (Hickie et al., 1999b); MDD/generalised anxiety disorder and
chronic fatigue = 0.60 (Narusyte et al., 2016); and depression and fatigue = 0 (Burri
et al., 2015)). Therefore, further investigation is required to characterise the
mechanism of the observed comorbidity between fatigue and depression.
Chapter 2: Literature Review 45
Table 2.12. Heritability estimates (and their 95% confidence intervals) of the unique additive genetic factors (A), common environmental factors (C), and unique
environmental factors (E) from previous trivariate and multivariate common factor twin models, which include a fatigue and depression phenotype.
Phenotypes Population Sex Number of twin pairs
(MZ / DZ)
Mean age ±
standard
deviation
(Age range)
Common between phenotypes Specific to an individual phenotype
Latent factor 1 Latent factor 2
A C E A E A C E
Multivariate modelling of major depressive disorder (MDD), generalised anxiety disorder, headaches, irritable bowel syndrome, chronic fatigue, and chronic widespread pain (Kato et al., 2009).
MDD
Sweden M & F 3260 / 8988 & 127 twin
pairs of unknown
zygosity
53.7 ± 5.7
(41-64)
42 - 16 0 0 0 - 42 Generalised anxiety disorder 34 - 13 0 0 0 - 53
Headaches 5 - 2 12 15 24 - 42
Irritable bowel syndrome 7 - 2 15 20 8 - 48 Chronic fatigue 12 - 4 20 26 9 - 29
Chronic widespread pain 5 - 2 31 39 16 - 7
Trivariate modelling of psychological symptoms, fatigue symptoms, and somatic symptoms (Ball et al., 2011).
Psychological symptoms
Sri Lanka
M
1805 twin pairs & 137 singleton twins
33.9 ± 13.4 (15-85)
5 4 9 - - 4 19 59
Fatigue symptoms 14 13 30 - - 4 0 39
Somatic symptoms 11 10 24 - - 10 0 45 Psychological symptoms
F
3 8 8 - - 14 3 64
Fatigue symptoms 7 21 21 - - 0 0 51
Somatic symptoms 6 19 18 - - 22 0 35 Trivariate twin modelling of depression, fatigue, and insomnia (Hur et al., 2012).
Depression
UK F 893 / 884 & 204
singleton twins
50
(18-81)
16 (13-21) - 17 (13-21) - - 18 (14-23) - 49 (44-53)
Fatigue 26 (21-32) - 27 (21-33) - - 11 (6-16) - 36 (30-40) Insomnia 19 (15-24) - 19 (15-24) - - 11 (6-16) - 51 (45-50)
Note: 95% confidence intervals are reported as described in the original studies therefore, if they are absent they were not reported in the original publication. Populations are of European decent unless otherwise
specified. M: male; F: female; M & F: male and female.
46 Chapter 2: Literature Review
Table 2.13. Heritability estimates (and their 95% confidence intervals) of the unique additive genetic factors (A), common environmental factors (C), and unique
environmental factors (E) from previous bivariate, trivariate, and multivariate Cholesky twin models, which include a fatigue and depression phenotype.
Phenotypes Population Sex
Number of
twin pairs
(MZ / DZ)
Mean age ±
standard
deviation
(Age range)
A1
(%)
C1
(%)
E1
(%)
A2
(%)
C2
(%)
E2
(%)
A3
(%)
C3
(%)
E3
(%)
A4
(%)
E4
(%)
Bivariate Twin Modelling of Depression and Fatigue (Fowler et al., 2006).
Depression
English and
Welsh M & F 1468 (8-17)
59 (37-66)
0 (0-17)
41 (34-50)
- - - - - - - -
Short-duration fatigue 8
(0-21)
13
(0-63)
1
(0-7)
52
(0-80)
0
(0-63)
26
(11-49) - - - - -
Depression 59
(37-63)
0
(0-17)
41
(34-49) - - - - - - - -
Prolonged fatigue 11
(0-28)
34
(0-75)
1
(0-9)
30
(0-81)
0
(0-7)
24
(9-51) - - - - -
Multivariate modelling of psychological distress, anxiety, depression, and fatigue (Hickie et al., 1999b).
Psychological distress
Australian M & F 533 / 471 61.9
(> 50)
36 - 64 - - - - - - - -
Anxiety 23 - 24 11 - 42 - - - - -
Depression 25 - 30 9 - 8 - - 28 - - Fatigue 20 - - 5 - - - - - 20 55
Trivariate modelling of major depressive disorder (MDD)/generalised anxiety disorder, CF and disability pension due to neurotic diagnoses (Narusyte et al., 2016).
MDD or generalised anxiety disorder
Sweeden F
1776 / 2358
& 1717
singleton twins
53.2 ± 5.7
(< 65)
42 (41-48)
- 58
(52-61) - - - - - -
Chronic fatigue 15
(12-16) -
1
(0-4)
27
(14-38) -
57
(48-69) - - -
Disability pension due to mental diagnoses 14
(13-27) -
3
(2-4)
0
(0-9) -
6
(3-15)
31
(30-48) -
46
(30-50)
Multivariate modelling of dehydroepiandrosterone sulfate, fatigue, depression, and chronic widespread musculoskeletal pain (Souza-Teodoro et al., 2016).
Dehydroepiandrosterone sulfate
UK F
219 / 324 &
642
singleton
twins
58.4 ± 11.1
(26-82)
80 - 20 - - - - - - - -
Fatigue - - - 40 - 60 - - - - -
Depression 7 - 15 - - - 31 - 38 - -
Chronic widespread musculoskeletal pain 6 - - 40 - 9 20 - - - 34
Trivariate modelling of psychological distress, fatigue, and immune responsiveness (Hickie et al., 1999a).
Psychological distress Australian M & F 79 / 45
46.6
(31-84)
61 - 39 - - - - - - Fatigue 30 - - 22 - 48 - - -
Immune responsiveness 9 - 8 - - 2 - 21 60
Table 2.13 footnote on page 47.
Chapter 2: Literature Review 47
Table 2.14. Continued Heritability estimates (and their 95% confidence intervals) of the unique additive genetic factors (A), common environmental factors (C), and unique
environmental factors (E) from previous bivariate, trivariate, and multivariate Cholesky twin models, which include a fatigue and depression phenotype.
Phenotypes Population Sex
Number of
twin pairs
(MZ / DZ)
Mean age ±
standard
deviation
(Age range)
A1
(%)
C1
(%)
E1
(%)
A2
(%)
C2
(%)
E2
(%)
A3
(%)
C3
(%)
E3
(%)
A4
(%)
E4
(%)
Multivariate modelling of fatigue and immunological factors (Hickie et al., 2001).
Fatigue
Australian M & F 79 / 45 46.9
47 - 53 - - - - - - - - IL-4 4 - - 28 35 33 - - - - -
IFN-ɣ 7 - 2 - 39 7 1 - 44 - -
cSD23 9 - 2 - 28 3 - - 20 10 28
Note: 95% confidence intervals are reported as described in the original studies therefore, if they are absent they were not reported in the original publication. Populations are of European decent unless otherwise specified. M: male; F: female; M & F: male and female.
48
Although high levels of comorbidity have consistently been observed between fatigue
and depression, the presenting symptoms of the phenotypes partially overlap. To
date, the differences in depression symptoms reported by fatigued individuals and
fatigue symptoms reported by depressed individuals has never been investigated.
Additionally, the contribution of the overlapping fatigue and depression symptoms to
the high comorbidity observed between the traits is unknown. In order to determine
the quantitative differences in presenting symptoms and the role of overlapping
symptoms the following chapter aimed to investigate the co-occurrence and
symptomatology of fatigue and depression.
Chapter 3: Co-occurrence and Symptomatology of Fatigue and Depression 49
Chapter 3: Co-occurrence and
Symptomatology of Fatigue and
Depression
This chapter comprises the following published article:
Corfield, E. C., Martin, N. G., & Nyholt, D. R. (2016). Co-occurrence and
symptomatology of fatigue and depression. Comprehensive Psychiatry, 71, 1-10.
doi:10.1016/j.comppsych.2016.08.004
50 Chapter 3: Co-occurrence and Symptomatology of Fatigue and Depression
QUT Verified Signature
QUT Verified Signature
Chapter 3: Co-occurrence and Symptomatology of Fatigue and Depression 51
3.1 ABSTRACT
Fatigue and depression are highly comorbid phenotypes with partially overlapping
symptoms. The main aims of the present study are to: i) identify the risk of current
fatigue and depression; ii) determine if the depression symptoms experienced by
individuals who are fatigued (N = 766) and non-fatigued (N = 1,849) are different;
and iii) identify if the fatigue symptoms experienced by depressed (N = 275) and
non-depressed (N = 2,340) individuals are different, in a community-based sample of
Australian twins aged over 50. Fatigue and depression symptom profiles and
classifications were generated using the Schedule of Fatigue and Anergia (SOFA);
the General Health Questionnaire; and the Delusions-Symptoms-States Inventory,
States of Anxiety and Depression questionnaires. The association between co-
occurring fatigue and depression was assessed using prevalence ratios. Differences in
the preponderance of fatigue and depression symptoms were assessed using logistic
regression modelling. Individuals with either fatigue or depression have an
approximately two-fold increased risk for comorbid presentation of both traits,
compared to the general population. Logistic regression analysis indicated fatigued
individuals were significantly more likely to report all of the Diagnostic and
Statistical Manual of Mental Disorders (DSM) depression symptoms assessed in the
study. Similarly, depressed individuals were significantly more likely to report all
SOFA fatigue symptoms. Fatigue and depression are highly correlated traits within
the community. Depression symptomatology and prevalence are significantly
increased in fatigued individuals. Fatigue and especially the symptoms of insomnia
and poor concentration are strong predictors of depression. Notably, the association
between fatigue and depression is independent of their overlapping symptomatology.
Therefore, presentation with fatigue, and in particular the symptoms of insomnia and
poor concentration, should be considered as warning signs of depression in older
adults.
52 Chapter 3: Co-occurrence and Symptomatology of Fatigue and Depression
3.2 INTRODUCTION
Fatigue is a multidimensional symptom, which is highly prevalent in medical
practice, and difficult to quantify (Kroenke et al., 1988). Numerous classifications
exist for fatigue, which are based on arbitrary durations and severities, as a result of
its continuous nature (Wessely et al., 1997). Fatigue is associated with numerous
physical and psychiatric diagnoses, potentially due to the physical, cognitive, and
emotional dimensions the symptoms comprise (Arnold, 2008). Causation of fatigue
has been associated with numerous predisposing, precipitating, and perpetuating
factors (Sharpe & Wilks, 2002). A common predisposing factor is sex; with females
1.5 times as likely to experience fatigue as males (Chen, 1986). Additionally,
increased age has been associated with fatigue, in both males and females (Loge et
al., 1998). Comparison of fatigue symptoms based on sex has found females report a
higher prevalence of tiring easily and needing rest (David et al., 1990). However,
knowledge of the biological mechanisms underlying fatigue, which could account for
the differences between the sexes, is limited. Reduced health outcomes and quality of
life are associated with fatigue, which is commonly linked to psychiatric disorders,
particularly major depressive disorder (MDD) (Kroenke et al., 1994; Lyon et al.,
2014).
MDD is classified according to the Diagnostic and Statistical Manual of
Mental Disorders (DSM), which requires the presence of at least one major
depressive episode (American Psychiatric Association, 2013). The criterion for a
major depressive episode requires a two-week period where at least five of nine
symptoms are exhibited and either depressed mood or anhedonia (an inability to feel
pleasure in normally pleasurable activities) is reported. The symptoms of a major
depressive episode are: 1) depressed mood, 2) anhedonia, 3) a change in weight or
appetite, 4) insomnia (difficulty sleeping) or hypersomnia (excessive sleeping), 5)
psychomotor (i.e., thought and physical movement) agitation or retardation, 6)
fatigue or loss of energy, 7) feelings of worthlessness or excessive guilt, 8) inability
to concentrate or make decisions, and 9) thoughts about death, suicidal thoughts,
suicidal plans, or suicidal attempts (American Psychiatric Association, 2013).
thoughts about death, suicidal thoughts, suicidal plans, or suicidal attempts
(American Psychiatric Association, 2013). Minor depressive disorder (MiDD) is also
classified using the criterion for a major depressive episode (American Psychiatric
Chapter 3: Co-occurrence and Symptomatology of Fatigue and Depression 53
Association, 2000, 2013). However, only two to four symptoms occurring over a
two-week period are required for diagnosis, of which at least depressed mood or
anhedonia must be exhibited. Differences in the prevalence of depression occur over
the lifespan, with the prevalence increasing from puberty before declining after the
age of approximately 60 years (Centers for Disease Control and Prevention, 2010;
Kessler et al., 2003). The preponderance of depression in females has been
frequently investigated with numerous risk factors attributed to the increased
prevalence observed compared to males (Kuehner, 2003).
Investigation of differences in depression symptom prevalence (assessed using
the Composite International Diagnostic Interview) of individuals with depression in
Sri Lanka based on sex, revealed males report more hypersomnia and fewer thoughts
about death than females (Ball et al., 2010a). Furthermore, in the Netherlands, males
reported increased levels of anhedonia and psychomotor symptoms, while females
reported higher levels of mid-nocturnal insomnia, increases in weight, and somatic
complaints (the depression symptoms were assessed by the 30 item Inventory of
Depressive Symptomatology) (Schuch et al., 2014). Depression symptom profiles
have been investigated in individuals with seasonal affective disorder, and
differential symptoms have been identified among patients with unipolar, bipolar I,
and bipolar II depression (Goel et al., 2002). Finally, individuals with depression
were able to be distinguished from those with Alzheimer’s disease based on items
from three depression scales using regression modelling (Purandare et al., 2001).
Identification of differential symptoms between disorders facilitates increased
accuracy of diagnosis, thereby enabling utilisation of the most effective treatment
options. Medically unexplained symptoms are associated with depressive disorders
in 50-75% of patients (Kroenke, 2003). Furthermore, fatigue or loss of energy is the
second most frequently reported criterion of the DSM classification, experienced by
87.2% of MDD patients (Zimmerman et al., 2015). The co-occurrence of fatigue and
depression is likely due, in part, to their overlapping symptomatology. Therefore,
identification of symptoms which enable differential diagnosis would assist
physicians in distinguishing between fatigue and depression, thereby facilitating
symptom-guided management (Rosenthal et al., 2008).
Depression has a polythetic definition—whereby categorical diagnosis occurs
based on an arbitrarily defined threshold of symptoms being reached from a specified
54 Chapter 3: Co-occurrence and Symptomatology of Fatigue and Depression
criteria list, of which not all are required; therefore, the DSM classification is highly
heterogeneous, enabling a diagnosis of MDD in patients with entirely different
symptom profiles (Krueger & Bezdjian, 2009). The minimum requirement of 5
symptoms, of which at least one is depressed mood or anhedonia, enables 227
potential symptom profiles and allows the diagnosis of MDD in a subgroup of
individuals who are non-fatigued (Østergaard et al., 2011).
Fatigue and depression are highly comorbid, with fatigued individuals
reporting higher levels of depression than the general population (Cathébras et al.,
1992; Walker et al., 1993). Individuals with medically unexplained fatigue are
approximately 11 times (OR = 10.9) more likely to have a lifetime diagnosis of
depression than non-fatigued individuals, within the community (Addington et al.,
2001). Furthermore, the prevalence of co-occurring fatigue and psychological
distress within primary care is approximately 23% (Van Der Linden et al., 1999).
The high prevalence of comorbid fatigue and depression and idiopathic fatigue cases
often results in fatigue being perceived as a purely psychological symptom.
However, a subgroup of fatigued individuals exists which are not depressed (Harvey
et al., 2009; Hickie et al., 1999c; Van Der Linden et al., 1999). Although,
longitudinally (over a thirteen year period), individuals with remitted (relative risk
[RR] = 4.5), incident (RR = 53.2), and recurrent (RR = 28.4) medically unexplained
fatigue have significantly increased risk of new onset depression compared to
individuals who have never been fatigued (RR = 1.0) (Addington et al., 2001).
Therefore, understanding the relationship between fatigue and depression is vital to
facilitating diagnosis and enhanced treatment outcomes.
Initially, the present study will investigate the risk of co-occurring fatigue and
depression. Logistic regression modelling will then be utilised to determine if the
proportion of specific depression symptoms differs between individuals who are
fatigued and non-fatigued. Furthermore, the full symptom model will be investigated
to identify the distinguishing depression symptoms between fatigued and non-
fatigued individuals. The same approach will be utilised to assess if differential
fatigue symptoms are experienced by depressed and non-depressed individuals.
Finally, these analyses will identify the specific symptoms most strongly associated
with comorbid fatigue and depression.
Chapter 3: Co-occurrence and Symptomatology of Fatigue and Depression 55
3.3 METHODS
3.3.1 Sample and Questionnaires
Data from the over 50’s (aged) study conducted by the Genetic Epidemiology group
within QIMR Berghofer Medical Research Institute (QIMRB) was used in this study.
Informed written consent was obtained from each participant and the study was
approved by the Human Research Ethics Committee (HREC) of QIMRB. The study
was conducted from 1993 to 1996, with 2,281 twin pairs from the Australian Twin
Registry aged over 50 asked to complete a mailed Health and Lifestyle Questionnaire
(Bucholz et al., 1998; Mosing et al., 2012). The survey contained numerous self-
report questionnaires, of which the Schedule of Fatigue and Anergia (SOFA), the
twelve-item General Health Questionnaire (GHQ), and the fourteen-item Delusions-
Symptoms-States Inventory, States of Anxiety and Depression (DSSI/sAD), were
used throughout this study (Bedford & Deary, 1997; Goldberg & Blackwell, 1970;
Hickie et al., 1996).
The SOFA was originally designed to identify chronic fatigue syndrome cases.
Therefore, physical (i.e., muscular pain or tiredness), neurocognitive (i.e., memory
and concentration problems), and neurovegetative (i.e., sleep problems) fatigue
symptoms are assessed by the questionnaire. Consequently, the fatigued state
identified by the SOFA is distinct from the fatigue experienced within a major
depressive episode. Ten questions are contained in the SOFA; however, a shorter
eight-item version was included in the survey due to two questions being replicated
within the GHQ. The SOFA questions contained within the survey had a binary
yes/no response set, which was scored as 1-0. Throughout the GHQ there are two
response sets: 1) “not at all”, “no more than usual”, “rather more than usual”, and
“much more than usual”; and 2) “more so than usual”, “same as usual”, “less than
usual”, and “much less than usual”. Standard scoring of 0-0-1-1 was used for both
response sets of the GHQ. Responses to the DSSI/sAD questionnaire were
dichotomised, with the scores 0-0-1-1 representing the answers “not at all”, “a little”,
“a lot”, and “unbearably”, respectively.
Responses to the eight SOFA items and the two overlapping GHQ questions
(Table 3.1) were summed to give an overall score out of ten, which was used to
assess fatigue. Individuals with three or more positive self-report responses were
classified as fatigued. Previous studies have utilised shortened versions of the SOFA
56 Chapter 3: Co-occurrence and Symptomatology of Fatigue and Depression
scale to reliably assess fatigue in community cohorts (Bennett et al., 2004; Hickie et
al., 1999b; Kirk et al., 1999a; Kirk et al., 1999b). Furthermore, similar modifications
of the SOFA have been used and shown to have internal consistency (Tritt et al.,
2010). Although the current modified scale, utilising two questions from the GHQ
has not previously been used, 97.1% of the study cohort would have received the
same fatigue classification if the 8-item SOFA scoring had been used. The use of the
2 GHQ items enabled us to assess the full range of fatigue symptoms originally
included in the SOFA.
Table 3.1. Questionnaire items used to assess fatigue.
Abbreviated fatigue symptom Questionnaire and
question number Questiona
Muscle pain at rest SOFA 10 I get muscle pain even at rest Post-exertional muscle pain SOFA 6 I get muscle pain after physical activity
Post-exertional muscle fatigue SOFA 3 My muscles feel tired after physical activity
Post-exertional fatigue SOFA 1 I feel tired for a long time after physical activity Hypersomnia SOFA 5 I need to sleep for long periods
Insomnia GHQ 2 Lost much sleep over worry
Poor concentration GHQ 1 Been able to concentrate on what you’re doing Speech problems SOFA 8 I have problems with my speech
Poor memory SOFA 9 My memory is poor
Headaches SOFA 4 I get headaches
SOFA: Schedule of Fatigue and Anergia; GHQ: General Health Questionnaire. aParticipants were asked to respond with relation to their health, in general, over the past few weeks
MDD and MiDD were classified using the nine criteria of a major depressive
episode, as defined by the DSM (version IV) criteria (American Psychiatric
Association, 2000). A combination of questions from the GHQ and DSSI/sAD were
used to assess depression (Table 3.2), through the assignment of specific questions to
the appropriate criterion of the major depressive episode criteria. If a question did not
assess any of the criteria of a major depressive episode it was not used in the
analysis. When multiple questions assessed a criterion at least one positive response
indicated the individual exhibited a symptom from the specific criterion. Each
criterion was assessed by assigning one to the criterion if a symptom was exhibited
by the individual and zero if none of the symptoms for the criterion were met. The
survey did not contain any assessment of change in weight or appetite; therefore, the
third criterion of a major depressive episode (“a change in weight or appetite”) was
not assessed. The scores of the eight criteria assessed were summed if the individual
scored positively on criteria 1) or 2), otherwise the individual was assigned a score of
zero. Individuals were designated MDD, MiDD, or non-depressed, if they had a self-
report score of five or more, two to four, or less than two, respectively. The
combination of GHQ and DSSI/sAD, used to map the self-reported depression
Chapter 3: Co-occurrence and Symptomatology of Fatigue and Depression 57
symptoms to the DSM criteria has not previously been used. However, 93.4% of the
cohort would have received the same depression classification if the standard
DSSI/sAD scoring had been used (which is a valid and reliable measure of
depression). Furthermore, utilisation of DSM symptomatic criteria enabled us to
investigate minor and major depression cases which would be impossible using the
DSSI/sAD measure of depression.
58 Chapter 3: Co-occurrence and Symptomatology of Fatigue and Depression
Table 3.2. Questionnaire items used to assess the criteria of a major depressive episode.
DSM Major depressive episode
criteria
Questionnaire and
question number Questiona
Depressed mood GHQ 9 Been feeling unhappy and depressed
DSSI/sAD 5 Recently, I have been depressed without knowing why Anhedonia GHQ 7 Been able to enjoy your normal day-to-day activities
DSSI/sAD 12 Recently, I have lost interest in just about everything
Insomnia DSSI/sAD 2 Recently, I have been so miserable that I have had difficulty with my sleep DSSI/sAD 11 Recently, worrying has kept me awake at night
Psychomotor agitation DSSI/sAD 4 Recently, I have been so ‘worked up’ that I couldn’t sit still
Loss of energy DSSI/sAD 8 Recently, I have been so low in spirits that I have sat for ages doing absolutely nothing Feeling worthless GHQ 3 Felt that you are playing a useful part in things
GHQ 6 Felt that you couldn’t overcome your difficulties
GHQ 11 Been thinking of yourself as a worthless person DSSI/sAD 10 Recently, the future has seemed hopeless
Inability to concentrate GHQ 4 Felt capable of making decisions about things
DSSI/sAD 13 Recently, I have been so anxious that I couldn’t make up my mind about the simplest thing Suicidal thoughts DSSI/sAD 6 Recently, I have gone to bed not caring if I never woke up
DSSI/sAD 14 Recently, I have been so depressed that I have thought of doing away with myself
DSM: Diagnostic and Statistical Manual of Mental Disorders; GHQ: General Health Questionnaire; DSSI/sAD: Delusions-Symptoms-States Inventory, States of Anxiety and
Depression. aParticipants were asked to respond with relation to their health, in general, over the past few weeks.
Chapter 3: Co-occurrence and Symptomatology of Fatigue and Depression 59
3.3.2 Statistical Analysis
Prevalence Ratios
The association between fatigue and depression was investigated using contingency
tables to assess the prevalence of co-occurrence within the cohort. The likelihood of
a fatigued individual having comorbid depression compared to non-fatigued
individuals and the total cohort was assessed using the prevalence ratio (PR) measure
of association and its 95% confidence interval (CI). The PR has the same
interpretation as the relative risk (RR) with respect to its null value of 1 and values
greater or less than 1. The PR is the ratio of the prevalence rate in one group divided
by the prevalence rate in a second group. For example, the prevalence of depression
in fatigued individuals was divided by the prevalence of depression in non-fatigued
individuals. Similarly, the prevalence of fatigue in depressed individuals was divided
by the prevalence of fatigue in non-depressed individuals. To assist interpretation of
the numerous PR estimates, we also calculated PRs for specific groups relative to the
total sample, by dividing the prevalence in the specific group by the prevalence in
total sample.
The fatigued individuals likelihood of experiencing depression was re-
calculated in the subgroup of individuals without (screening negative for)
overlapping DSM depression symptoms (i.e., insomnia, poor concentration, and
hypersomnia) to remove the effect of overlapping symptoms. Likewise, the
likelihood of depressed individuals experiencing fatigue was re-calculated in the
subgroup of individuals without (screening negative for) fatigue symptoms (i.e.,
insomnia, inability to concentrate, and loss of energy).
Multiple Test Correction
The matrix spectral decomposition (matSpD) web-based tool
(http://neurogenetics.qimrberghofer.edu.au/matSpD/) estimates the effective number
of independent variables from a pairwise correlation matrix (Cheverud, 2001; Li &
Ji, 2005; Nyholt, 2004; R Development Core Team, 2003). To identify the effective
number of independent variables, matSpD analyses the eigenvalues of the correlation
matrix (after spectral decomposition—factorisation of a matrix into a canonical
form). Briefly, to retain an experiment-wide type I error rate of 5%, the significance
thresholds for analysing the full set of fatigue and depression symptom measures was
calculated by dividing the nominal significance threshold of p-value = 0.05, by the
60 Chapter 3: Co-occurrence and Symptomatology of Fatigue and Depression
effective number of independent measures estimated by matSpD analysis of the
pairwise correlation matrix calculated using R (R Core Team, 2014) for the fatigue
and depression symptom measures.
Logistic Regression Modelling
Demographic differences in age and sex, with respect to fatigue and depression
classification, were initially assessed by logistic regression in R (R Core Team,
2014).
Binomial logistic regression modelling was used to compare the depression
symptoms between the fatigued and the non-fatigued groups (R Core Team, 2014).
The depression symptoms were assessed individually (univariable analysis) and as
part of the full model (multivariable analysis) containing all eight symptoms. The
Akaike information criterion (AIC) was used to assess the parsimony of the
depression symptom model compared to the null model, with lower AIC indicating a
better fit (Akaike, 1973, 1974). To account for relatedness, an exchangeable
conditional covariance matrix was used (i.e., we allowed for correlated residuals
between members of the same family) and tests were based on the robust (sandwich-
corrected) standard errors, using the rms package in R (R Core Team, 2014).
Analysis of deviance containing the chi-squared test was used to assess statistical
differences between the logistic regression of the fatigued and non-fatigued groups.
The eight depression symptoms were compared between the fatigued and non-
fatigued groups using a two-tailed p-value and odds ratio (OR) with their 95% CI.
The approach was replicated to compare the ten fatigue symptoms between
depressed and non-depressed individuals. Additionally, ordinal logistic regression,
using rms, was utilised to compare the use of a broad, two-category depression
classification (non-depressed, MiDD/MDD) to an ordered three-category depression
classification (non-depressed, MiDD, MDD).
To obtain subgroup specific odds ratios, multinomial logistic regression
modelling was used to compare the fatigue symptoms between the MDD, MiDD, and
non-depressed groups. Relatedness was not accounted for due to its negligible effect
on the binomial logistic regression results. The fatigue symptoms were assessed
individually and as part of the full model. Multinomial regression modelling
conducted throughout the study followed the protocol defined by Morris et al.
(2010), using the mlogit package within R (Morris et al., 2010; R Core Team, 2014).
Chapter 3: Co-occurrence and Symptomatology of Fatigue and Depression 61
Parsimony of the model was assessed using the AIC and statistical differences
between the MDD, MiDD, and non-depressed groups were identified using analysis
of deviance containing the chi-squared test. The fatigue symptoms were compared
between the depression groupings using a two-tailed p-value and OR with their 95%
CI. Fatigue and depression symptoms which were significantly different were
identified using the thresholds obtained from matSpD.
3.4 RESULTS
3.4.1 Study Population
The over 50’s (aged) study consisted of 4,562 participants. However, 1,947
individuals returned incomplete responses to SOFA, GHQ, and/or DSSI/sAD
questionnaire items utilised to assess depression and fatigue in the present study and
were therefore excluded. The remaining 2,615 individuals with complete responses
comprised the study cohort which was used in all analyses (Table 3.3).
Supplementary Table 3.1 lists the number of individuals reporting each specific
symptom. The study cohort (including 496 complete monozygotic twin pairs, 440
complete dizygotic twin pairs, 5 complete twin pairs of unknown zygosity, and 733
unpaired twin singles), had a mean age of 60.5 years (range = 50-92), which was not
significantly different from the non-responders. As typically found, significantly
higher response rates (p < 2 × 10-16) were observed for females (71.9%) compared to
males (58.7%).
Depressed individuals had a two-fold (PR = 2.18, 95% CI = 1.96-2.43)
increase in risk of fatigue, compared to the total population sample. Stratification of
depressed individuals revealed the increased risk of fatigue was slightly (although
not significantly) higher in MDD cases (PR = 2.32, 95% CI = 1.90-2.83) compared
to individuals with MiDD (PR = 2.15, 95% CI = 1.92-2.42). Meanwhile, non-
depressed individuals had a reduced risk of fatigue (PR = 0.86, 95% CI = 0.79-0.94).
Significantly, depressed individuals risk of fatigue was significantly increased,
independent of insomnia, concentration problems, and hypersomnia (PR = 2.27, 95%
CI = 1.51-3.39). Similarly, fatigued individuals had a two-fold (PR = 2.18, 95% CI =
1.84-2.59) increased risk of depression, compared to the total population sample.
Furthermore, stratification of fatigued individuals risk of depression revealed
fatigued individuals had a slightly (although not significantly) higher risk of MDD
62 Chapter 3: Co-occurrence and Symptomatology of Fatigue and Depression
(PR = 2.32, 95% CI = 1.51-3.56) than MiDD (PR = 2.15, 95% CI = 1.77-2.62).
Meanwhile, non-fatigued individuals had a reduced risk of depression (PR = 0.51,
95% CI = 0.41-0.64). Notably, fatigued individuals risk of depression was
significantly increased, independent of insomnia, concentration problems and loss of
energy (PR = 2.27, 95% CI = 1.33-3.86).
Interestingly, the risk of depression (PR = 4.29, 95% CI = 3.40-5.41) in
fatigued compared to non-fatigued individuals is approximately two-fold greater than
the risk of fatigue (PR = 2.54, 95% CI = 2.27-2.84) in depressed compared to non-
depressed individuals.
3.4.2 Fatigued Individuals Report a Higher Proportion of Depression Symptoms
The matSpD analysis indicated moderate intercorrelation between the eight
depression symptom measures, and estimated them to be equivalent to six effectively
independent measures. Therefore, to keep type I error rate at 5%, the significance
threshold used for univariable analysis of the eight depression symptoms was
adjusted for six independent tests (i.e., Bonferroni adjusted experiment-wide
significant threshold, p = 0.05 / 6 = 8.3 × 10-3).
Analysis of age and sex revealed no significant differences between fatigued
and non-fatigued individuals. Therefore, the age and sex variables were not included
as covariates in the logistic regression analysis of fatigue symptoms.
Notably, all eight depression symptoms were significantly different
(univariable p < 8.3 × 10-3) between fatigued and non-fatigued individuals (Table
3.4). Furthermore, the full logistic regression model (AIC = 2960.5) comparing
fatigued versus non-fatigued individuals, was more parsimonious than the null model
(AIC = 3164.8). Therefore, the results provided (in Table 3.4) are for the more
parsimonious model. Comparison of the fatigued and non-fatigued groups (Table
3.4) revealed an overall significant difference in depression symptoms (χ2 = 220.32,
p < 2.2 × 10-16). In particular, the proportion of fatigued cases reporting anhedonia,
insomnia, and feeling worthless, were significantly higher than non-fatigued
individuals.
Chapter 3: Co-occurrence and Symptomatology of Fatigue and Depression 63
Table 3.3. Prevalence ratios of fatigue and depression.
Counts (%) PR (95% CI) of fatigue PR (95% CI) of depression
Non-fatigued Fatigued Total Non-depresseda Totalb Non-fatiguedc Totald
All symptoms
Non-depressed 1750 (66.9) 590 (22.6) 2340 (89.5) NA 0.86 (0.79-0.94) NA 0.51 (0.41-0.64) Depressed 99 (3.8) 176 (6.7) 275 (10.5) 2.54 (2.27-2.84) 2.18 (1.96-2.43) 4.29 (3.40-5.41) 2.18 (1.84-2.59)
Total 1849 (70.7) 766 (29.3) 2615 (100.0)
Non-overlapping symptoms
Non-depressed 1531 (83.4) 253 (13.8) 1784 (97.2) NA 0.96 (0.82-1.13) NA 0.78 (0.51-1.20)
Depressed 34 (1.9) 17 (0.9) 51 (2.8) 2.35 (1.57-3.52) 2.27 (1.51-3.39) 2.90 (1.64-5.11) 2.27 (1.33-3.86) Total 1565 (85.3) 270 (14.7) 1835 (100.0)
All symptoms: all individuals; Non-overlapping symptoms: individuals without the fatigue and depression overlapping symptoms; PR: prevalence ratio; CI: confidence interval; NA: not applicable. aPrevalence ratio of
fatigue in depressed compared to non-depressed individuals. bPrevalence ratio of fatigue in depressed individuals compared to the total cohort. cPrevalence ratio of depression in fatigued compared to non-fatigued
individuals. dPrevalence ratio of depression in fatigued individuals compared to the total cohort.
Table 3.4. Logistic regression, unadjusted and and adjusted for, relatedness, comparing the depression symptoms exhibited by fatigued individuals (N = 766) to non-fatigued
(N = 1,849) individuals.
Depression symptomsa
Univariable Multivariableb
Unadjusted for relatedness Adjusted for relatedness Unadjusted for relatedness Adjusted for relatedness
OR (95% CI) p-value OR (95% CI) p-value OR (95% CI) p-value OR (95% CI) p-value
Depressed mood 3.88 (2.96-4.97) < 2.00 × 10-16 3.83 (2.96-4.98) < 2.00 × 10-16 1.42 (1.02-1.99) 0.04 1.42 (1.01-2.00) 0.04
Anhedonia 3.89 (3.04-4.98) < 2.00 × 10-16 3.89 (3.05-4.97) < 2.00 × 10-16 1.97 (1.46-2.65) 7.67 × 10-6 1.97 (1.46-2.66) 1.02 × 10-5 Insomnia 6.60 (4.36-9.98) < 2.00 × 10-16 6.60 (4.33-10.05) < 2.00 × 10-16 2.14 (1.29-3.53) 3.10 × 10-3 2.14 (1.23-3.73) 0.01
Psychomotor agitation 9.69 (4.96-18.93) 2.92 × 10-11 9.69 (5.02-18.71) 1.30 × 10-11 2.75 (1.26-6.00) 0.01 2.75 (1.23-6.16) 0.01
Loss of energy 4.47 (2.36-8.45) 4.11 × 10-6 4.47 (2.37-8.44) 3.95 × 10-6 0.72 (0.32-1.60) 0.42 0.72 (0.29-1.75) 0.47 Feeling worthless 4.25 (3.32-5.46) < 2.00 × 10-16 4.25 (3.33-5.43) < 2.00 × 10-16 2.12 (1.56-2.87) 1.22 × 10-6 2.12 (1.56-2.79) 1.24 × 10-6
Inability to concentrate 4.31 (3.00-6.20) 2.78 × 10-15 4.31 (3.02-6.16) 8.88 × 10-16 1.75 (1.14-2.68) 0.01 1.75 (1.10-2.79) 0.02
Suicidal thoughts 6.02 (3.06-11.87) 2.10 × 10-7 6.02 (3.08-11.77) 1.48 × 10-7 0.85 (0.38-1.93) 0.70 0.85 (0.35-2.05) 0.72
OR: odds ratio; CI: confidence interval. aDefined by the Diagnostic and Statistical Manual of Mental Disorders (DSM). bMultivariable model includes all 8 depression symptoms.
64 Chapter 3: Co-occurrence and Symptomatology of Fatigue and Depression
3.4.3 Depressed Individuals Report Higher Proportions of Fatigue Symptoms
The matSpD analysis of the ten fatigue symptom measures revealed minimal
intercorrelation being equivalent to nine effectively independent measures.
Therefore, to keep type I error rate at 5%, the significance threshold used for
analyses involving all the fatigue symptoms was 5.6 × 10-3 (p = 0.05 / 9).
Demographic analysis of the difference in age and sex revealed no significant
differences between depressed and non-depressed individuals. Therefore, the age and
sex variables were not included as covariates in the logistic regression analysis of
fatigue symptoms.
Interestingly, all ten fatigue symptoms were significantly different (univariable
p < 5.6 × 10-3) between depressed and non-depressed individual (Table 3.5).
Furthermore, the full symptom model (AIC = 1342.0) comparison of the fatigue
symptoms endorsed by depressed versus non-depressed individuals was more
parsimonious than the null model of no differences between the groups (AIC =
1760.7). The comparison revealed an overall significant difference in the depression
symptoms (χ2 = 438.77, p < 2 × 10-16) experienced by depressed and non-depressed
individuals (Table 3.5). In particular, the proportion of depression cases reporting
insomnia, poor concentration, and headaches, were significantly higher than non-
depressed individuals. Results were comparable between the binomial and ordinal
logistic regression modelling (Table 3.6).
Chapter 3: Co-occurrence and Symptomatology of Fatigue and Depression 65
Table 3.5. Logistic regression, both unadjusted and adjusted for relatedness, of fatigue symptoms exhibited by depressed (N = 275) and non-depressed (N = 2,340)
individuals.
Fatigue symptomsa
Univariable Multivariableb
Unadjusted for relatedness Adjusted for relatedness Unadjusted for relatedness Adjusted for relatedness
OR (95% CI) p-value OR (95% CI) p-value OR (95% CI) p-value OR (95% CI) p-value
Binomial logistic regression Muscle pain at rest 2.79 (2.04-3.80) 8.40 × 10-11 2.79 (2.05-3.79) 7.70 × 10-11 1.35 (0.88-2.07) 0.17 1.35 (0.84-2.18) 0.22
Post-exertional muscle pain 2.32 (1.80-2.99) 9.91 × 10-11 2.32 (1.79-3.00) 1.54 × 10-10 1.11 (0.77-1.60) 0.57 1.11 (0.76-1.63) 0.59
Post-exertional muscle fatigue 2.48 (1.93-3.19) 1.78 × 10-12 2.48 (1.92-3.21) 5.50 × 10-12 1.05 (0.71-1.54) 0.81 1.05 (0.70-1.56) 0.82 Post-exertional fatigue 3.46 (2.66-4.49) < 2.20 × 10-16 3.46 (2.65-4.51) < 2.00 × 10-16 1.77 (1.20-2.60) 3.70 × 10-3 1.77 (1.17-2.67) 0.01
Hypersomnia 2.31 (1.74-3.07) 6.24 × 10-9 2.31 (1.73-3.09) 1.83 × 10-8 1.05 (0.72-1.53) 0.79 1.05 (0.71-1.56) 0.80
Insomnia 11.68 (8.67-15.74) < 2.20 × 10-16 11.68 (8.62-15.83) < 2.00 × 10-16 8.08 (5.78-11.29) < 2.20 × 10-16 8.08 (5.68-11.50) < 2.20 × 10-16 Poor concentration 12.12 (8.95-16.41) < 2.20 × 10-16 12.12 (9.04-16.25) < 2.00 × 10-16 6.92 (4.85-9.87) < 2.20 × 10-16 6.92 (4.79-9.99) < 2.20 × 10-16
Speech problems 2.01 (1.48-2.72) 7.08 × 10-6 2.01 (1.48-2.73) 8.15 × 10-6 1.01 (0.67-1.52) 0.97 1.01 (0.67-1.51) 0.97
Poor memory 2.38 (1.79-3.16) 2.66 × 10-9 2.38 (1.78-3.18) 4.57 × 10-9 1.09 (0.74-1.62) 0.66 1.09 (0.72-1.66) 0.68 Headaches 2.78 (2.14-3.63) 3.73 × 10-14 2.78 (2.13-3.564) 1.01 × 10-13 1.76 (1.28-2.44) 5.92 × 10-4 1.76 (1.24-2.50) 1.41 × 10-3
OR: odds ratio; CI: confidence interval. aAssessed by the Schedule of Fatigue and Anergia (SOFA). bMultivariable model includes all 10 fatigue symptoms.
66 Chapter 3: Co-occurrence and Symptomatology of Fatigue and Depression
The analyses comparing MDD (N = 50), MiDD (N = 225), and non-depressed
(N = 2,340) individuals exhibited comparable trends to the results of the ‘complete’
depressed cohort. All ten fatigue symptoms were significantly different between
MiDD and non-depressed individuals (Table 3.6). Similarly, all the fatigue
symptoms except post-exertional muscle pain are significantly different between
MDD and non-depressed individuals.
Table 3.6. Logistic regression of fatigue symptoms exhibited by individuals with major depressive
disorder (N = 50), minor depressive disorder (N = 225), and are non-depressed (N = 2,340).
Fatigue symptoma Univariable Multivariableb OR (95% CI) p-value OR (95% CI) p-value
MiDD versus non-depressed
Muscle pain at rest 2.77 (1.98-3.88) 2.75 × 10-9 1.33 (0.85-2.06) 0.21
Post-exertional muscle pain 2.42 (1.83-3.20) 4.62 × 10-10 1.22 (0.84-1.79) 0.30
Post-exertional muscle fatigue 2.38 (1.80-3.13) 7.75 × 10-10 0.99 (0.66-1.48) 0.95 Post-exertional fatigue 3.35 (2.52-4.45) < 2.20 × 10-16 1.80 (1.21-2.70) 3.98 × 10-3
Hypersomnia 2.15 (1.58-2.94) 1.46 × 10-6 1.00 (0.68-1.48) 0.98
Insomnia 8.50 (6.12-11.81) < 2.20 × 10-16 6.08 (4.23-8.73) < 2.20 × 10-16
Poor concentration 10.03 (7.22-13.93) < 2.20 × 10-16 6.13 (4.21-8.92) < 2.20 × 10-16
Speech problems 1.89 (1.35-2.65) 2.03 × 10-4 0.96 (0.62-1.48) 0.86
Poor memory 2.35 (1.72-3.20) 8.16 × 10-8 1.15 (0.76-1.73) 0.51 Headaches 2.89 (2.17-3.86) 4.44 × 10-13 1.87 (1.34-2.61) 2.31 × 10-4
MDD versus non-depressed
Muscle pain at rest 2.84 (1.46-5.51) 2.01 × 10-3 1.53 (0.62-3.75) 0.36
Post-exertional muscle pain 1.91 (1.08-3.39) 0.03 0.63 (0.29-1.40) 0.26
Post-exertional muscle fatigue 3.00 (1.70-5.28) 1.40 × 10-4 1.51 (0.68-3.37) 0.31 Post-exertional fatigue 3.97 (2.26-5.27) 1.78 × 10-6 1.61 (0.71, 3.65) 0.26
Hypersomnia 3.11 (2.73, 7.00) 1.54 × 10-4 1.47 (0.67, 3.20) 0.34
Insomnia 48.38 (25.10, 93.26) < 2.20 × 10-16 33.64 (16.80, 67.34) < 2.20 × 10-16 Poor concentration 27.27 (15.05, 49.41) < 2.20 × 10-16 12.98 (6.54, 25.77) 2.41 × 10-14
Speech problems 2.58 (1.37, 4.83) 3.20 × 10-3 1.33 (0.59, 2.98) 0.49
Poor memory 2.53 (1.37, 4.68) 3.13 × 10-3 0.983 (0.36, 1.90) 0.66 Headaches 2.32 (1.28, 4.21) 0.01 1.20 (0.58, 2.48) 0.61
Ordinal logistic regression Muscle pain at rest 2.77 (2.04-3.77) 8.62 × 10-11 1.30 (0.86-1.98) 0.22
Post-exertional muscle pain 2.30 (1.79-2.97) 1.35 × 10-10 0.99 (0.69-1.42) 0.97
Post-exertional muscle fatigue 2.49 (1.93-3.20) 1.44 × 10-12 1.17 (0.81-1.69) 0.41 Post-exertional fatigue 3.46 (2.66-4.49) < 2.20 × 10-16 1.66 (1.14-2.41) 0.01
Hypersomnia 2.33 (1.76-3.09) 4.25 × 10-9 1.09 (0.76-1.56) 0.64
Insomnia 12.79 (9.49-17.23) < 2.20 × 10-16 8.72 (6.31-12.05) < 2.20 × 10-16 Poor concentration 12.57 (9.32-16.95) < 2.20 × 10-16 7.10 (5.03-10.01) < 2.20 × 10-16
Speech problems 2.02 (1.49-2.74) 5.61 × 10-6 1.04 (0.70-1.55) 0.84
Poor memory 2.38 (1.79-3.16) 2.56 × 10-9 1.00 (0.68-1.48) 0.99 Headaches 2.76 (2.12-3.59) 5.55 × 10-14 1.62 (1.18-2.22) 2.68 × 10-3
MiDD versus non-depressed: results from multinomial logistic regression analysis for MiDD subgroup compared to non-depressed group;
MDD versus non-depressed: results from multinomial logistic regression analysis for MDD subgroup compared to non-depressed group; Ordinal logistic regression: results from ordinal logistic regression analysis of the three (MDD, MiDD and non-depressed) subgroups; OR:
odds ratio; CI: confidence interval. aAssessed by the Schedule of Fatigue and Anergia (SOFA). bMultivariable (“full”) model includes all 10
fatigue symptoms.
The full fatigue symptom model (AIC = 1585.3) was more parsimonious than
the null model (AIC = 2023.5), comparing MDD, MiDD, and non-depressed
individuals. Comparison of the MDD and MiDD groups (Table 3.6) to the non-
depressed group revealed an overall significant difference in fatigue symptoms (χ2 =
478.28, p < 2.2 × 10-16). In particular, the proportion of MiDD cases reporting post-
exertional fatigue, insomnia, poor concentration, and headaches was significantly
higher than non-depressed individuals. Similarly, the proportion of MDD cases
Chapter 3: Co-occurrence and Symptomatology of Fatigue and Depression 67
reporting insomnia and poor concentration was higher than non-depressed
individuals.
3.5 DISCUSSION
The results demonstrate that individuals presenting with either fatigue or depression
have a two-fold increase in risk for a co-occurring presentation of both traits. The
risk of depression in fatigued individuals compared to non-fatigued individuals, is
two-fold greater than the risk of fatigue in depressed individuals compared to non-
depressed individuals, indicating that fatigue could be used as a predictor to facilitate
early detection of depression. This is particularly interesting considering that fatigue
severity has been identified as a good predictor of MDD within cancer patients
(Deckx et al., 2015). Although fatigue severity is subjective, the use of specific
fatigue symptoms might facilitate more accurate prediction of depression.
Significantly, fatigued individuals reported more depression symptoms than
non-fatigued individuals. These results are consistent with previous findings showing
fatigued individuals have higher depression levels (Cathébras et al., 1992; Walker et
al., 1993). However, a proportion of the fatigued individuals will not have comorbid
depression; although pure fatigue appears to be a dynamic state with numerous cases
exhibiting symptoms of psychological distress (Harvey et al., 2009; Van Der Linden
et al., 1999). That said, the analysis comparing non-depressed individuals with
depressed cases revealed significant differences for all ten fatigue symptoms.
Therefore, although fatigue and depression symptoms were reported in individuals
who were non-depressed and non-fatigued, respectively, the increased number of
symptoms exhibited by fatigued and depressed cases suggests an underlying
association.
Heritable associations have been identified between fatigue and depression
(Hickie et al., 1999b) in a twin sample that partially overlaps the present one. The
heritability of fatigue and depression are both estimated to have unique genetic and
environmental factors but no contribution of common environmental factors.
Multivariate twin modelling estimated a common additive genetic component
explained 36.0%, 23.3%, 25.0%, and 20.3% of the variance in psychological distress,
anxiety, depression, and fatigue, respectively. Moreover, a second common additive
genetic component explained 11.0%, 9.0%, and 5.1% of the variance in anxiety,
68 Chapter 3: Co-occurrence and Symptomatology of Fatigue and Depression
depression, and fatigue, respectively. Additionally, a third additive genetic
component (independent of psychological distress, anxiety, and depression) was
found to explain a further 20.3% of variance in fatigue. Furthermore, depression and
fatigue were both estimated to have independent unique environmental factors which
explained 28% and 54.3% of their variance, respectively (Hickie et al., 1999b).
Therefore, the observed comorbidity between fatigue and depression may be
explained, in part, by shared underlying genetically determined disease mechanisms.
Insomnia was assessed as both a fatigue and depression symptom. Therefore,
the identification of insomnia as a distinguishing symptom between fatigued and
non-fatigued individuals is unsurprising. Although poor concentration was also
assessed as a symptom of both fatigue and depression, it is not a distinguishing
symptom between fatigued and non-fatigued individuals. However, concentration
problems may not have reached significance in the full symptom model due to
differences in the wording of the fatigue and depression questions for its assessment
potentially resulting in different responses by individuals. Therefore, insomnia is a
key indicator of co-occurring fatigue and depression. Considering depression
diagnosis is particularly difficult within older adults, insomnia and to a lesser extent
poor concentration, should be considered as warning signs of depression. Indeed,
Deckx and colleagues have previously shown fatigue to be an indicator of depression
in older cancer patients (Deckx et al., 2015); whereas, our results demonstrate the
broader applicability of fatigue, and in particular insomnia, as an indicator of
depression within older adults in the community. Evidence for overlapping molecular
mechanisms between fatigue, depression, and insomnia has been provided by
heritability estimates within females (Hur et al., 2012). Common and symptom-
specific additive genetic and unique environmental factors were identified which
explain the variance of insomnia, fatigue, and depression. Therefore, overlapping
genetic factors could explain the high levels of insomnia in fatigued and depressed
individuals and potentially account for a proportion of the high comorbidity of
fatigue and depression.
The present study is the first to investigate both fatigue and depression
symptoms experienced by depressed and fatigued individuals, respectively. A
possible limitation of our study lies in the relatively small number of individuals with
MDD and inability to assess the third DSM criterion of a major depressive episode—
Chapter 3: Co-occurrence and Symptomatology of Fatigue and Depression 69
change in weight or appetite. Although re-running the analysis removing the small
proportion (6.4%) of non-depressed individuals who report either depressed mood or
anhedonia (and could therefore be depression cases if they reported a change in
weight or appetite) did not change the study findings (data not shown). Also, the
large number of individuals who did not complete the questions used throughout this
study could potentially be due to a reduced likelihood of depressed individuals
completing the survey. Although the increased age of the present cohort has possibly
contributed to the lower levels of depression observed, we note that comparable
prevalence estimates have been reported for individuals over 65 in the United States.
The Centers for Disease Control and Prevention (2010) reported the prevalence of a
current diagnosis of MDD and MiDD in adults at 4.1% and 5.1%, respectively,
compared to 2.1% and 4.8%, respectively, in individuals over 65 (Centers for
Disease Control and Prevention, 2010). Furthermore, symptomatic differences have
been identified between younger and older adults with depression (Hybels et al.,
2012). Therefore, investigating fatigue and depression in older adults is clinically
significant; particularly considering the increased prevalence of fatigue in this age
group—although the age of participants increased the likelihood of medically
explainable fatigue within the cohort, thereby potentially reducing the specificity of
the study. However, fatigue and depression were assessed independently using
validated self-report questionnaires; allowing the utilisation of consistent assessment
measures throughout the complete study cohort, enabling comparable classifications
between individuals. Furthermore, utilising a current depression status was
advantageous because it enabled investigation of self-reported co-occurring fatigue
and depression. Finally, the use of a community study cohort removed potential
confounding with medical healthcare-seeking behaviour.
In summary, increased preponderance of depression and fatigue symptoms in
fatigued and depressed cases, respectively, indicates that an underlying association
exists between the two entities. Furthermore, the polythetic definition of depression
and the spectrum of fatigue symptoms imply that the underlying genetics of both
entities are heterogeneous. Therefore, utilisation of distinguishing symptoms could
facilitate the selection of more homogeneous subgroups, potentially enabling
identification of risk loci associated with varying phenotype presentations. Future
analyses should investigate the comorbidity of fatigue and depression by
70 Chapter 3: Co-occurrence and Symptomatology of Fatigue and Depression
characterising the type of relationship which exists between the two entities and their
underlying genetics.
71
Chapter 3 provided evidence that an underlying association likely exists between
fatigue and depression. Additionally, previous studies have indicated fatigue and
depression have genetic contributions. To determine if a shared genetic contribution
explains a proportion of the variation between fatigue and depression and
characterise the type of relationship that exists between the traits, the heritability of
the individual phenotypes was initially investigated. The following chapter aimed to
investigate the familiality and heritability of fatigue.
72 Chapter 4: Familiality and Heritability of Fatigue in an Australian Twin Sample
Chapter 4: Familiality and Heritability of
Fatigue in an Australian Twin
Sample
This chapter comprises the following prepared manuscript:
Corfield, E. C., Martin, N. G., & Nyholt, D. R. (In press, accepted 20 March 2017).
Familiality and heritability of fatigue in an Australian twin sample. Twin Research
and Human Genetics.
Chapter 4: Familiality and Heritability of Fatigue in an Australian Twin Sample 73
QUT Verified Signature
QUT Verified Signature
74 Chapter 4: Familiality and Heritability of Fatigue in an Australian Twin Sample
4.1 ABSTRACT
Familial factors have previously been implicated in the etiology of fatigue, of which,
a significant proportion is likely attributable to genetic influences. However, family
studies have primarily focused on Chronic Fatigue Syndrome, while univariate twin
studies have investigated broader fatigue phenotypes. The results for similar fatigue
phenotypes vary between studies, particularly, with regard to sex-specific
contributions to the heritability of the traits. Therefore, the current study aims to
investigate the familiality and sex-specific effects of fatigue experienced over the
past few weeks, in an older Australian population of 660 monozygotic (MZ) twin
pairs, 190 MZ singleton twins, 593 dizygotic (DZ) twin pairs, and 365 DZ singleton
twins. Higher risks for fatigue were observed in MZ compared to DZ co-twins of
probands with fatigue. Univariate heritability analyses indicated fatigue has a
significant genetic component, with a heritability (h2) estimate of 40%. Sex-specific
effects did not significantly contribute to the heritability of fatigue, with similar
estimates for males (h2 = 41%, 95% confidence interval [CI] = 18-62%) and females
(h2 = 40%, 95% CI = 27-52%). These results indicate that fatigue experienced over
the past few weeks has a familial contribution, with additive genetic factors playing
an important role in its etiology.
Chapter 4: Familiality and Heritability of Fatigue in an Australian Twin Sample 75
4.2 INTRODUCTION
Fatigue is a highly prevalent trait with multidimensional symptoms. The broad
symptom spectrum is associated with quantifiable difficulties, resulting in fatigue
classifications based on arbitrarily defined durations and severities. Commonly
utilised classifications include prolonged fatigue, chronic fatigue (CF), idiopathic
chronic fatigue (ICF), and chronic fatigue syndrome (CFS). Prolonged fatigue is
classified as self-reported persistent or relapsing fatigue experienced for at least one
month (Fukuda et al., 1994), and has an estimated population prevalence of 6.16-
28.00% (EvengÅRd et al., 2005; Hamaguchi et al., 2011; Jason et al., 1999; Kim et
al., 2005; Njoku et al., 2007). CF is classified as self-reported persistent or relapsing
fatigue experienced for at least six months (Fukuda et al., 1994), and has an
estimated population prevalence of 2.00-12.20% (Bierl et al., 2004; Cho et al., 2009;
EvengÅRd et al., 2005; Friedberg et al., 2015; Hamaguchi et al., 2011; Jason et al.,
1995; Jason et al., 1999; Kim et al., 2005; Loge et al., 1998; Njoku et al., 2007; Patel
et al., 2005; Steele et al., 1998; Wessely et al., 1995; Wessely et al., 1997; Wong &
Fielding, 2010). ICF is classified as clinically evaluated, medically unexplained CF,
with insufficient symptom presentation for diagnosis with CFS (Fukuda et al., 1994),
and has an estimated population prevalence of 1.00-9.00% (Hamaguchi et al., 2011;
Kim et al., 2005; Wessely et al., 1997).
The original CFS classification was published in 1988, by the Centres for
Disease Control (Holmes et al., 1988). This CFS classification required the presence
of new onset unexplained CF and either six or more symptom criteria (mild fever or
chills, sore throat, painful lymph nodes, muscle weakness, muscle discomfort or
myalgia, post-exertional fatigue, headaches, migratory arthralgia, neuropsychologic
complaints, sleep disturbance, and acute onset) and two physical criteria (low grade
fever, nonexudative pharyngitis, and palpable or tender lymph nodes), or at least
eight of the symptom criteria. In 1994, the Centres for Disease Control published a
revision to the CFS classification which has become the standard definition utilised
worldwide (Fukuda et al., 1994). The 1994, CFS classification requires clinically
evaluated, medically unexplained CF, with four or more physical symptoms (sore
throat, tender lymph nodes, headaches, cognitive difficulties, unrefreshing sleep,
multijoint pain, muscle pain, and post-exertional malaise) experienced over a six-
month period, which have not pre-dated the fatigue (Fukuda et al., 1994). The
76 Chapter 4: Familiality and Heritability of Fatigue in an Australian Twin Sample
population prevalence of CFS has been estimated at 0.07-2.60% (Cho et al., 2009;
Hamaguchi et al., 2011; Jason et al., 1999; Kawakami et al., 1998; Kim et al., 2005;
Lindal et al., 2002; Nacul et al., 2011; Njoku et al., 2007; Reyes et al., 2003; Vincent
et al., 2012; Wessely et al., 1995; Wessely et al., 1997).
Familial studies of fatigue have mainly focused on CFS. In 1991, Bell and
colleagues (1991) showed, children (aged 6-17) with CFS (based on the original CFS
classification) were significantly more likely to have family members with CFS
symptoms than asymptomatic controls (relative risks [RR] = 48.60, 95% confidence
interval [CI] = 9.43-587.22). Although the degree of relatedness investigated by the
authors is unclear. In 2001, Walsh and colleagues (2001) showed first-degree
relatives of CFS cases (with a mean age of 37.6 years) have an increased risk of
prolonged fatigue (RR = 2.18, 95% CI = 0.88-3.48) and CFS (RR = 9.22, 95% CI =
7.84-10.60). Additionally, Buchwald and colleagues (2001) showed monozygotic
(MZ) twin pairs have higher concordance rates compared to dizygotic (DZ) twin
pairs for CF and ICF (within a cohort with a mean age of 46 years). In 2006,
adolescents, aged 12-18, with CFS and their mothers were shown to have shared
symptom complexes, which were not exhibited by their fathers (van de Putte et al.,
2006). Finally, in 2011, Albright and colleagues (2011) showed CFS cases’ first (RR
= 2.70, 95% CI = 1.56-4.66), second (RR = 2.34, 95% CI = 1.32-4.19), and third (RR
= 1.93, 95% CI = 1.21-3.07) degree relatives had an increased risk of CFS compared
to controls. These family studies indicate genetic and common environmental factors
likely contribute to CFS.
Univariate twin studies have been utilised to estimate the contribution of
additive genetic (also known as narrow-sense heritability [h2]), common
environmental, and unique environmental factors to the variation observed in the
population of interfering fatigue (tiredness or fatigue experienced for at least five
days), abnormal tiredness, prolonged fatigue, CF, ICF, and CFS, in adults (see Table
4.1 for a summary) (Buchwald et al., 2001; Schur et al., 2007; Sullivan et al., 2003;
Sullivan et al., 2005). Interfering fatigue has an estimated heritability of 6% in males
and 26% in females (Sullivan et al., 2003). Similarly, abnormal tiredness has an
estimated heritability of 30% in males and 26% in females (Sullivan et al., 2005).
Prolonged fatigue has an estimated heritability of 34-51% in males and 18-27% in
females (Schur et al., 2007; Sullivan et al., 2005). CF has an estimated heritability of
Chapter 4: Familiality and Heritability of Fatigue in an Australian Twin Sample 77
30-47% in males and 12-32% in females (Buchwald et al., 2001; Schur et al., 2007;
Sullivan et al., 2005). Finally, ICF and CFS both have an estimated heritability’s of
51% in females (Buchwald et al., 2001; Schur et al., 2007). Notably, the heritability
estimates for males and females were similar within the Swedish cohort, which had
an age range of 42-64 years. Meanwhile, the American cohorts with mean ages of
32.4 years and approximately 35 years have greater differences in heritability
estimates between the sexes.
To date, only two studies have been conducted which included children or
adolescents and utilised univariate twin modelling to investigate the contribution of
genetic and environmental factors in fatigue phenotypes. The first study investigated
the heritability of short-duration fatigue (fatigue experienced for at least one week)
and prolonged fatigue within children (aged 5-17) from South Wales (Farmer et al.,
1999). However, sex-specific effects were not investigated and confidence intervals
were not reported for the heritability estimates. Nonetheless, short-duration fatigue
had an estimated additive genetic, common environmental, and unique
environmental contribution of 42%, 38%, and 20%, respectively. Similarly,
prolonged fatigue had an estimated additive genetic, common environmental, and
unique environmental contribution of 54%, 19%, and 26%, respectively. Meanwhile,
the heritability of fatigue severity (a continuous scale of the 11 core fatigue items and
2 muscle pain items of the Chalder Fatigue Questionnaire (Chalder et al., 1993)) and
abnormal fatigue (assessed by the 11 core fatigue items of the Chalder Fatigue
Questionnaire) was investigated in a Sri Lankan population of adolescents and adults
(aged ≥ 15) (Ball et al., 2010b). Fatigue severity had an estimated additive genetic,
and unique environmental contribution of 30% (95% CI = 24-35%) and 70% (95%
CI = 65-76%), respectively. Similarly, abnormal fatigue had an estimated additive
genetic, and unique environmental contribution of 39% (95% CI = 29-49%) and 61%
(95% CI = 51-71%), respectively.
78 Chapter 4: Familiality and Heritability of Fatigue in an Australian Twin Sample
Table 4.1. Previously published variance estimates (with their 95% confidence intervals) for varying fatigue classifications, in adults, from univariate structural equation
modelling.
Fatigue classification Study Population Mean age ± standard deviation
(Age range)
Number of
twin pairs
Male (M) Female (F)
A C E A C E
Interfering fatigue Sullivan et al. (2003) USA Case: 34.9 ± 9.3
Control: 35.1 ± 9.2
M = 3422
F = 3104
0.06
(0.00-0.46)
0.21
(0.00-0.25)
0.73
(0.54-0.90)
0.26
(0.00-0.44)
0.01
(0.00-0.30)
0.73
(0.56-0.92)
Abnormal tiredness Sullivan et al. (2005) Swedish (42-64) M =11293
F = 12813
0.30
(0.11-0.40)
0.00
(0.00-0.14)
0.70
(0.60-0.80)
0.26
(0.08-0.33)
0.00
(0.00-0.14)
0.74
(0.67-0.82)
Prolonged fatigue Sullivan et al. (2005) Swedish (42-64) M =11293
F = 12813
0.34
(0.03-0.45)
0.00
(0.00-0.25)
0.66
(0.55-0.79)
0.27
(0.06-0.35)
0.00
(0.00-0.16)
0.73
(0.65-0.82)
Schur et al. (2007) USA 32.4 ± 14.7
(18-90) M = 1468 F = 2272
0.51 (0.13-0.69)
0.00 (0.00-0.33)
0.49 (0.31-0.71)
0.18 (0.00-0.54)
0.23 (0.00-0.48)
0.59 (0.46-0.74)
Chronic fatigue Buchwald et al. (2001) USA 46 F = 146 - - - 0.19
(0.00-0.56)
0.69
(0.32-0.89)
0.12
(0.07-0.19)
Sullivan et al. (2005) Swedish (42-64) M =11293
F = 12813
0.30
(0.02-0.44)
0.00
(0.00-0.23)
0.70
(0.56-0.86)
0.32
(0.11-0.41)
0.00
(0.00-0.16)
0.68
(0.59-0.78)
Schur et al. (2007) USA 32.4 ± 14.7
(18-90) M = 1468 F = 2272
0.47 (0.00-0.68)
0.00 (0.00-0.39)
0.53 (0.32-0.79)
0.12 (0.00-0.48)
0.26 (0.00-0.48)
0.62 (0.47-0.78)
Idiopathic chronic fatigue Buchwald et al. (2001) USA 46 F = 146 - - - 0.51
(0.07-0.96)
0.42
(0.00-0.85)
0.08
(0.04-0.13)
Chronic fatigue syndrome Schur et al. (2007) USA 32.4 ± 14.7
(18-90)
M = 1444
F = 2222 - - -
0.51
(0.00-0.82)
0.12
(0.00-0.72)
0.36
(0.18-0.65)
A: additive genetic component; C: common environmental component E: unique environmental component.
Chapter 4: Familiality and Heritability of Fatigue in an Australian Twin Sample 79
Additional studies have utilised multivariate twin modelling to investigate the
contribution of shared genetic and environmental factors to numerous traits that are
comorbid or hypothesised to be associated with fatigue. The traits investigated within
previous multivariate studies include various fatigue definitions (i.e., fatigue
symptoms, short-duration fatigue, abnormal fatigue, fatigue, prolonged fatigue, and
CF) and major depressive disorder, insomnia, psychological distress, anxiety,
depression, psychological symptoms, somatic symptoms, generalised anxiety
disorder, disability pension due to neurotic diagnoses, headaches, irritable bowel
syndrome, chronic widespread pain, immune responsiveness, and the immunological
factors IL-4, IFN-γ, and sCD23. The heritability of the various fatigue measures
ranged from 7% to 60%, and a number of these studies reported significant evidence
for shared genetic factors between fatigue and other traits, in particular strong genetic
correlations (rg) were observed between prolonged fatigue and depression (rg = 0.53),
CF and depression or anxiety (rg = 0.60), fatigue and psychological distress (rg =
0.67), and fatigue and immune responsiveness (rg = 0.76) (Ball et al., 2011; Fowler et
al., 2006; Hickie et al., 1999a; Hickie et al., 1999b; Hickie et al., 2001; Hur et al.,
2012; Kato et al., 2009; Narusyte et al., 2016).
Given the large variation in both the definition of fatigue and estimates of
heritability produced from a relatively small number of univariate twin studies
(conducted in Swedish and American cohorts), the current study aimed to investigate
the heritability of fatigue experienced over the past few weeks, in a cohort of
Australian twin pairs. While previously published family studies have focused on
CFS, we assessed the familiality of fatigue experienced over a shorter time period.
4.3 METHODS
4.3.1 Study Cohort and Fatigue Classification
The present study utilises data from the over 50’s (aged) study, conducted by the
genetic epidemiology group within the QIMR Berghofer Medical Research Institute
(QIMRB), between 1993 and 1996. The study invited 2,281 twin pairs, aged over 50,
from the Australian Twin Registry to complete a 16-page mailed Health and
Lifestyle Questionnaire (Bucholz et al., 1998; Mosing et al., 2012). Informed written
consent was obtained from each participant, and the study was approved by the
Human Research Ethics Committee (HREC) of QIMRB.
80 Chapter 4: Familiality and Heritability of Fatigue in an Australian Twin Sample
The fatigue classification utilised throughout this study was assessed by the
Schedule of Fatigue and Anergia (SOFA) (Hickie et al., 1996). Ten questions are
contained in the SOFA; however, a shorter eight-item version was used in the Health
and Lifestyle Questionnaire, due to two items being replicated within the General
Health Questionnaire (GHQ) (Goldberg & Blackwell, 1970), that was also
administered to the participants. Responses to the eight SOFA and two GHQ items
were used to assess fatigue within the cohort, as previously detailed (Corfield et al.,
2016a). Individuals were classified as fatigued if they reported three or more of the
ten fatigue symptoms (muscle pain at rest, post-exertional muscle pain, post-
exertional muscle fatigue, post-exertional fatigue, hypersomnia, insomnia, poor
concentration, speech problems, poor memory, and headaches), over the past few
weeks. Fatigue was dichotomized rather than using symptom counts because the
SOFA was originally designed to identify chronic fatigue syndrome cases and
assesses physical, neurocognitive, and neurovegetative fatigue symptoms. Hopefully
the utilisation of case-control classifications enabled us to identify fatigued
individuals with similar underlying pathophysiology as CFS cases and prevented
confounding with other traits.
4.3.2 Statistical Analysis
Familial clustering of fatigue was investigated by calculating RR, measured by the
prevalence ratio, with their 95% CI in complete MZ and DZ twin pairs. RR were
calculated relative to non-fatigued individuals. Within MZ and same-sex DZ twin
pairs RR were calculated by averaging over using twin 1 or twin 2 as the proband.
Tetrachoric correlations were calculated for fatigue, within MZ and DZ twin
pairs and singleton twins, using the polycor package in R (R Core Team, 2014). The
tetrachoric correlation assumes that underlying the observed binary distribution of
affection status, there exists a continuous, normally distributed latent (non-
observable) liability (Kendler, 1993). That is, the tetrachoric correlation is an
estimate of the correlation between two latent variables, where each latent variable is
assumed to have a bivariate normal distribution. Comparison of the correlations
between MZ and DZ twins was used to provide information on the importance of
genetic and environmental factors contributing to the heritability of fatigue.
Correlations that are larger in MZ compared to DZ twins indicate the phenotype has
Chapter 4: Familiality and Heritability of Fatigue in an Australian Twin Sample 81
a genetic contribution. While correlations that are similar in MZ and DZ twins
indicate environmental factors explain the majority of variation in the phenotype.
Structural equation modelling (SEM), including the threshold model, was
utilised to investigate the heritability of fatigue. The threshold model posits that
distinct traits represent a single, normally distributed, severity continuum. A single
threshold was used to separate non-fatigued and fatigued individuals. SEM was used
to estimate the contribution of additive genetic (A), non-additive (dominance)
genetic (D), common environmental (C), and unique environmental (E) variance
components (Neale et al., 1992). Adjustments for (linear) age and sex effects were
included in the model. Significance of the variance components was assessed by
comparing the fit of the full model (ACE/ADE) to the nested models (AE, CE, and
E) where the effect was dropped, using OpenMx in R (Boker et al., 2011).
Additionally, sex-limitation modelling was conducted to determine if sex-specific
effects contribute to the heritability of fatigue. Initially, a non-scalar sex-limitation
model was fitted which included variance components for females (i.e., Af, Cf, and
Ef) and males (i.e., Am, Cm, and Em), as well as an additional additive genetic
component specific to males (A′m). Restricted non-scalar sex-limitation modelling
was then conducted, whereby, A′m was removed. Evidence for sex-specific genetic
effects was formally tested by determining if the genetic correlation within opposite-
sex DZ twin pairs significantly differs from 0.5. The goodness of fit parameters used
to assess the differences in the twin models were the likelihood-ratio chi-square test
(χ2) and the p-value. Additionally, model fit was compared utilising Akaike’s
Information Criteria (AIC); with the lowest AIC indicating the most parsimonious
model (Akaike, 1973, 1974).
Tetrachoric correlations and SEM were estimated using full information
maximum likelihood (FIML), whereby both complete twin pairs and incomplete twin
pairs (singleton twins) were included in the analyses. The inclusion of singleton
twins provides more accurate estimation of the thresholds and may correct for
participation bias.
4.4 RESULTS
Within the over 50’s study, 473 twin pairs and 555 singleton twins returned
incomplete responses to the SOFA and GHQ questionnaire items utilised to assess
82 Chapter 4: Familiality and Heritability of Fatigue in an Australian Twin Sample
fatigue within the present study and were therefore excluded. The remaining 1,253
complete twin pairs and 555 singleton twins with fatigue data comprised the cohort
utilised within the present study. The study cohort contained 660 MZ twin pairs (504
female-female and 156 male-male twin pairs) and 190 MZ singleton twins (109
females and 81 males) with a mean age of 61.3 ± 8.9 (range = 50-92), and 593 DZ
twin pairs (272 female-female, 76 male-male, 137 female-male, and 108 male-female
twin pairs) and 365 DZ singleton twins (260 females and 105 males) with a mean age
of 61.0 ± 8.5 (range = 50-94). The prevalence of fatigue defined as above was 30.7%
(31.7% of females and 28.3% of males).
An increased risk of fatigue in co-twins of fatigued probands was observed,
indicating a significant familial contribution. Strong evidence for a genetic
contribution to fatigue is provided by the higher RR observed in MZ compared to DZ
twin pairs (Table 4.2). In particular, the risk of fatigue in co-twins of fatigued
probands was 2.20 (95% CI = 1.77-2.75) in MZ twin pairs compared to 1.32 (95% CI
= 1.01-1.73) in DZ twin pairs (applicable to first-degree relatives in the general
population). Analysis of familial clustering within males and females indicated a
similar pattern of risks.
Table 4.2. Relative riska of fatigue within complete monozygotic (MZ), same-sex dizygotic (DZss),
and opposite-sex dizygotic (DZos) twin pairs.
Zygosity Number of complete twin pairs RR (95% CI)
MZ 660 2.20 (1.77-2.75)
MZ [F-F] 504 2.14 (1.68-2.74)
MZ [M-M] 156 2.28 (1.38-3.78) DZ total 593 1.32 (1.01-1.73)
DZss 348 1.16 (0.83-1.62)
DZss [F-F] 272 1.14 (0.78-1.66) DZss [M-M] 76 1.23 (0.61-2.49)
DZos 241 1.59 (1.03-2.45)
F: female; M: male; aRelative risks and 95% confidence intervals were calculated with respect to non-depressed or non-fatigued status in twin 1.
Same-sex twin pair tables were made symmetrical by averaging over
using twin 1 or twin 2 as the proband.
The tetrachoric correlations for fatigue were approximately three times larger
in MZ compared to DZ twin pairs (Table 4.3). Overall, the observed MZ > DZ
correlations, indicate additive genetic factors contribute to the variation in fatigue.
Chapter 4: Familiality and Heritability of Fatigue in an Australian Twin Sample 83
Table 4.3. Tetrachoric correlations (r) with their 95% confidence intervals (CI) for fatigue according
to zygosity.
Zygosity r (95% CI)
MZ 0.43 (0.32-0.54)
MZ [F-F] 0.43 (0.30-0.56) MZ [M-M] 0.41 (0.17-0.66)
DZ total 0.14 (0.01-0.28)
DZss 0.08 (-0.10-0.26) DZss [F-F] 0.07 (-0.14-0.27)
DZss [M-M] 0.12 (-0.26-0.51)
DZos 0.24 (0.02-0.46)
F: female; M: male.
Initially, full univariate ACE and ADE models were fitted, however,
systematic dropping of A, C, and D effects was used to determine if the effect of the
individual variance components was significant (Table 4.4). Dropping C (i.e., AE
model), from the ACE model, did not worsen the model fit. However, dropping A
(i.e., CE model) or both A and C (i.e., E model) was significant (p = 4.76 × 10-3 and
8.91 × 10-11, respectively)—indicating A is an important source of variance in the
heritability of fatigue. Meanwhile, dropping D (i.e., AE model), from the ADE
model, did not worsen the model fit. However, dropping both A and D (i.e., E model)
was significant (p = 6.46 × 10-11)—indicating genetic factors play an essential role in
the heritability of fatigue. Therefore, the AE model was selected as the most
parsimonious model based on fit statistics. No differences in threshold distributions
were observed within twin pairs and singleton twins, or across zygosity and sex
groups.
Additive genetic factors were estimated to explain approximately 40% of the
heritability of fatigue. No significant evidence for sex-specific genetic effects was
observed within the cohort. Results of the non-scalar sex-limitation modelling
indicated the restricted model was the most parsimonious (AIC = -2405.41) with
similar heritability estimates for fatigue in males, at 41% (95% CI = 18-62%; E =
59%, 95% CI = 38-82%), compared to females, at 40% (95% CI = 27-52%; E =
60%, 95% CI = 48-73%).
84 Chapter 4: Familiality and Heritability of Fatigue in an Australian Twin Sample
Table 4.4. Fit statistics and variance estimates (with their 95% confidence intervals) from univariate
structural equation modelling.
Model -2LL p-value (ACE) p-value (ADE) AIC A C (or D) E
ACE 3704.60 NA NA -2407.40 0.40
(0.15-0.50)
0.00
(0.00-0.20)
0.60
(0.50-0.71)
ADE 3703.96 NA NA -2408.05 0.17
(0.00-0.50)
0.25
(0.00-0.53)
0.58
(0.47-0.70)
AE 3704.60 1.00 0.42 -2409.40 0.40
(0.29-0.50) -
0.60
(0.50-0.71)
CE 3712.57 4.76 × 10-3 NA -2401.43 - 0.30
(0.21-0.39)
0.71
(0.61-0.79)
E 3750.88 8.91 × 10-11 6.46 × 10-11 -2365.12 - - 1.00
(1.00-1.00)
-2LL: minus two log-likelihood; A: additive genetic component; C: common environmental component E: unique environmental component. The best-fitting model is indicated in bold. p-value compares -2LL for the full ACE or ADE model
to the reduced (AE, CE, DE, and E) models.
4.5 DISCUSSION
Findings from the present study indicate fatigue, in older adults, is familial, and has a
genetic contribution with no significant sex-specific effects.
The familial clustering analysis revealed co-twins of fatigued probands were at
an increased risk of fatigue. These results indicate the familial contribution of fatigue
is not specific to CFS. Although, in 2001, first-degree relatives of CFS cases were
shown to have an increased risk of prolonged fatigue and MZ twin pairs were shown
to have higher concordance rates than DZ twin pairs for CF and ICF (Buchwald et
al., 2001; Walsh et al., 2001). However, to our knowledge, this is the first study to
characterise the familial clustering of fatigue experienced for less than six months.
The higher risk observed in MZ twin pairs compared to DZ twin pairs indicates
genetic factors likely contribute to the etiology of fatigue. These results are reflective
of the conclusions drawn from previous family studies of CF, ICF, and CFS
(Buchwald et al., 2001; van de Putte et al., 2006). However, the similar pattern of
risks observed within males and females indicates the underlying etiology of fatigue
is likely independent of sex. This finding opposes the results of van de Putte and
colleagues (2006), who found an increase of CFS symptoms in mothers of children
with CFS, but not fathers. Indicating the etiology of fatigue may differ with age.
The differences in heritability between males and females identified within
previous twin studies are larger in the cohorts comprised of younger adults compared
to cohorts of older cohorts. In contrast, results from the present study indicate that
the underlying etiology of fatigue is independent of sex in older adults. Sex-
limitation modelling revealed males (h2 = 41%, 95% CI = 18-62%) and females (h2 =
40% (95% CI = 27-52%) had very similar heritability estimates. Furthermore, based
Chapter 4: Familiality and Heritability of Fatigue in an Australian Twin Sample 85
on fit statistics the most parsimonious model was the univariate AE twin model—
which did not include sex-specific effects. These results support the suggestion of
Sullivan and colleagues (2005), that females and males have similar genetic and
environmental contributions for varying fatigue classifications—despite the higher
prevalence of fatigue in females. In comparison, Schur and colleagues (2007)
suggested further investigations are required to understand the differences in fatigue
etiology between the sexes. Within the present study higher tetrachoric correlations
were observed in opposite-sex DZ twin pairs compared to same-sex twin pairs. The
same pattern of correlations has been reported for other complex phenotypes (Vink et
al., 2012). One possible explanation is that males and females respond differently on
self-report questionnaires (Sigmon et al., 2005). Based on our results and previous
findings we suggest further investigation into the heritability of fatigue across the
lifespan is required.
A possible limitation of our study is the utilisation of self-report rather than
interview based data. However, considering prolonged fatigue and CF classifications
are based on self-report the utilisation of questionnaire-based data is valid.
Additionally, this prevented confounding within the study by healthcare seeking
behaviour, due to the population-based structure of the cohort. Another potential
limitation of the study was the utilisation of a non-standard fatigue duration—due to
the ambiguity of the questionnaire, which assessed fatigue symptoms experienced
“over the past few weeks”. However, considering the SOFA was designed to assess
CFS symptoms, fatigue is representative of a spectrum, and previous studies have
looked at similar fatigue definitions—our findings still offer valid insights into the
underlying etiology of fatigue.
In summary, we have shown that fatigue experienced over the past few weeks
is familial, with additive genetic factors explaining a substantial proportion of its
variance in older adults. Future research aimed at identifying the specific genes and
risk loci associated with fatigue (e.g., via genome-wide association studies), will
increase our understanding of its underlying biological mechanisms.
86
In Chapter 4, fatigue experienced over the past few weeks, in older adults, was
shown to have a familial component, with additive genetic factors explaining 40% of
the traits’ variation. A similar analysis was conducted within Chapter 5 to determine
if the familiality and heritability of MDD within the study cohort was similar to
previously published analyses. Additionally, Chapter 5 aimed to investigate the
familiality and heritability of MiDD and determine the validity of utilising a broad
depression phenotype comprising MDD and MiDD cases.
Chapter 5: A Continuum of Genetic Liability for Minor and Major Depression 87
Chapter 5: A Continuum of Genetic
Liability for Minor and Major
Depression
This chapter comprises the following submitted article:
Corfield, E. C., Yang, Y., Martin, N. G., & Nyholt, D. R. (In Press, accepted 4 April
2017). A continuum of genetic liability for minor and major depression.
Translational Psychiatry.
88 Chapter 5: A Continuum of Genetic Liability for Minor and Major Depression
QUT Verified Signature
QUT Verified Signature
Chapter 5: A Continuum of Genetic Liability for Minor and Major Depression 89
5.1 ABSTRACT
The recent success of a large genome-wide association (GWA) study—analysing
130,620 major depression cases and 347,620 controls—in identifying the first single
nucleotide polymorphism (SNP) loci robustly associated with major depression in
Europeans, confirms that immense sample sizes are required to identify risk loci for
depression. Given the phenotypic similarity between major depressive disorder
(MDD) and the less severe minor depressive disorder (MiDD), we hypothesised that
broadening the case definition to include MiDD may be an efficient approach to
increase sample sizes in GWA studies of depression. By analysing two large twin
pair cohorts, we show that minor depression and major depression lie on a single
genetic continuum, with major depression more severe but not etiologically distinct
from minor depression. Furthermore, we estimate heritabilities of 37% for minor
depression, 46% for major depression, and 48% for ‘minor or major depression’ in a
cohort of older adults (aged 50-92). While, the heritability of ‘minor or major
depression’ was estimated at 40% in a cohort of younger adults (aged 23-38).
Moreover, two robust major depression risk SNPs nominally associated with major
depression in our Australian GWA dataset, produced more significant evidence for
association with ‘minor or major depression’. Hence, broadening the case phenotype
in GWA studies to include sub-threshold definitions such as MiDD, should facilitate
the identification of additional genetic risk loci for depression.
90 Chapter 5: A Continuum of Genetic Liability for Minor and Major Depression
5.2 INTRODUCTION
Major depressive disorder (MDD) is a common, complex trait with an estimated
heritability of approximately 40% (Sullivan et al., 2000). However, until recently,
genome-wide association (GWA) studies in large European samples have failed to
robustly identify genetic variants contributing to MDD (Bosker et al., 2011; Kohli et
al., 2011; Lewis et al., 2010; Major Depressive Disorder Working Group of the
Psychiatric GWAS Consortium et al., 2013; Muglia et al., 2010; Rietschel et al.,
2010; Shi et al., 2011; Shyn et al., 2011; Sullivan et al., 2009; Wray et al., 2012). In
July of 2015, two genome-wide significant (p < 5 × 10-8) single nucleotide
polymorphism (SNP) loci (rs12415800 near the SIRT1 gene, p = 2.53 × 10-10 and
rs35936514 in the intron of LHPP, p = 6.45 × 10-12) were reported to be associated
with severe and recurrent MDD, in a sample of Han Chinese women (5,282 cases,
5,220 controls) (CONVERGE Consortium, 2015); however, these SNPs were not
associated in the Psychiatric Genomics Consortium (PGC) GWA study of 9,240
European MDD cases and 9,519 controls (Major Depressive Disorder Working
Group of the Psychiatric GWAS Consortium et al., 2013).
In August 2016, the first SNP loci robustly associated with major depression in
Europeans were reported (Hyde et al., 2016). This landmark study analysed a
combined cohort of 130,620 self-reported and clinically evaluated lifetime major
depression cases and 347,620 controls, and identified 17 genome-wide significant
SNPs within 15 independent genomic regions. The implicated SNP risk loci
comprise rs10514299 in an intron of TMEM161B-AS1 (p = 9.99 × 10-16), rs1518395
in an intron of VRK2 (p = 4.32 × 10-12), rs2179744 in an intron of L3MBTL2 (p =
6.03 × 10-11), rs11209948 downstream of NEGR1 (p = 8.38 × 10-11), rs454214
upstream of MEF2C (p = 1.09 × 10-9), rs301806 in an intron of RERE (p = 1.90 × 10-
9), rs1475120 in an intron of LIN28B (p = 4.17 × 10-9), rs10786831 in an intron of
SORCS3 (p = 8.11 × 10-9), rs12552 in the 3′ UTR of OLFM4 (p = 8.16 × 10-9),
rs6476606 in an intron of PAX5 (p = 1.20 × 10-8), rs8025231 in an intergenic region
between MEIS2 and TMCO5A (p = 1.23 × 10-8), rs12065553 in an intergenic region
on chromosome 1 (p = 1.32 × 10-8), rs1656369 in the intergenic region between
RSRC1 and MLF1 (p = 1.34 × 10-8), rs4543289 in an intergenic region on
chromosome 5 (p = 1.36 × 10-8), rs2125716 upstream of SLC6A15 (p = 3.05 × 10-8),
rs2422321 downstream of NEGR1 (p = 3.18 × 10-8), and rs7044150 in the intergenic
Chapter 5: A Continuum of Genetic Liability for Minor and Major Depression 91
region between KIAA0020 and RFX3 (p = 4.31 × 10-8). An important implication of
this study is that immense sample sizes are required to identify a relatively modest
number of MDD risk loci in Europeans (compared to other traits of similar
prevalence and heritability) (Visscher et al., 2012).
Additional insights into the molecular mechanisms of depression, in
Europeans, were identified in 2016 through the investigation of depressive symptoms
and a broad depression phenotype. In April, rs7973260 in an intron of KSR2 (p = 1.8
× 10-9) and rs62100776 in an intron of DCC (p = 8.5 × 10-9) were associated with
depressive symptoms (Okbay et al., 2016). Meanwhile, in December, rs9825823
located in the intron of FHIT (p = 8.2 × 10-9) was associated with a broad depression
phenotype—including MDD and depressive symptoms (Direk et al., 2016). Most
recently, MDD in adults aged over 27 years was associated with the intergenic SNP
rs7647854, located on chromosome 3 (p = 5.2 × 10-11) (Power et al., 2017).
Given the phenotypic similarity between MDD and the less severe minor
depressive disorder (MiDD), we hypothesised that broadening the case definition to
include sub-threshold definitions such as MiDD may provide an efficient means to
increase sample sizes in GWA studies of depression.
In contrast to MDD, the heritability and molecular genetics of MiDD have not
been well investigated. The only difference in diagnosis between MDD and MiDD is
the number of presenting symptoms of the Diagnostic and Statistical Manual of
Mental Disorders (DSM-IV) criteria (NB, within the DSM-V, MiDD individuals
would be diagnosed as ‘Unspecified Depressive Disorder’); with MDD requiring at
least five symptoms and MiDD requiring two to four symptoms (American
Psychiatric Association, 2000, 2013). This phenotypic similarity coupled with a
reported increased risk of MDD in first degree relatives and patients with MiDD
(Chen et al., 2000; Cuijpers et al., 2004; Cuijpers & Smit, 2004; Judd et al., 1997;
Lewinsohn et al., 2003; Rapaport et al., 2002) suggests that a depression continuum
exists and that MiDD may be a relevant trait which could be utilised to elucidate the
underlying mechanisms associated with MDD. Therefore, the present study utilises
relative risks (RR) to ensure a similar pattern of familiality exists within the study
cohorts. In addition, heritability estimates and the liability threshold model were
utilised to investigate whether minor and major depression lie on a single genetic
continuum. Finally, the association signal of the 17 SNPs robustly associated with
92 Chapter 5: A Continuum of Genetic Liability for Minor and Major Depression
major depression in Europeans was examined utilising both a narrow major
depression case phenotype and a broader depression phenotype including minor
depression (i.e., ‘minor or major depression’ cases).
5.3 MATERIALS AND METHODS
5.3.1 Study Cohorts
Two independent, community-based cohorts of Australian twin pairs were analysed
within the current study. Initially, the analysis was conducted within an older adult
cohort, the over 50’s (aged) study, before being replicated in a young adult cohort,
the Twin 89 (TE) study. Informed written consent was obtained from each
participant, and the study was approved by the Human Research Ethics Committee
(HREC) of the QIMR Berghofer Medical Research Institute (QIMRB).
The over 50’s cohort (Bucholz et al., 1998; Mosing et al., 2012) contained
1,220 twin pairs with complete self-report depression classifications (non-depressed,
minor depression, major depression). Current depression classifications were
obtained utilising a combination of responses from the twelve-item General Health
Questionnaire (GHQ) (Goldberg & Blackwell, 1970) and the fourteen-item
Delusions-Symptoms-States Inventory, States of Anxiety and Depression
(DSSI/sAD) (Bedford & Deary, 1997) questionnaires. As previously detailed
(Corfield et al., 2016a), specific questions from the GHQ and DSSI/sAD were
assigned to the appropriate DSM-IV major depressive episode criteria (American
Psychiatric Association, 2000). If an individual exhibited at least five of the DSM-IV
symptom criteria, of which either depressed mood or anhedonia was reported, they
were assessed as suffering major depression. Similarly, if an individual exhibited two
to four of the DSM-IV symptom criteria (depressed mood, anhedonia, a change in
weight or appetite, insomnia or hypersomnia, psychomotor agitation or retardation,
fatigue or loss of energy, feelings of worthlessness or excessive guilt, inability to
concentrate or make decisions, and thoughts about death, suicidal thoughts, suicidal
plans, or suicidal attempts), of which either depressed mood or anhedonia was
reported, they were assessed as suffering minor depression. The remaining
individuals were assessed as non-depressed. Meanwhile, the Twin 89 cohort (Heath
et al., 2001) contained 2,363 twin pairs, with complete lifetime self-report depression
classifications. Minor depression and major depression classifications required an
Chapter 5: A Continuum of Genetic Liability for Minor and Major Depression 93
individual to report depressed mood and/or anhedonia. In addition, individuals
reporting a total of two to four and five or more depression symptoms for a period of
two or more weeks, across their lifespan, were classified as minor depression and
major depression, respectively. In depth explanation of depression assessment
utilised within the Twin 89 cohort is provided by Yang and colleagues (Yang et al.,
2016). Depression can be conceptualised as the extreme of a trait that is widely
distributed in the general population. Utilisation of depression symptom count is one
way to investigate the genetic architecture of the complete distribution. However, the
DSM diagnosis for depression requires the presence of depressed mood or
anhedonia, which would not be accounted for if symptom-counts were used.
Therefore, within the present study a three-category depression classification was
used to focus on traits which were as close as possible to clinical diagnoses.
5.3.2 Statistical Analysis
Familial clustering of major depression, minor depression, and depression (minor or
major depression) was investigated by calculating RR with their 95% confidence
intervals (CI) in monozygotic (MZ) and dizygotic (DZ) twin pairs. RR were
calculated relative to non-depressed individuals. Within MZ and same-sex DZ twin
pairs RR were calculated by averaging over using twin 1 or twin 2 as the proband
(Nyholt et al., 2004).
A major goal of the genetic analysis was to test the multiple threshold model,
which asserts that different syndromes reflect different levels of severity on a single
dimension, rather than distinct etiologies (Reich et al., 1972). The fit of the multiple
threshold model was tested by calculating the polychoric correlation for the three-
category depression (non-depressed, minor depression, major depression)
classification using POLYCORR (Uebersax J.S., 2007). The polychoric correlation,
assumes that underlying the observed polychotomous distribution of affection status,
there exists a continuous, normally distributed latent (non-observable) liability
(Kendler, 1993). That is, the polychoric correlation is an estimate of the correlation
between two latent variables, where each latent variable is assumed to have a
bivariate normal distribution. A 2 goodness-of-fit test is used to test whether the
multiple threshold model provides a good fit to the observed data (i.e., compares the
observed frequencies to those predicted by the model). Additionally, polychoric
correlations were calculated for minor depression (excluding major depression
94 Chapter 5: A Continuum of Genetic Liability for Minor and Major Depression
cases), major depression (excluding minor depression cases), two-category
depression (non-depressed, minor or major depression), and three-category
depression within MZ and DZ twin pairs. Comparisons of the correlations between
MZ and DZ twin pairs was used to provide information on the importance of genetic
and environmental factors contribution to the heritability of depression.
Structural equation modelling (SEM) was utilised to investigate the heritability
of minor depression (excluding major depression cases), major depression (excluding
minor depression cases), two-category depression, and three-category depression.
SEM was used to estimate the contribution of additive genetic (A), non-additive
(dominance) genetic (D), common environmental (C), and unique environmental (E)
variance components (Neale et al., 1992). Adjustments for (linear) age and sex
effects were included in the model. Significance of the variance components was
assessed by comparing the fit of the full model (ACE/ADE) to the nested models
(AE, CE, and E) where the effect was dropped, using OpenMx in R (Boker et al.,
2011). Evidence for sex-specific genetic effects was formally tested by determining
if the genetic correlation within opposite-sex DZ twin pairs significantly differs from
0.5. The goodness of fit parameters used to assess the differences in the twin models
were the likelihood-ratio chi-square test (χ2) and the p-value. Additionally, model fit
was compared utilising Akaike’s Information Criteria (AIC); with the lowest AIC
indicating the most parsimonious model (Akaike, 1973, 1974).
An association analysis was conducted for the 17 genome-wide significant loci
associated with major depression, in Europeans (Hyde et al., 2016), within the
Australian GWA dataset. The Australian GWA dataset is a community cohort which
contained 3,664 unrelated major depression cases, 620 unrelated minor depression
cases, and 7,113 unrelated controls, of European ancestry. In depth explanation of
the genotyping and quality control methods utilised within the GWA cohort have
previously been detailed by Medland and colleagues (Medland et al., 2009). Briefly,
standard quality control measures were utilised, whereby SNPs with BeadStudio
GenCall scores < 0.7, call rate < 0.95, Hardy-Weinberg equilibrium p-values < 1 ×
10-6, and minor allele frequencies < 0.01 were excluded. Imputation was then
conducted utilising HapMap samples of European ancestry. If multiple cases were
present within a family the most severe case was selected based on the number of
reported DSM depression symptoms, or an individual was randomly selected if
Chapter 5: A Continuum of Genetic Liability for Minor and Major Depression 95
numerous individuals reported the same number of depression symptoms. Similarly,
if multiple controls were available within a family the single ‘best’ control was
selected based on the lowest number of depression symptoms reported or an
individual was randomly selected if numerous individuals within a family reported
the same number of depression symptoms. Finally, a single population control was
randomly selected from the remaining families for whom genotyping data but no
depression phenotype data was available.
The association analysis was conducted on 3,664 major depression cases and
7,113 controls (2,381 unaffected + 4,732 population-based controls), using logistic
regression with sex as a covariate, using PLINK (Purcell et al., 2007). The
association analysis was then repeated using a broad depression phenotype, of 4,284
cases (3,664 major depression + 620 minor depression cases). The results were then
compared to ascertain if the evidence for association was increased by the addition of
minor depression cases; thus reflecting an increase in power.
5.4 RESULTS
The over 50’s (aged) study cohort consisted of 643 MZ twin pairs (491 female-
female [F-F] and 152 male-male [M-M]) and 577 DZ twin pairs (263 F-F, 73 M-M,
136 female-male [F-M], and 105 male-female [M-F]), with a mean age of 61.30 ±
8.60 (range = 50-92). The prevalence of minor depression, major depression, and
two-category depression was 8.98%, 2.05%, and 11.02%, respectively (9.61%,
2.12%, 11.72% in females; 7.38%, 1.88%, 9.26% in males). Meanwhile, the Twin 89
(TE) cohort consisted of 1,005 MZ twin pairs (609 F-F and 396 M-M) and 1,358 DZ
twin pairs (455 F-F, 349 M-M, 301 F-M, and 253 M-F) with a mean age of 29.80 ±
2.49 (range = 23-38). The prevalence of minor depression, major depression, and two
category depression was 7.70%, 37.41%, and 45.11%, respectively (7.23%, 42.80%,
50.04% in females; 8.32%, 30.33%, 38.65% in males).
Familial clustering of minor depression and major depression cases was
observed (Table 5.1), with co-twins of minor depression probands having an
increased risk of major depression, and vice versa, within both cohorts.
96 Chapter 5: A Continuum of Genetic Liability for Minor and Major Depression
Table 5.1. Relative riska of depression and fatigue within monozygotic (MZ), same-sex dizygotic (DZss), and opposite-sex dizygotic (DZos) twin pairs.
Aged TE
Probandco-twin MZ (643 pairs) DZss (336 pairs) DZos F-M (241 pairs) DZos M-F (241 pairs) MZ (1,005 pairs) DZss (804 pairs) DZos F-M (554 pairs) DZos M-F (554 pairs)
MiDMiD 2.85 (1.62-5.00) 2.67 (1.09-6.51) 1.13 (0.27-4.66) 1.15 (0.29-4.51) 0.91 (0.37-2.23) 1.71 (0.79-3.72) 1.40 (0.49-3.98) 1.05 (0.38-2.85)
MiDMD 7.32 (2.47-21.67) 2.04 (0.25-16.54) 4.24 (0.82-21.99) 4.02 (0.44-36.69) 1.64 (1.22-2.22) 1.17 (0.82-1.68) 1.06 (0.66-1.70) 1.40 (1.04-1.88)
MiDMiD/MD 3.48 (2.19-5.54) 2.54 (1.15-5.61) 1.79 (0.66-4.83) 1.51 (0.50-4.53) 1.48 (1.14-1.92) 1.27 (0.95-1.69) 1.12 (0.75-1.66) 1.34 (1.04-1.71)
MDMiD 5.45 (2.69-11.03) 2.22 (0.35-14.09) 3.53 (0.60-20.66) 3.44 (1.03-11.46) 1.17 (0.75-1.83) 1.06 (0.63-1.80) 1.72 (0.95-3.11) 0.99 (0.53-1.82)
MDMD 6.02 (0.79-45.56) - - - 2.16 (1.82-2.56) 1.46 (1.22-1.76) 1.16 (0.89-1.50) 1.17 (0.95-1.45)
MDMiD/MD 5.53 (2.99-10.22) 1.76 (0.28-11.05) 2.79 (0.48-16.07) 3.01 (0.91-9.94) 1.94 (1.68-2.24) 1.39 (1.19-1.63) 1.26 (1.01-1.56) 1.14 (0.95-1.36)
MiD/MDMiD 3.32 (2.04-5.42) 2.58 (1.12-5.94) 1.46 (0.45-4.75) 1.72 (0.64-4.60) 1.12 (0.73-1.73) 1.17 (0.72-1.91) 1.67 (0.94-2.97) 1.00 (0.57-1.76)
MiD/MDMD 7.09 (2.50-20.05) 1.64 (0.20-13.42) 3.66 (0.70-19.08) 3.01 (0.33-27.85) 2.07 (1.75-2.45) 1.41 (1.18-1.69) 1.14 (0.89-1.46) 1.22 (1.01-1.48)
MiD/MDMiD/MD 3.85 (2.54-5.84) 2.39 (1.12-5.06) 1.92 (0.78-4.76) 1.88 (0.79-4.48) 1.86 (1.62-2.14) 1.37 (1.18-1.60) 1.23 (1.00-1.52) 1.18 (1.00-1.39) aRelative risks and 95% confidence intervals were calculated with respect to non-depressed or non-fatigued status in twin 1. Same-sex twin pair tables were made symmetrical by averaging over using twin 1 or twin 2 as the proband. MiD: minor depression; MD: major depression.
Chapter 5: A Continuum of Genetic Liability for Minor and Major Depression 97
No differences in threshold liability distributions were observed within twin
pairs, and across sex and zygosity groups, in either study cohort. Importantly, none
of the multiple-threshold model goodness-of-fit tests (one for each zygosity group)
were significant at the 5% level within the over 50’s study cohort (Table 5.2).
Similarly, only one multiple-threshold model goodness-of-fit test was nominally
significant (M-M DZ twin pairs, p = 0.01) in the Twin 89 cohort (Table 5.2);
however, considering goodness-of-fit tests were performed for each of the 5 zygosity
groups and 4 additional combined groupings, this finding is not considered study-
wide significant. Therefore, these results support the validity of the multiple
threshold model for the DSM-IV classifications for minor and major depression, and
indicate that they can be conceptualised as different levels of severity on a single
dimension of liability.
Table 5.2. Liability threshold model fit p-values.
Twin pair Aged TE
Complete pairs 0.41 0.53
MZ 0.73 0.75 MZf 0.64 0.81
MZm 0.62 0.92
DZ 0.46 0.43 DZss 0.66 0.66
DZf 0.34 0.21
DZm 0.64 0.01 DZos 0.46 0.46
Aged: over 50’s (aged) cohort; TE: Twin 89
(TE) cohort; MZ: monozygotic; MZf: MZ female; MZm: MZ male; DZ: dizygotic; DZss:
DZ same sex; DZf: DZ female; DZm: DZ male;
DZos: DZ opposite-sex.
The polychoric correlations for the varying depression classifications were
approximately two times larger in MZ compared to DZ twin pairs within the over
50’s cohort (Table 5.3). Similarly, with the exception of the MZ non-depressed–
minor depression correlation (due to small cell counts), the polychoric correlations
for the major depression, two-category depression, and three-category depression
were at least two to three times higher in MZ compared to DZ twin pairs within the
Twin 89 cohort (Table 5.3). The observed MZ > DZ correlations, indicate additive
genetic factors contribute to the heritability of depression.
98 Chapter 5: A Continuum of Genetic Liability for Minor and Major Depression
Table 5.3. Polychoric correlations with their 95% confidence intervals for depression according to zygosity.
Aged TE
Depression classification MZ DZ MZ DZ
non-depressed, minor depression 0.37 (0.17-0.56) 0.21 (-0.04-0.45) 0.05 (-0.22-0.31) 0.17 (-0.04-0.39)
non-depressed, major depression 0.46 (-0.01-0.93) - 0.49 (0.41-0.58) 0.19 (0.10-0.28)
non-depressed, minor or major depression 0.49 (0.33-0.64) 0.25 (0.05-0.46) 0.43 (0.34-0.51) 0.18 (0.10-0.27)
non-depressed, minor depression, major depression 0.48 (0.33-0.62) 0.24 (0.04-0.43) 0.43 (0.35-0.51) 0.17 (0.09-0.25)
Aged: over 50’s (aged) cohort; TE: Twin 89 (TE) cohort; MZ: monozygotic; DZ: dizygotic
Chapter 5: A Continuum of Genetic Liability for Minor and Major Depression 99
The best-fitting model for all depression classifications in the over 50’s cohort
was the AE model (Table 5.4). Similarly, the best-fitting model for major depression,
two-category depression, and three-category depression was the AE model, in the
Twin 89 cohort.
Within the over 50’s cohort unique additive genetic factors were estimated to
explain approximately 37% of the heritability of minor depression. Similarly, unique
additive genetic factors were estimated to explain approximately 46% and 45% of
the heritability of major depression, within the over 50’s and Twin 89 cohorts,
respectively (Table 5.4). Significantly, the heritability of the two-category model
(combining minor depression and major depression) was estimated at 48% (95% CI:
33–62%), which was almost indistinguishable to the three-category model estimate
of 47% (95% CI: 33–60%), in the over 50’s cohort. The observed indistinguishability
of the estimates for unique additive genetic factors between two-category depression
(A: 40%, 95% CI: 23-48%) and three-category depression (A: 40%, 95% CI: 32-
47%) was replicated within the Twin 89 cohort. No significant evidence for sex-
specific genetic effects was observed within the over 50’s or Twin 89 cohorts.
Of the 17 loci robustly associated with major depression in Europeans (Hyde et
al., 2016), two were nominally (p < 0.1) associated with major depression in our
Australian dataset, SNP rs10514299 between TMEM161B and MEF2C (allele T:
odds ratio [OR] = 1.10, 95% CI = 1.03-1.17; p = 0.006) and SNP rs11209948 near
NEGR1 (allele T: OR = 1.07, 95% CI = 1.01-1.12; p = 0.05). Broadening our case
phenotype to include an additional 620 unrelated minor depression cases (providing a
total of 4,284 unrelated cases), increased the evidence for association with depression
at both loci, producing more significant p-values of 0.003 and 0.03, respectively.
Comparable results were observed utilising a quantitative, three-category depression
classification with p-values of 0.005 and 0.04, respectively.
100 Chapter 5: A Continuum of Genetic Liability for Minor and Major Depression
Table 5.4. Fit statistics and variance estimates (with their 95% confidence intervals) from univariate structural equation modelling.
Aged TE
Model -2LL p-value
(ACE)
p-value
(ADE) AIC A C (or D) E -2LL
p-value
(ACE)
p-value
(ADE) AIC A C (or D) E
non-depressed, minor depression
ACE 1446.58 - - -3323.42 0.32
(0.00-0.54)
0.05
(0.00-0.43)
0.63
(0.46-0.84) 2203.82 - - -3702.18
0.00
(0.00-0.35)
0.14
(0.00-0.30)
0.87
(0.65-1.00)
AE 1446.61 0.87 1.00 -3325.40 0.37
(0.18-0.54) -
0.63
(0.46-0.82) 2204.58 0.39 1.00 -3703.43
0.15
(0.00-0.37) -
0.85
(0.63-1.00)
CE 1447.58 0.32 - -3324.42 - 0.30
(0.13-0.45)
0.70
(0.46-0.82) 2203.82 1.00 - -3704.18 -
0.14
(0.00-0.30)
0.87
(0.70-1.00)
E 1460.62 8.90 × 10-4 9.00 × 10-4 -3313.38 - - 1.00
(1.00-1.00) 2206.15 0.31 0.46 -3703.85 - -
1.00
(1.00-1.00)
ADE 1446.61 - - -3323.40 0.37
(0.00-0.54) 0.00
(0.00-0.54) 0.63
(0.45-0.82) 2204.58 - - -3701.43
0.15 (0.00-0.37)
0.00 (0.00-0.36)
0.85 (0.63-1.00)
non-depressed, major depression
ACE 475.73 - - -3956.27 0.46
(0.00-0.83) 0.00
(0.00-0.62) 0.54
(0.17-1.00) 5717.52 - - -2996.48
0.45 (0.29-0.53)
0.00 (0.00-0.12)
0.55 (0.47-0.63)
AE 475.73 1.00 0.49 -3958.27 0.46
(0.00-0.83) -
0.54
(0.17-1.00) 5717.517 1.00 0.28 -2998.48
0.45
(0.37-0.53) -
0.55
(0.47-0.63)
CE 476.73 0.32 - -3957.27 - 0.29
(0.00-0.66)
0.71
(0.34-1.00) 5735.63 2.1 × 10-5 - -2980.37 -
0.31
(0.24-0.38)
0.69
(0.62-0.76)
E 478.03 0.32 0.25 -3957.97 - - 1.00
(1.00-1.00) 5813.19 1.7 × 10-21 9.4 × 10-22 -2904.81 - -
1.00 (1.00-1.00)
ADE 475.27 - - -3956.74 0.00
(0.00-0.81)
0.51
(0.00-0.85)
0.49
(0.15-1.00) 5716.37 - - -2997.64
0.25
(0.00-0.52)
0.23
(0.00-0.55)
0.52
(0.44-0.62) Two-category depression (non-depressed, minor depression or major depression)
ACE 1655.41 - - -3214.59 0.44
(0.00-0.61)
0.04
(0.00-0.45)
0.52
(0.38-0.69) 6353.91 - - -3088.09
0.40
(0.22-0.48)
0.00
(0.00-0.14)
0.60
(0.52-0.68)
AE 1655.44 0.88 1.00 -3216.56 0.48
(0.33-0.62) -
0.52
(0.38-0.67) 6353.91 1.00 0.56 -3090.09
0.40
(0.32-0.48) -
0.50
(0.52-0.68)
CE 1658.31 0.09 - -3213.69 - 0.51
(0.39-0.49) 0.61
(0.49-0.74) 6368.07 2.0 × 10-4 - -3075.93 -
0.28 (0.22-0.34)
0.72 (0.66-0.78)
E 1690.12 2.90 × 10-8 2.90 × 10-8 -3183.88 - - 1.00
(1.00-1.00) 6445.22 1.5 × 10-20 1.3 × 10-20 -3000.78 - -
1.00
(1.00-1.00)
ADE 1655.44 - - -3214.56 0.48
(0.00-0.62)
0.00
(0.00-0.60)
0.52
(0.38-0.67) 6353.57 - - -3088.43
0.31
(0.00-0.48)
0.11
(0.00-0.47)
0.58
(0.50-0.67)
Table 5.4 footnote on page 101.
Chapter 5: A Continuum of Genetic Liability for Minor and Major Depression 101
Table 5.4. Continued Fit statistics and variance estimates (with their 95% confidence intervals) from univariate structural equation modelling.
Aged TE
Model -2LL p-value
(ACE)
p-value
(ADE) AIC A C (or D) E -2LL
p-value
(ACE)
p-value
(ADE) AIC A C (or D) E
Three-category depression (non-depressed, minor depression, major depression)
ACE 1912.56 - - -2955.44 0.47
(0.00-0.60)
0.01
(0.00-0.41)
0.53
(0.40-0.69) 8281.43 - - -1158.57
0.40
(0.24-0.47)
0.00
(0.00-0.00)
0.60
(0.53-0.68)
AE 1912.56 0.99 1.00 -2957.44 0.47
(0.32-0.60) -
0.53
(0.40-0.67) 8281.43 1.00 0.42 -1160.57
0.40
(0.32-0.47) -
0.60
(0.53-0.68)
CE 1916.14 0.06 - -2953.86 - 0.37
(0.25-0.49)
0.53
(0.51-0.75) 8298.23 4.2 × 10-5 - -1143.78 -
0.28
(0.22-0.33)
0.72
(0.67-0.78)
E 1948.85 1.30 × 10-8 1.30 × 10-8 -2923.15 - - 1.00
(1.00-1.00) 8381.43 1.9 × 10-22 1.4 × 10-22 -1062.57 - -
1.00
(1-00-1.00)
ADE 1912.56 - - -2955.44 0.47
(0.00-0.60) 0.01
(0.00-0.59) 0.53
(0.40-0.67) 8280.79 - - -1159.21
0.27 (0.00-0.46)
0.14 (0.00-0.14)
0.59 (0.51-0.67)
Aged: over 50’s (aged) cohort; TE: Twin 89 (TE) cohort; -2LL: minus two log likelihood; A: additive genetic component; C: common environmental component E: unique environmental component
102 Chapter 5: A Continuum of Genetic Liability for Minor and Major Depression
5.5 DISCUSSION
Findings from the present study of two independent twin cohorts, indicate the
heritability of minor depression has a genetic contribution. Although, the heritability
of minor depression appears larger in the over 50’s cohort compared to the Twin 89
cohort. In contrast, the heritability estimates for major depression were comparable at
approximately 46% in individuals over 50 and 45% in 23-38 year olds. Within each
cohort the contribution of additive genetic factors was comparable between the two-
category and three-category depression classification. However, the heritability
estimates were larger at 47-48% in the over 50’s cohort compared to 40% in the
Twin 89 cohort. The differences in heritability estimates between the cohorts are
potentially attributable to the time periods assessed within each study cohort (i.e.,
current depression in the over 50’s study compared to lifetime depression in the Twin
89 cohort). Additionally, the difference in depression classifications between the
study cohorts explains the higher major depression prevalence within the Twin 89
cohort. Kendler and colleagues have reported a comparable elevation in lifetime
major depression prevalence within an independent cohort (Kendler et al., 1992).
The authors postulated the higher prevalence of major depression was likely
attributable to a lower average cohort age than national population cohorts and use of
self-report rather than highly structured psychiatric interview—which may
underestimate population rates of major depression. In 2016, Zeng and colleagues
showed self-declared depression is a valid alternative to MDD in genetic studies,
reporting common genetic effects were highly correlated with significant genetic
contributions associated with both classifications (Zeng et al., 2016).
The near identical heritability estimates of the two-category and three-category
depression classifications and the results of the liability threshold model indicate
minor depression and major depression lie on a single genetic liability continuum,
with major depression more severe but not etiologically distinct from minor
depression. Although, we note that the evidence for a genetic contribution to the
heritability of minor depression in younger adults is weak; this is likely due to a
relative lack of power (i.e., low number of minor depression cases) within the Twin
89 cohort. Indeed, the heritability estimate for the broad, two-category depression
classification indicates the applicability of a broad depression phenotype is not
specific to the over 50’s cohort. Therefore, broadening the depression phenotype in
Chapter 5: A Continuum of Genetic Liability for Minor and Major Depression 103
genetic studies by including individuals with a diagnosis of ‘minor depression or
major depression’ should facilitate the identification of genetic risk factors associated
with depression due to improved power via increased sample size. Such improved
power will be readily provided through re-analysis of existing GWA datasets (which
currently exclude minor depression-like cases from analysis), and more cost-
effective collection of depression cases in future studies. As previously outlined by
Sullivan (Sullivan, 2012) GWA studies will continue to be of great importance for
identification of the underlying biology and genetic architecture of psychiatric
disorders. Indeed, the MDD working group of the PGC have previously emphasised
the absence of reference to the underlying biology or pathophysiology within the
MDD diagnosis (Major Depressive Disorder Working Group of the Psychiatric
GWAS Consortium et al., 2013).
Previous MDD GWA analyses have discussed possible approaches to increase
power and enable identification of genetic risk loci associated with MDD (Wray et
al., 2012). The first approach involves utilising more homogenous MDD case
samples. In 2015, the CONVERGE consortium utilised this method, by selecting
5,303 Han Chinese women with recurrent MDD (of which 85% have the severe
melancholic subtype) and 5,337 Han Chinese female controls screened to exclude
MDD, to identify the first SNP loci robustly associated with severe recurring MDD
(CONVERGE Consortium, 2015). Furthermore, in 2017, Power and colleagues
utilised additional phenotypic data to stratify cases and thereby reduce heterogeneity,
which enabled the identification of a genetic risk locus associated with MDD onset in
adults aged over 27 years (Power et al., 2017). Stratification of MDD cases based on
symptom dimensions represents an alternative method of utilising phenotypic data to
reduce heterogeneity within GWA studies; with Pearson and colleagues (Pearson et
al., 2016) showing common SNPs explain varying proportions of the variation in the
depression symptom dimensions of core depression symptoms, insomnia, appetite,
and anxiety symptoms (SNP-based heritability = 14.3%, 30.3%, 29.6%, and 4.7%,
respectively). Meanwhile, a complementary approach is to obtain larger sample sizes
which are more representative of the general population. This approach can be
achieved by broadly defining depression, to detect the common variation of small
effect given the relatively high prevalence and low heritability of MDD (Wray et al.,
2012).
104 Chapter 5: A Continuum of Genetic Liability for Minor and Major Depression
To demonstrate the utility of using a broader depression phenotype, we
examined the association signal of the 17 SNPs reported by Hyde and colleagues
(Hyde et al., 2016), in Table 2, reaching genome-wide significant association with
major depression, utilising our Australian GWA dataset. In our Australian sample of
3,664 unrelated major depression cases and 7,113 unrelated controls, SNP
rs10514299 between TMEM161B and MEF2C and SNP rs11209948 near NEGR1,
were nominally associated with major depression (p ≤ 0.05).
TMEM161B encodes the transmembrane protein 161B and is expressed in the
brain (cortex, hypothalamus, anterior cingulate cortex (BA24), and cerebellum).
Similarly, MEF2C encodes myocyte enhancer factor 2C and has been associated
with phenotypes involved in the central nervous system (mental retardation,
stereotypic movements, epilepsy and cerebral malformations). Meanwhile, NERG1
encodes neuronal growth regulator 1, is expressed in the brain (cerebellar
hemisphere, cerebellum, and pituitary), and associated with body mass index,
subcutaneous adipose tissue, weight, and age-at -onset of menarche. Our current
knowledge of the functional role of TMEM161B, MEF2C, and NEGR1 substantiates
the central nervous systems involvement in the pathophysiology of depression.
Broadening our case phenotype to include an additional 620 unrelated minor
depression cases (providing a total of 4,284 unrelated cases), increased the statistical
evidence for association with depression at both loci. Although a subset (1,450 cases
and 1,711 controls) of the 3,664 Australian major depression cases and 7,113
controls were part of the PGC MDD GWA (Major Depressive Disorder Working
Group of the Psychiatric GWAS Consortium et al., 2013) that was meta-analysed in
Hyde et al (Hyde et al., 2016), these results provide proof-of-principle for using a
broader depression phenotype to increase power in genetic association studies of
depression. In addition, the study by Hyde and colleagues, provides evidence that
utilising large self-report depression data, which broadens the MDD phenotype due
to the lack of restriction to clinically validated MDD cases, is an effective strategy
for overcoming the large heterogeneity of depression (Hyde et al., 2016). Further
evidence for the utility of broad depression phenotypes in genetic studies is provided
by the investigation of ‘depression symptoms’ conducted by Okbay and colleagues
(2016) and ‘MDD or depression symptoms’ by Direk and colleagues (2016).
Chapter 5: A Continuum of Genetic Liability for Minor and Major Depression 105
Continued use of broad depression phenotypes and large cohorts without
detailed clinical evaluation, such as from large ongoing commercial (e.g., 23andMe
and Kaiser Permanente) and public (e.g., UK Biobank and Generation Scotland)
datasets should therefore identify additional genetic risk factors, and provide the
crucial clues to further elucidating the complex molecular pathways underlying
MDD—which can then be characterised with respect to particular features of
depression via the study of specific patient subgroups in deeply-phenotyped clinical
cohorts.
106
In Chapter 5, minor depression and major depression, were shown to lie on a single
genetic continuum. Additionally, minor depression, major depression, and a broad
depression phenotype (minor or major depression) were all shown to have
significant, additive genetic contributions. Similarly, fatigue was shown to have a
significant additive genetic contribution in Chapter 4. Building on the results from
Chapter 4 and Chapter 5, the following chapter aims to determine if shared genetic
factors explain a proportion of the variation in depression and fatigue and
characterise the type of relationship that exists between the traits, in older adults.
Chapter 6: Shared Genetic Factors in the Co-Occurrence of Depression and Fatigue 107
Chapter 6: Shared Genetic Factors in the
Co-Occurrence of Depression
and Fatigue
This chapter comprises the following published article:
Corfield, E. C., Martin, N. G., & Nyholt, D. R. (2016). Shared Genetic Factors in the
Co-Occurrence of Depression and Fatigue. Twin Research and Human Genetics, 1-9.
doi:10.1017/thg.2016.79
108 Chapter 6: Shared Genetic Factors in the Co-Occurrence of Depression and Fatigue
QUT Verified Signature
QUT Verified Signature
Chapter 6: Shared Genetic Factors in the Co-Occurrence of Depression and Fatigue 109
6.1 ABSTRACT
Depression and fatigue have previously been suggested to share an underlying
genetic contribution. The present study aims to investigate and characterise the
familiality and genetic relationship between depression and fatigue. The familiality
of depression and fatigue was assessed by calculating relative risks, measured by the
prevalence ratio, within 643 monozygotic (MZ) and 577 dizygotic (DZ) twin pairs.
Bivariate twin modelling was utilised to assess the magnitude of shared heritability
between depression and fatigue. Finally, the relationship between depression and
fatigue was investigated using the co-twin control method, to determine whether the
association is explained by causal or non-causal models. We observed an increased
risk of fatigue in co-twins of probands with depression and increased risk of
depression in co-twins of probands with fatigue. Higher risks were observed in MZ
compared to DZ twin pairs and bivariate heritability analyses indicated significant
genetic components for depression and fatigue, with heritability estimates of 48%
and 41%, respectively. Importantly, a significant additive genetic correlation of 0.71
(95% confidence interval [CI] = 0.51-0.92) and bivariate heritability of 21% (95% CI
= 10-35%) was observed between depression and fatigue. Furthermore, results from
the co-twin control method indicate a non-causal genetic relationship likely explains
the association between depression and fatigue. Notably, the contribution of shared
genetic factors remained significant independent of the overlapping symptoms,
indicating the relationship between co-occurring depression and fatigue is primarily
due to shared genetic factors rather than overlapping symptomatology.
110 Chapter 6: Shared Genetic Factors in the Co-Occurrence of Depression and Fatigue
6.2 INTRODUCTION
Depression and fatigue are highly prevalent traits and are associated with a
considerable reduction in quality of life. The heritability of depression and fatigue
has been estimated to range from 17-78% and 18-51%, respectively (Schur et al.,
2007; Sullivan et al., 2000; Sullivan et al., 2005). The wide ranges are likely
attributable to differences in ethnicity and gender distribution within the study
populations. Shared genetic aetiologies have been implicated in studies investigating
the heritability of psychological distress, anxiety, depression, and fatigue, and
insomnia, fatigue, and depression, respectively (Hickie et al., 1999b; Hur et al.,
2012). However, the underlying mechanisms associated with depression and fatigue
which could explain the high levels of comorbidity are poorly understood.
Two studies have tested for a shared genetic influence to depression and
fatigue. In the first study, the heritability of lifetime-ever disabling fatigue (assessed
by parental report using the disabling fatigue measure (Farmer et al., 1999)) and
depression within the past three months (assessed by the mood and feelings
questionnaire in individuals over 11 (Costello & Angold, 1988)) was investigated in
children (aged 8-17) (Fowler et al., 2006). The second study examined the genetic
relationship between abnormal fatigue (assessed by the Chalder Fatigue
Questionnaire (Chalder et al., 1993)) and an indicator of lifetime-ever depression
(assessed by two screening questions of the Composite International Diagnostic
Interview (World Health Organization, 1990) regarding depressed mood and loss of
interest—the two core symptoms of a major depressive episode, as defined by the
Diagnostic and Statistical Manual of Mental Disorders (DSM) (American Psychiatric
Association, 2013))—in a Sri Lankan population (aged ≥ 15) (Ball et al., 2010b).
Although both studies indicated depression and fatigue have a shared genetic
contribution, the genetic relationship between co-occurring depression and fatigue
was not well characterised and their results are not readily comparable due to the fact
that risk for depression differs by age and sex (Bijl et al., 2002; Centers for Disease
Control and Prevention, 2010; Kessler et al., 2003).
Differences in familial and genetic risk for depression have been identified
with age. Older adults exhibit the lowest prevalence of a current depression diagnosis
and a comparable risk of onset between males and females (Bebbington et al., 1998;
Faravelli et al., 2013). Additionally, symptomatology differences have been observed
Chapter 6: Shared Genetic Factors in the Co-Occurrence of Depression and Fatigue 111
between depression patients in different age ranges (Hybels et al., 2012; Wilkowska-
Chmielewska et al., 2013). However, little is known about the genetic relationship
between co-occurring depression and fatigue in older adults. Therefore, the present
study utilises relative risks and twin modelling to investigate the familiality and
heritability of depression and fatigue within older adults. Furthermore, the co-twin
control method was utilised to investigate whether the association between
depression and fatigue is explained by a causal model or shared underlying aetiology.
6.3 MATERIALS AND METHODS
6.3.1 Study Cohort
The present study was conducted using data from the over 50’s (aged) study
conducted by the genetic epidemiology group within QIMR Berghofer. The study
invited 2,281 twin pairs from the Australian twin registry to complete a mailed
Health and Lifestyle Questionnaire (Bucholz et al., 1998; Mosing et al., 2012). The
present study utilised responses to the Schedule of Fatigue and Anergia (SOFA), the
twelve-item General Health Questionnaire (GHQ), and the fourteen-item Delusions
Symptoms-States Inventory, States of Anxiety and Depression (DSSI/sAD)
questionnaires (Bedford & Deary, 1997; Goldberg & Blackwell, 1970; Hickie et al.,
1996). The study cohort utilised here, overlaps the cohort used by Hickie et al.
(1999b) to investigated the multivariate heritability of psychological distress,
anxiety, depression, and fatigue. However, the present study focuses on depression
and fatigue, including looking at MDD and MiDD.
6.3.2 Diagnosis of Depression and Fatigue
MDD and MiDD were classified using the nine criteria of a major depressive episode
(depressed mood, anhedonia, a change in weight or appetite, insomnia or
hypersomnia, psychomotor agitation or retardation, fatigue or loss of energy, feelings
of worthlessness or excessive guilt, inability to concentrate or make decisions, and
thoughts about death, suicidal thoughts, suicidal plans, or suicidal attempts), as
defined by the DSM-IV criteria (American Psychiatric Association, 2000). A
combination of questions from the GHQ and DSSI/sAD were used to assess
depression, through assignment of specific questions to the appropriate criterion of
the major depressive episode criteria. When multiple questions assessed a criterion at
least one positive response indicated the individual exhibited a symptom from the
112 Chapter 6: Shared Genetic Factors in the Co-Occurrence of Depression and Fatigue
specific criterion. Each criterion was assessed by assigning one to the criterion if a
symptom was exhibited by the individual and zero if none of the symptoms for the
criterion were met. The survey did not contain any assessment of change in weight or
appetite; therefore, this criterion of a major depressive episode was not assessed. The
scores of the remaining eight criteria assessed were summed if the individual
screened positive (score > 0) for depressed mood and/or anhedonia, otherwise the
individual was assigned a total score of zero. Individuals were classified as MDD,
MiDD, or non-depressed, if they had a score of five or more, two to four, or less than
two, respectively.
The SOFA was originally designed to identify chronic fatigue syndrome cases.
Therefore, physical, neurocognitive, and neurovegetative fatigue symptoms are
assessed by the questionnaire. Consequently, the fatigued state identified by the
SOFA is distinct from the fatigue experienced within a major depressive episode.
Ten questions are contained in the SOFA; however, a shorter eight-item version was
included in the survey due to two questions being replicated within the GHQ.
Individuals were classified as fatigued if they reported three or more of the ten
fatigue symptoms (muscle pain at rest, post-exertional muscle pain, post-exertional
muscle fatigue, post-exertional fatigue, hypersomnia, insomnia, poor concentration,
speech problems, poor memory, and headaches).
6.3.3 Familial Clustering
Familiality between depression and fatigue was assessed by calculation of relative
risks (RR), assessed by the prevalence ratio, with their 95% confidence intervals
(CI). Initially, cross-tabulation was utilised to assess the depression and fatigue status
within twin pairs based on zygosity groupings. The method by Nyholt and colleagues
(2004) was utilised to estimate the risk within the complete cohort and same-sex twin
pairs, where the cross-tabulations from using twin 1 or twin 2 as the proband were
averaged. RR were also calculated from the averaged cross-tabulations within same-
sex monozygotic (MZ) and dizygotic (DZ) twin pairs and opposite-sex DZ twin pairs
relative to non-depressed or non-fatigued status. Initially, the risk of fatigue in co-
twins of depressed probands was calculated. Similarly, the risk of depression in co-
twins of fatigued probands was calculated.
To assess the familiality of depression and fatigue independent of their
overlapping symptoms, the risk of fatigue in co-twins was also estimated in the
Chapter 6: Shared Genetic Factors in the Co-Occurrence of Depression and Fatigue 113
subgroup of depressed individuals without overlapping DSM depression symptoms
(i.e., insomnia, poor concentration, and hypersomnia). Similarly, the risk of
depression in co-twins was also estimated in the subgroup of fatigued individuals
without overlapping fatigue symptoms (i.e., insomnia, inability to concentrate, and
loss of energy).
6.3.4 Genetic Analysis
The association between depression and fatigue was assessed by looking at twin,
phenotypic, and cross-twin cross-trait correlations within MZ and DZ twin pairs.
Polychoric correlations were calculated (due to the binary coding utilised within the
cohort) using the polycor package in R (R Core Team, 2014). Twin correlations
assessed the association of a single trait across a twin pair, phenotypic correlations
assessed the association of two traits within individuals, and cross-twin cross-trait
correlations assessed the association of two traits across a twin pair. Twin and cross-
twin cross-trait correlations which are larger in MZ compared to DZ twin pairs
indicate the aetiology of the traits has a genetic contribution.
Bivariate twin models were calculated to estimate the relative contribution of
genetic and environmental factors on the covariation of depression and fatigue. The
bivariate twin modelling was conducted utilising the Cholesky decomposition which
allowed the genetic and environmental factors of the first trait to load onto the
second trait (Neale et al., 1992). The model contains another set of genetic and
environmental factors which are unique to the second trait. Such twin modelling
partitions the observed phenotypic variance into specific components. Briefly,
phenotypic differences between MZ and DZ twin pairs (MZ twin pairs have the same
genotype and common environment while DZ twin pairs only share 50% of their
genes but have the same common environment) are used to estimate the contribution
of additive genetic (A), dominant (non-additive) genetic (D), common environmental
(C), and unique environmental (E) variance components (Neale et al., 1992).
Additionally, the genetic (rg) and environmental (re) correlation between depression
and fatigue was calculated as a measure of the overlap in gene and environmental
sets, respectively.
Heritability estimates were calculated utilising structural equation modelling,
including the threshold model. The threshold model posits that distinct traits
represent a single, normally distributed, severity continuum. Initially, a single
114 Chapter 6: Shared Genetic Factors in the Co-Occurrence of Depression and Fatigue
threshold was utilised for depression, whereby individuals were separated into a
broad, two-category (non-depressed, MiDD/MDD) depression definition.
Additionally, two thresholds were used to assess the three-category (non-depressed,
MiDD, MDD) depression definition. A single threshold was used within all models
to separate non-fatigued and fatigued individuals. Corrections for (linear) age and
sex effects were included in all models which were fitted using the OpenMx package
in R (Boker et al., 2011; R Core Team, 2014). The significance of the variance
components was assessed by comparing the fit of the full model (ACE/ADE) to the
nested submodels (AE, CE, and E) where individual variance components were
dropped from the model. The goodness of fit parameters used to assess the
differences in the twin models were the likelihood-ratio chi-square test (χ2), the
difference in degrees of freedom (Δ df), and p-value. Additionally, model fit was
compared utilising Akaike’s Information Criteria (AIC); with the lowest AIC
indicating the most parsimonious model (Akaike, 1973, 1974).
Polychoric correlations and bivariate heritability estimates were also estimated
for depression and fatigue in the subgroup of twins without overlapping symptoms
(i.e., insomnia, concentration problems, hypersomnia, and loss of energy).
6.3.5 Relationship Analysis
The co-twin control method was utilised to determine the type of relationship that
exists between depression and fatigue (Kendler et al., 1993; Kendler et al., 1999). A
causal relationship exists when a risk factor directly causes a phenotype, without
familial confounding. Meanwhile, a non-causal model exists when familial (A and C)
factors completely explain the correlation between the risk factor and the trait.
Furthermore, non-causal relationships exist, where the association between the risk
factor and the trait is mediated by familial factors (A or C) (Kendler et al., 1993;
Kendler et al., 1999; McGue et al., 2010).
The co-twin control method conducted throughout the study followed the
protocol outlined by Ligthart et al (2010), where the odds ratio (OR) of the trait is
calculated based on the presence or absence of the risk factor within three cohorts:
MZ and DZ twin pairs with the trait that are discordant for the risk factor and a
general population sample. The over 50’s (aged) study contained 200 MZ twin pairs
and 215 DZ twin pairs with a measure of depression which were discordant for
fatigue. Similarly, 99 MZ twin pairs and 96 DZ twin pairs with a measure of fatigue
Chapter 6: Shared Genetic Factors in the Co-Occurrence of Depression and Fatigue 115
were discordant for depression. The general population sample of 1,266 individuals
was obtained by selecting all unpaired twin singles and randomly selecting a single
individual from each complete twin pair (i.e., one individual from each family)
excluding those discordant for fatigue or depression.
If a causal relationship exists between the risk factor and the trait of interest the
three cohorts are expected to have comparably elevated ORs (Figure 6.1). Similarly,
under non-causal models, the general population is expected to show increased odds
of exhibiting the trait given the presence of the risk factor. However, MZ and DZ
cohorts are expected to exhibit varying OR patterns, although the association should
always be smaller than within the general population. If the relationship between the
risk factor and the trait is non-causal no association is expected in the MZ cohort
while the DZ cohort is expected to exhibit a small association (Figure 6.1). Similarly,
if a non-causal relationship exists between the risk factor and the trait which is
mediated by shared environment both the MZ and DZ cohorts are expected to exhibit
a similar association (Figure 6.1). Finally, under a non-causal model mediated by
genetic factors the MZ cohort is expected to exhibit a smaller association than the
DZ cohort (Figure 6.1).
Figure 6.1. Expected outcomes of the co-twin control method under the causal, non-causal, non-
causal shared environment, and non-causal genetic models within the general population (light grey),
discordant DZ twin pairs (grey) who share 50% of their genetics and 100% of their common
environment, and discordant MZ twin pairs (dark grey) who share 100% of their genetics and
common environment. Under a causal model an association is expected within all three groups. Under
a non-causal model, an association is expected within the general population, discordant DZ cohort
will have a small association, and discordant MZ cohort will have no association. Similarly, under the
non-causal shared environmental model, discordant DZ and MZ twin pairs have a small, equal
association. Finally, under the non-causal genetic model, discordant DZ twin pairs have an
association, whereas discordant MZ twin pairs have a smaller association.
116 Chapter 6: Shared Genetic Factors in the Co-Occurrence of Depression and Fatigue
6.4 RESULTS
The over 50’s (aged) study was a population-based cohort of 2,281 Australian twin
pairs. A total of 1,061 twin pairs were removed due to incomplete depression and
fatigue classifications for one or both twins. The remaining 1,220 twin pairs
consisted of 643 MZ twin pairs (491 female-female and 152 male-male twin pairs)
with a mean age of 61.5 ± 8.9 (range = 50-92) and 577 DZ twin pairs (263 female-
female, 73 male-male, 136 female-male, and 105 male-female twin pairs) with a
mean age of 61.2 ± 8.2 (range = 50-90). The prevalence of depression (either MDD
or MiDD) and fatigue was 11.0% (11.7% of females, 9.3% of males) and 29.6%
(32.5% of females, 24.7% of males), respectively.
6.4.1 Relative Risks
Initially, all individuals who participated in the over 50’s (aged) study were assessed
for both MDD and MiDD (Supplementary Table 6.1). The present study focused on a
two-category, broad definition of depression defined as either MDD or MiDD.
Cross-tabulation within MZ, same-sex DZ, and opposite-sex DZ twin pairs was
based on depressed or non-depressed and fatigued or non-fatigued classifications
(Table 6.1).
Table 6.1. Cross-tabulationa of two-category depression and fatigue status within twin pairs.
Non-depressed Depressed Total
MZ
Non-fatigued 411 35 446
Fatigued 157.5 39.5 197
Total 568.5 74.5 643
DZss
Non-fatigued 212 22.5 234.5
Fatigued 90.5 11 101.5
Total 302.5 33.5 336
DZos [F-M]
Non-fatigued 163 16 179
Fatigued 54 8 62
Total 217 24 241
DZos [M-F]
Non-fatigued 159 20 180
Fatigued 53 9 61
Total 212 29 241
MZ: monozygotic; DZss: same-sex dizygotic; DZos: opposite-sex dizygotic; F-M: female-male; M-F: male-female. aTables were made symmetrical in same-sex twin pairs by averaging over using either twin 1 or twin 2 as proband. For example,
within the complete twin pairs there were 155 twin pairs where twin 1 was fatigued and twin 2 was non-depressed and 160
twin pairs where twin 2 was fatigued and twin 1 was non-depressed. Therefore, the cross-tabulation averaging over twin 1 or twin 2 as proband is (155+160)/2=157.5.
We observed an increased risk of fatigue in co-twins of probands with
depression and increased risk of depression in co-twins of probands with fatigue,
indicating a significant familial association between the traits (Table 6.2). Strong
evidence for a genetic contribution is provided by the higher risk observed in MZ
compared to DZ twin pairs. In particular, the risk of fatigue in co-twins of depressed
probands was 1.91 (95% CI = 1.49-2.46) in MZ twin pairs compared to 1.10 (95% CI
Chapter 6: Shared Genetic Factors in the Co-Occurrence of Depression and Fatigue 117
= 0.66-1.84) in same-sex DZ twin pairs. Similarly, the risk of depression in co-twins
of fatigued probands was 2.56 (95% CI = 1.67-3.90) in MZ twin pairs compared to
1.13 (95% CI = 0.57-2.23) in same-sex DZ twin pairs. Analysis of familial clustering
within males and females indicated a similar pattern of risks (Supplementary Table
6.2). A comparable pattern of MZ to DZ RR was also observed for the separate
MiDD and MDD cases, but with MDD producing further increased RR
(Supplementary Table 6.3). Importantly, a similar pattern of risks was observed
when they were estimated independently of depression and fatigue overlapping
symptoms (Supplementary Table 6.7 and Supplementary Table 6.8).
118 Chapter 6: Shared Genetic Factors in the Co-Occurrence of Depression and Fatigue
Table 6.2. Relative riska of two-category depression and fatigue within monozygotic (MZ), same-sex dizygotic (DZss), and opposite-sex dizygotic (DZos) twin pairs.
Proband - co-twin MZ
(643 twin pairs)
DZss
(336 twin pairs)
DZos [F-M]
(241 twin pairs)
DZos [M-F]
(241 twin pairs)
Depressed -non-fatigued 0.65 (0.51-0.83) 0.96 (0.75-1.23) 0.92 (0.71-1.19) 0.89 (0.66-1.19)
Depressed - fatigued 1.91 (1.49-2.46) 1.10 (0.66-1.84) 1.24 (0.69-2.24) 1.34 (0.73-2.47)
Fatigued - non-depressed 0.87 (0.80-0.94) 0.99 (0.91-1.07) 0.96 (0.86-1.06) 0.96 (0.86-1.08)
Fatigued - depressed 2.56 (1.67-3.90) 1.13 (0.57-2.23) 1.44 (0.65-3.21) 1.30 (0.62-2.70)
MZ: monozygotic; DZss: same-sex dizygotic; DZos: opposite-sex dizygotic; F-M: female-male; M-F: male-female. aRelative risks were
calculated with respect to non-depressed or non-fatigued status in twin 1.
Chapter 6: Shared Genetic Factors in the Co-Occurrence of Depression and Fatigue 119
6.4.2 Polychoric Correlations
The twin correlations for depression and fatigue were approximately two and three
times larger in MZ compared to DZ twin pairs, respectively (Table 6.3). Similarly,
the cross-twin cross-trait correlations for depression and fatigue were over twice the
magnitude in MZ compared to DZ twin pairs. This observed MZ > DZ correlations,
indicate additive genetic factors contribute to the association between depression and
fatigue. A similar pattern of polychoric correlations was observed when the analyses
were repeated independently of the overlapping symptoms (Supplementary Table
6.9). Furthermore, comparable patterns of polychoric correlations were observed for
MiDD, MDD, and the three-category depression classification (non-depressed,
MiDD, MDD) (Supplementary Table 6.4).
Table 6.3. Polychoric correlations with their 95% confidence intervals for two-category depression
and fatigue according to zygosity.
Twin 1 Twin 2
Depression Fatigue Depression Fatigue
Monozygotic twin pairs (N = 643)
Twin 1 Depression 1.00
Fatigue 0.49 (0.35-0.62)a 1.00
Twin 2 Depression 0.49 (0.33-0.64)b 0.35 (0.21-0.50)c 1.00
Fatigue 0.33 (0.18-0.48)c 0.43 (0.31-0.54)b 0.49 (0.36-0.62)a 1.00
Dizygotic twin pairs (N = 577)
Twin 1 Depression 1.00 Fatigue 0.51 (0.37-0.65)a 1.00
Twin 2 Depression 0.25 (0.05-0.46)b 0.09 (-0.09-0.27)c 1.00
Fatigue 0.05 (-0.13-0.23)c 0.14 (0.001-0.28)b 0.58 (0.45-0.71)a 1.00 aPhenotypic correlation between depression and fatigue. bTwin correlation. cCross-twin cross-trait correlation.
6.4.3 Bivariate Heritability Estimates
Bivariate model fitting for depression and fatigue indicated that the AE model is the
most parsimonious (Table 6.4). No differences in depression or fatigue threshold
distributions were observed within twin pairs, and across zygosity and sex groups.
Overall, 48% (95% CI = 32-61%) and 41% (95% CI =0.30 – 0.51%) of the variance
in depression and fatigue, respectively, was explained by genetic factors. Also, 52%
(95% CI = 39-68%) and 59% (95% CI = 0.49 - 0.70%) of the variance in depression
and fatigue was explained by unique environmental factors, respectively (Figure 6.2).
Notably, 21% (95% CI = 10-35%) of the variance in depression due to genetic
factors also contributes to the heritability of fatigue (i.e., bivariate heritability of
21%). Also, 7% of the variance in depression due to unique environmental factors
also contributes to the variance in fatigue. The overlap in genetic and unique
environmental factors was supported by the rg of 0.71 (95% CI = 0.51-0.92) and re
120 Chapter 6: Shared Genetic Factors in the Co-Occurrence of Depression and Fatigue
of 0.35 (95% CI = 0.17-0.51) between depression and fatigue. Similar results were
obtained for bivariate heritability estimates between MDD or MiDD and fatigue
(Supplementary Table 6.5, Supplementary Figure 6.1, and Supplementary Figure
6.2). Almost identical bivariate heritability estimates were obtained between three-
category depression and fatigue (Supplementary Table 6.5 and Supplementary Figure
6.3). Importantly, the shared genetic contribution to depression and fatigue remained
significant independent of the overlapping symptoms (Supplementary Figure 6.4).
Table 6.4. Bivariate heritability model fits.
Model Minus two log-likelihood χ2 Δ df p-value AIC
ACE 4405.65 -5324.35 AE 4406.74 1.09 3 0.78 -5329.26
CE 4419.35 13.70 3 3.34 × 10-3 -5316.65
E 4483.87 78.22 6 8.35 × 10-15 -5258.13 ADE 4405.20 -0.46 0 NA -5324.80
Note: Fit statistics are compared to ACE model and the best-fitting model is indicated in bold. χ2: likelihood-ratio chi-squared
test; Δ df: difference in degrees of freedom.
Figure 6.2. Path diagram of the bivariate Cholesky model variance estimates (with their 95%
confidence intervals) for two-category depression and fatigue. The observed traits are shown in the
rectangles. Similarly, the latent variables (additive genetic factors: A, and unique environmental
factors: E) are depicted by circles. The arrows depict the relationship between the variables.
6.4.4 Co-twin Control
Assessment of depression as a risk factor for fatigue revealed the ORs in the general
population, discordant DZ twin pairs, and discordant MZ twin pairs were 7.20 (95%
CI: 4.49-11.56), 6.29 (95% CI: 3.35-11.81), and 1.92 (95% CI: 1.09-3.39),
respectively (Figure 6.3). Similarly, assessment of fatigue as a risk factor for
depression revealed the ORs in the general population, discordant DZ twin pairs, and
Chapter 6: Shared Genetic Factors in the Co-Occurrence of Depression and Fatigue 121
discordant MZ twin pairs were 7.20 (95% CI: 4.49-11.56), 5.21 (95% CI: 2.75-9.88),
and 1.98 (95% CI: 1.10-3.55), respectively (Figure 6.3). The pattern of OR exhibited
when fatigue was a risk factor for depression was comparable to when depression
was a risk factor for fatigue. The observed OR pattern indicates that a non-causal
genetic model best describes the relationship between depression and fatigue. A non-
causal genetic relationship was also indicated between fatigue and MiDD
(Supplementary Table 6.6). However, the co-twin control analysis could not be
replicated for fatigue and MDD due to lack of power in the smaller sub-sample.
Figure 6.3. Left: The observed odds ratios (OR) for a current diagnosis of fatigue given a current
diagnosis of depression in the general population (1,266 unrelated twin singles), 99 discordant DZ
twin pairs, and 96 discordant MZ twin pairs. Right: The observed OR for a current diagnosis of
depression given a current diagnosis of fatigue in the general population (1,266 unrelated twin
singles), 200 discordant DZ twin pairs, and 215 discordant MZ twin pairs. In both situations, the
observed OR patterns are consistent with a non-causal genetic model.
6.5 DISCUSSION
Three key findings were identified from the present study. Firstly, depression and
fatigue exhibit a familial component. Secondly, co-occurring depression and fatigue
have considerable genetic overlap. Finally, a non-causal genetic model likely
explains the association between depression and fatigue.
The familial clustering analysis revealed depression and fatigue likely have
shared underlying aetiologies. These results are supported by previous studies which
have reported higher levels of depression in fatigued individuals than the general
population (Cathébras et al., 1992; Walker et al., 1993). However, to our knowledge,
this is the first study to assess and characterise the familial clustering of depression
122 Chapter 6: Shared Genetic Factors in the Co-Occurrence of Depression and Fatigue
and fatigue, with regard to the risk of depression in fatigued individuals and the risk
of fatigue in depressed individuals.
Overlap in genetic factors between depression and fatigue has been indicated in
previous twin studies (Ball et al., 2010b; Fowler et al., 2006; Hickie et al., 1999b;
Hur et al., 2012). Previous multivariate twin studies have identified genetic and
environmental factors which are unique to fatigue. Hur et al (2012) established
common and symptom-specific genetic and environmental factors contributed to the
heritability of self-reported insomnia, fatigue, and depression experienced within a
twelve month period. Similarly, Hickie et al (1999b) determined 44% of the
heritability and 100% of the environmental contribution of fatigue is independent of
psychological distress, anxiety, and depression. Although, the proportion of genetic
factors contributing to fatigue which are independent of depression appears to be
smaller in older adults than in children. Bivariate modelling reported by Fowler et al
(2006) established that 87% and 73% of the heritability for lifetime ever short-
duration fatigue (fatigue experienced for at least one week) and lifetime ever
prolonged fatigue (fatigue experienced for at least one month), respectively, was
independent of depression within the last three months in children. The high
proportion of heritability specific to short-duration fatigue and prolonged fatigue was
substantiated by rg of 0.36 and 0.53, respectively. However, Ball et al (2010b)
indicated unique environmental variance components explained a larger proportion
of the overlap in heritability between fatigue and an indicator of lifetime depression
than familial factors, in Sri Lanka.
Bivariate twin modelling results from the current study indicated 50% of the
heritability and 12% of the environmental contribution of fatigue is shared with
depression. The rg of 0.71 between co-occurring depression and fatigue within adults
(aged over 50) was considerably higher than previous bivariate studies, indicating
larger genetic overlap exists between co-occurring depression and fatigue in older
adults than depression within the past three months and lifetime-ever disabling
fatigue in children. A small environmental overlap between co-occurring depression
and fatigue was also observed (re = 0.35). Differences in the contribution of genetic
and environmental factors to the shared heritability of depression and fatigue
between adults in Australia and Sri Lanka are likely attributable to variation in
phenotypic classification, ethnicity, age, and cultural differences. In particular, the
Chapter 6: Shared Genetic Factors in the Co-Occurrence of Depression and Fatigue 123
measure of depression, utilised within the Sri Lankan population, only assessed the
two core symptoms of the DSM criteria for a major depressive episode. Therefore,
the genetic correlation of depression and fatigue potentially increases with the
number of depression symptoms.
Skapinakis and colleagues (2004) described four possible explanations for the
association between depression and unexplained fatigue observed within their
international study: the causal (depression causes fatigue), reverse causality
(depression is caused by an unknown, disabling illness, such as chronic fatigue
syndrome), common etiology (common risk factors explain the comorbidity between
depression and fatigue), and overlapping criteria (the observed comorbidity between
depression and fatigue is a result of overlapping symptomatology) hypotheses. The
bivariate twin modelling and co-twin control method used within this study enabled
us to investigate which of the four proposed hypotheses drives the association
between depression and fatigue. Our results substantiate the common aetiology
hypothesis, whereby depression and fatigue share common risk factors. In particular,
our results indicate shared genetic factors explain the majority of the correlation. The
larger overlap in heritability between co-occurring depression and fatigue is also
supported by the results of the co-twin control analysis. That is, the determined non-
causal genetic relationship between depression and fatigue adds further support to the
comorbidity between depression and fatigue being primarily due to shared genetic
factors.
Our findings lead us to suggest that overlapping genetic factors could also
underlie the relationship between depression relapse and residual fatigue. Residual
fatigue has a prevalence of 63-98% and 22-49% in partial responders and remitted
patients, respectively, after antidepressant treatment (Fava et al., 2014). Additionally,
fatigue as a symptom of depression has been associated with higher health care
utilisation, 10-20% greater annual health care cost, increased medication uses, and
lower quality of life (Robinson et al., 2015). Furthermore, residual fatigue has been
shown to lead to higher levels of functional impairment and depression relapse.
Considering that currently available antidepressant therapies have been shown to
inadequately treat residual fatigue, we believe research should focus on
understanding the shared mechanisms of depression and fatigue. Such research holds
great potential to facilitate the development of enhanced treatment outcomes, which
124 Chapter 6: Shared Genetic Factors in the Co-Occurrence of Depression and Fatigue
are targeted to shared mechanisms of depression and fatigue. The effective treatment
of depression, focussing on symptomatic treatment of residual fatigue could lower
remission levels, thereby reducing the global burden of depression.
The present study is the first to investigate the type of relationship between co-
occurring depression and fatigue, utilising a two-category and three-category
depression status. Additionally, it is the only study which determined whether a
causal relationship exists between depression and fatigue. A possible limitation of
the study is that depression and fatigue were assessed by self-report not interview
based. However, this allowed depression and fatigue to be assessed independently
without introducing interviewer bias. Furthermore, the study was not confounded by
healthcare seeking behaviour due to the population-based structure of the cohort.
Additionally, the current study focused on an older age group for which some
evidence suggests the risk of depression is similar between the sexes (Bebbington et
al., 1998; Faravelli et al., 2013).
In summary, our results indicate depression and fatigue are familial with shared
genetic factors explaining a substantial proportion of the comorbidity between the
traits in adults. Research focusing on the underlying pathways which are shared by
depression and fatigue will facilitate the elucidation of the mechanisms driving the
association.
125
In Chapter 6, a non-causal genetic relationship was implicated to explain the
association between depression and fatigue. This result was analogous with the
significant additive genetic correlation of 0.71 and bivariate heritability of 21%
identified between depression and fatigue. Importantly, the contribution of shared
genetic factors to the heritability of depression and fatigue was independent of their
overlapping presenting symptoms. Therefore, the relationship between depression
and fatigue is primarily due to shared genetic factors.
As detailed in Chapter 4, the first SNP loci robustly associated with MDD in
Europeans was identified in 2016. However, the underlying mechanisms associated
with fatigue remain unknown. Therefore, the following chapter evaluates the
previously implicated genes and risk loci from candidate gene association analyses
and genome-wide association analyses of CFS.
126 Chapter 7: Systematic Evaluation of Risk Loci from Candidate Gene and Genome-wide Association Studies of Fatigue
Chapter 7: Systematic Evaluation of Risk
Loci from Candidate Gene and
Genome-wide Association
Studies of Fatigue
7.1 ABSTRACT
Chronic fatigue syndrome (CFS) is a complex, neurological disorder of unknown
pathophysiology. However, a genetic contribution has been implicated by a limited
number of candidate gene association (CGA) and genome-wide association (GWA)
studies, the majority of which involved very small sample sizes. Replication studies
are required to confirm these exploratory findings. Therefore, the present study aims
to evaluate the findings from CGA and GWA studies. Additionally, the selected
genes will be evaluated in a more general fatigue phenotype. Thirty-nine genes with
nominal evidence for association were selected from published CGA studies while
524 SNPs and 319 genes were selected from published CFS GWA studies. Finally, 3
SNPs and 51 genes from a GWA and gene-based analysis of tiredness were selected.
Previously implicated SNPs and genes were investigated within GWA and gene-
based analyses, respectively. Initially, the analysis was conducted in a CFS cohort of
47 cases and 55 controls. The analysis was replicated in a fatigue cohort of 307 cases
and 744 controls. Bonferroni-corrected significance thresholds were calculated to
assess the association of SNPs with CFS and fatigue. Limited evidence was
identified supporting previously implicated SNPs or genes association with CFS or
fatigue. However, within the CFS cohort 3 SNPs and 3 genes of interest were
identified. Similarly, 6 genomic locations of interest which are likely associated with
fatigue were identified. Although encouraging, these results indicate GWA studies of
CFS and fatigue require larger sample sizes to identify robustly associated SNPs and
genes.
Chapter 7: Systematic Evaluation of Risk Loci from Candidate Gene and Genome-wide Association Studies of Fatigue 127
7.2 INTRODUCTION
Chronic fatigue syndrome (CFS) is a complex, neurological disorder of unknown
pathophysiology. Diagnosis occurs based on medical exclusion and requires
presentation with persistent or relapsing fatigue experienced over a six-month period
(Fukuda et al., 1994). Additionally, concurrent occurrence of at least four physical
symptoms (sore throat, tender lymph nodes, muscle pain, multi-joint pain, headaches,
unrefreshing sleep, cognitive difficulties, and post-exertional malaise), over a six-
month period which have not predated the fatigue are required. If insufficient
physical symptoms are present individuals can be diagnosed with idiopathic chronic
fatigue (ICF).
Although little is known about the underlying etiology of CFS previous studies
have indicated molecular genetics may contribute to the pathophysiology. The
additive heritability of CFS has been estimated at 51% in females, using classical
twin studies. Furthermore, the underlying genetic contribution of CFS has been
implicated through candidate gene association (CGA) studies and genome-wide
association (GWA) studies. However, to date, the association studies conducted have
comprised small sample sizes. Genes from the immune, nervous, and endocrine
systems have been implicated from CGA studies. Candidate genes which have
previously been associated with CFS from CGA studies include: BMP2K, CHRM1,
CHRM2, CHRM3, CHRM5, CHRNA10, CHRNA2, CHRNA3, CHRNA4, CHRNA5,
CHRNA9, CHRNB1, CHRNB4, CHRND, CHRNE, CHRNG, DISC1, EIF3A, FAM126B,
HTR2A, IFNG, IL-17F, IL6ST, METTL3, NR3C1, PEX16, SORL1, TCF3, TNF, TRPA1,
TRPC2, TRPC4, TRPC6, TRPM3, TRPM4, TRPM8, TRPV2, TRPV3, and UBTF (Carlo-
Stella et al., 2006; Fukuda et al., 2010; Marshall-Gradisnik et al., 2015a; Marshall-
Gradisnik et al., 2016a; Marshall-Gradisnik et al., 2016b; Marshall-Gradisnik et al.,
2015b; Metzger et al., 2008; Rajeevan et al., 2007; Shimosako & Kerr, 2014; Smith
et al., 2008). These genes were selected based on the current understanding and
hypotheses about the pathophysiology of CFS.
The first GWA study of CFS was conducted in 2011 and investigated 116,204
SNPs in 40 cases and 40 non-fatigued controls (Smith et al., 2011). No genome-wide
significant associations were identified, however, the authors suggested GRIK2 and
NPAS2 were candidate genes which warranted further investigation. The second
GWA study of CFS was conducted in 2015 and investigated 11,000 immune and
128 Chapter 7: Systematic Evaluation of Risk Loci from Candidate Gene and Genome-wide Association Studies of Fatigue
inflammation SNPs, in 50 CFS cases and 121 non-fatigued controls (Rajeevan et al.,
2015). No genome-wide significant associations were identified, however, the
authors identified 32 SNPs of potential functional importance. Finally, in February
2016, a GWA study of 42 CFS cases and 38 healthy controls investigated 659,094
SNPs (Schlauch et al., 2016). Ninety-two genome-wide significant SNPs were
reported using a Bonferroni adjusted threshold of 7.5 × 10-8 (of which 88 reached the
standard genome-wide significance threshold of 5 × 10-8). However, methodological
concerns within this study brings to question the validity of the reported results. In
particular, 288 of the 442 SNPs (407 autosomal SNPs with p < 3.3 × 10-5 and 35 X-
chromosome SNPs with p < 1.0 × 10-5) reported with suggestive associations did not
meet the minor allele frequency (MAF) > 5% or Hardy-Weinberg equilibrium
(HWE) χ2 p-value > 8 × 10-4 thresholds, reported by the authors, in cases, controls, or
the complete cohort separately (Supplementary Table 7.1). Furthermore, of the 92
reported genome-wide significant SNPs only 18 meet the described MAF and HWE
thresholds (rs254577, p = 2.35 × 10-11; rs2200706, p = 5.48 × 10-10; rs17255510, p =
6.61 × 10-10; rs6892217, p = 6.61 × 10-10; rs16826918, p = 1.13 × 10-9; rs5974598, p
= 1.55 × 10-9; rs689462, p = 2.08 × 10-9; rs6675622, p = 5.94 × 10-9; rs12391243, p =
6.68 × 10-9; rs4022211, p = 9.09 × 10-9; rs16883408, p = 1.06 × 10-8; rs16902672, p
= 1.77 × 10-8; rs4473594, p = 1.81 × 10-8; rs10737169, p = 2.51 × 10-8; rs2748997, p
= 2.76 × 10-8; rs2882361, p = 3.02 × 10-8; rs17133553, p = 4.74 × 10-8; and
rs2816936, p = 4.91 × 10-8).
In 2017, the most powerful genetic analysis of a fatigue phenotype was
conducted in the UK Biobank sample (Deary et al., 2017). Self-reported tiredness
was analysed as a quantitative trait with four levels of classification based on
individuals response to the question “Over the last two weeks, how often have you
felt tired or had little energy?”; 6,948 individuals responded “nearly every day”,
6,404 individuals responded “more than half the days”, 44,208 individuals responded
“several days”, and 51,416 individuals responded “not at all”. The investigation
identified one SNP reaching genome-wide significance (Affymetrix ID
1:64178756_C_T, p = 1.36 × 10-11) and two suggestive peaks on chromosome 1 and
17 (top SNPs within each peak: rs142592148, p = 5.88 × 10-8 and rs2555592, p =
6.86 × 10-8). Within the study a gene-based analysis was also conducted which
identified five genome-wide significant (p < 2.768 × 10-6) genes (DRD2, p = 2.94 ×
Chapter 7: Systematic Evaluation of Risk Loci from Candidate Gene and Genome-wide Association Studies of Fatigue 129
10-7; PRRC2C, p = 1.43 × 10-6; C3orf84, p = 1.45 × 10-6; ANO10, p = 1.52 × 10-6;
and ASXL3, p = 2.67 × 10-6) (Deary et al., 2017).
Although only a few studies with limited power have attempted to identify
genetic risk loci associated with CFS, immense effort has gone into the elucidation of
the molecular genetics of depression. Depression and fatigue commonly co-occur,
with 10.7-14.5% of CFS patients diagnosed with major depressive disorder (MDD)
(Cella et al., 2013; Janssens et al., 2015). Furthermore, twin studies indicate shared
genetic factors likely contribute to the high levels of comorbidity observed between
varying levels of fatigue and depression (Burri et al., 2015; Corfield et al., 2016b;
Fowler et al., 2006; Hickie et al., 1999b; Hur et al., 2012; Kato et al., 2009; Narusyte
et al., 2016). In 2016 and 2017, 17 SNPs within 15 independent genomic locations
which are robustly associated with major depression (Hyde et al., 2016), one SNP
was associated with MDD and age of onset ≥ 27 years (Power et al., 2017), two
SNPs were associated with depressive symptoms (Okbay et al., 2016), and one SNP
was associated with a broad depression phenotype (MDD and depressive symptoms)
(Direk et al., 2016), in Europeans. However, to date nothing is known about the
molecular genetics underlying the comorbidity between fatigue and depression.
Considering the limited power and methodological concerns associated with
previous CFS CGA and GWA analyses, replication studies are required to confirm
the exploratory findings. Therefore, the present study aims to evaluate the findings
from CGA and GWA studies of fatigue phenotypes. Additionally, the association
signal of robustly associated depression risk loci will be investigated to determine if
they contribute to fatigue phenotypes. These results will be evaluated in a CFS and
more general fatigue phenotype.
7.3 METHODS
7.3.1 Previously Implicated Genes
A literature search was conducted to identify CGA and GWA studies were the
primary phenotype investigated was fatigue. Only autosomal SNP markers and genes
were investigated within the present study. Thirty-nine genes (and 151 SNPs within
these genes) with nominal evidence (p < 0.05) for an allelic association were selected
from published CFS CGA studies (Table 7.1) (Carlo-Stella et al., 2006; Fukuda et
al., 2010; Marshall-Gradisnik et al., 2015a; Marshall-Gradisnik et al., 2016a;
130 Chapter 7: Systematic Evaluation of Risk Loci from Candidate Gene and Genome-wide Association Studies of Fatigue
Marshall-Gradisnik et al., 2016b; Marshall-Gradisnik et al., 2015b; Metzger et al.,
2008; Rajeevan et al., 2007; Shimosako & Kerr, 2014; Smith et al., 2008;
Sommerfeldt et al., 2011). While 524 SNPs (and their 319 assigned genes) were
selected from published CFS GWA studies (Table 7.2) (Rajeevan et al., 2015;
Schlauch et al., 2016; Smith et al., 2011). Additionally, one SNP reaching genome-
wide significance and the two top SNPs (and the two associated genes) from
suggestive peaks in a GWA study of self-reported tiredness, within the UK biobank
sample were investigated (Table 7.3) (Deary et al., 2017). Finally, five genes
reaching genome-wide significance (p < 2.768 × 10-6) and 44 genes suggestively
associated (p < 1 × 10-4) with self-reported tiredness in a gene-based association
analysis, within the UK biobank sample were investigated (Table 7.4) (Deary et al.,
2017).
In addition, a literature search was conducted to identify SNPs (and their
assigned genes) which have been robustly associated with a depression phenotype
from GWA analyses, in Europeans. Twenty-one SNPs (and their eight assigned
genes) were selected for investigation within the present study (Table 7.5).
Chapter 7: Systematic Evaluation of Risk Loci from Candidate Gene and Genome-wide Association Studies of Fatigue 131
Table 7.1. Summary of genes from candidate gene association studies for fatigue traits.
Gene SNP Chr SNP position RA OA Frequency of RA
OR (95% CI) Allelic p-value Powerb
Case Control 1000 Genomesa
Study by Carlo-Stella and colleagues (2006) in an Italian cohort of 47 CFS cases and 104 controls.
IFNG rs2430561 12 68552522 T A 0.550 0.630 0.538 1.43 (0.89-2.28) 0.0440 0.9771
Study by Carlo-Stella and colleagues (2006) in an Italian cohort of 80 CFS cases and 224 controls.
TNF rs1799724 6 31542482 T C 0.300 0.190 0.094 1.80 (1.20-2.72) 0.0040 0.9816
Study by Rajeevan and colleagues (2007) in a European cohort of 40 CFS and 55 ICF cases and 42 controls.
NR3C1 rs6188 5 142680344 C A 0.700 0.570 0.682 1.76 (0.83-3.73) 0.0383 0.9999 rs852977 5 142687494 A G 0.700 0.570 0.680 1.76 (0.83-3.73) 0.0365 0.9999
rs860458 5 142696036 G A 0.850 0.730 0.855 2.10 (0.87-5.07) 0.0180 0.9997 rs1866388 5 142759785 A G 0.710 0.570 0.681 1.85 (0.87-3.93) 0.0335 1.0000
rs2918419 5 142722353 T C 0.850 0.730 0.854 2.10 (0.87-5.07) 0.0164 0.9997
Study by Smith and colleagues (2008) in a European cohort of 40 CFS cases and 42 controls.
HTR2A rs6311 13 47471478 A G 0.487 0.274 0.437 2.52 (1.00-6.30) 0.0065 1.0000
rs6313 13 47469940 T C 0.475 0.281 0.436 2.32 (0.93-5.78) 0.0150 1.0000
rs1923884 13 47421836 C T 0.787 0.595 0.874 2.51 (0.95-6.67) 0.0100 1.0000 Study by Metzger and colleagues (2008) in a European cohort of 89 CFS cases and 56 controls.
IL-17F rs763780 6 52101739 T C 0.955 0.839 0.942 4.07 (1.70-9.71) 0.0018 0.9997
Study by Fukuda and colleagues (2010) in a Japanese cohort of 108 CFS cases and 68 controls.
DISC1 rs821616 1 232144598 T A 0.140 0.100 0.287 (0.076)c 1.50 (1.02-2.19) 0.0370 0.9885
Study by Shimosako and Kerr (2014) in a UK cohort of 108 CFS cases and 68 controls.
BMP2K rs1426139 4 79766677 A T 0.056 0.051 0.046 1.10 (0.28-4.29) 0.0091 0.0723 rs3775516 4 79744066 G A 0.951 0.946 0.955 1.11 (0.28-4.35) 0.0025 0.0768
EIF3A rs10787901 10 120819453 A G 0.561 0.511 0.189 1.22 (0.67-2.25) < 0.0001 0.4032
FAM126B rs11895568 2 201847877 G A 0.009 0.000 0.018 NC 0.0110 NC IL6ST rs1373998 5 55255565 T C 0.12 0.081 0.133 1.55 (0.54-4.41) 0.0130 0.9282
METTL3 rs3752411 14 21968876 A G 0.131 0.059 0.147 2.40 (0.76-7.62) 0.0310 1.0000
PEX16 rs3802758 11 45936035 C T 0.319 0.118 0.085 3.50 (1.51-8.11) < 0.0001 1.0000 SORL1 rs3737529 11 121477816 T C 0.052 0.007 0.025 7.78 (0.40-152.39) 0.0280 0.9737
TCF3 rs1860661 19 1650134 G A 0.225 0.045 0.415 6.16 (1.80-21.13) < 0.0001 1.0000
UBTF rs2071167 17 42287519 A G 0.319 0.199 0.260 1.89 (0.92-2.25) 0.0240 1.0000 Study by Marshall-Gradisnik and colleagues (2015b) in an Australian cohort of 115 CFS cases and 90 controls.
TRPA1 rs2383844 8 72961252 G A 0.505 0.398 0.427 1.54 (1.04-2.29) 0.0400 0.9981
rs4738202 8 72940861 A G 0.369 0.253 0.311 1.73 (1.12-2.65) 0.0180 0.9999 TRPC4 rs655207 13 38368012 G T 0.505 0.381 0.39 1.66 (1.11-2.46) 0.0180 0.9999
rs6650469 13 38367949 T C 0.505 0.380 0.607 1.66 (1.12-2.48) 0.0160 0.9998
TRPM3 rs1160742 9 73314011 A G 0.470 0.333 0.442 1.78 (1.19-2.66) 0.0080 1.0000 rs1328153 9 73416062 C T 0.240 0.137 0.207 1.99 (1.18-3.35) 0.0130 1.0000
rs1504401 9 73916953 C T 0.900 0.827 0.905 1.88 (1.06-3.36) 0.0410 0.9645
rs3763619 9 73225802 A C 0.440 0.316 0.422 1.70 (1.13-2.56) 0.0140 1.0000
Table 7.1 footnotes on page 135.
132 Chapter 7: Systematic Evaluation of Risk Loci from Candidate Gene and Genome-wide Association Studies of Fatigue
Table 7.1. Continued Summary of genes from candidate gene association studies for fatigue traits.
Gene SNP Chr SNP position RA OA Frequency of RA
OR (95% CI) Allelic p-value Powerb Case Control 1000 Genomesa
rs4454352 9 73410410 C T 0.240 0.137 0.207 1.99 (1.18-3.35) 0.0130 1.0000
rs7865858 9 73204431 A G 0.450 0.331 0.445 1.65 (1.10-2.48) 0.0210 0.9998
rs10115622 9 73306551 C A 0.665 0.565 0.661 1.53 (1.02-2.29) 0.0500 0.9903
rs11142508 9 73231662 C T 0.445 0.298 0.423 1.89 (1.25-2.85) 0.0040 1.0000
rs12682832 9 73220691 A G 0.444 0.293 0.424 1.93 (1.27-2.91) 0.0030 1.0000
Study by Marshall-Gradisnik and colleagues (2015a) in an Australian cohort of 115 CFS cases and 90 controls.
CHRM3 rs589962 1 239989964 T C 0.758 0.608 0.694 2.02 (1.32-3.09) 0.0035 1.0000
rs726169 1 239794277 A G 0.717 0.599 0.669 1.70 (1.12-2.56) 0.0235 0.9998 rs1072320 1 239982376 G A 0.324 0.184 0.272 2.12 (1.33-3.39) 0.0037 1.0000
rs4463655 1 239984294 C T 0.692 0.533 0.570 1.97 (1.32-2.96) 0.0028 1.0000
rs6429157 1 239981643 G A 0.522 0.408 0.457 1.59 (1.07-2.35) 0.0375 0.9991 rs6661621 1 239984803 C G 0.302 0.171 0.255 2.10 (1.30-3.39) 0.0054 1.0000
rs6669810 1 240068629 C G 0.579 0.453 0.521 1.66 (1.12-2.45) 0.0236 0.9999
rs7520974 1 240067260 A G 0.580 0.447 0.522 1.71 (1.15-2.53) 0.0167 1.0000 rs7543259 1 239979186 A G 0.319 0.184 0.269 2.07 (1.30-3.31) 0.0051 1.0000
CHRNA10 rs2672211 11 3690278 C T 0.374 0.243 0.339 1.85 (1.20-2.86) 0.0107 1.0000
rs2672214 11 3691512 C T 0.371 0.240 0.339 1.87 (1.21-2.88) 0.0108 0.9999
rs2741862 11 3687985 C T 0.286 0.184 0.257 1.77 (1.10-2.84) 0.0304 0.9999
rs2741868 11 3690183 T A 0.369 0.240 0.661 1.85 (1.20-2.86) 0.0119 1.0000
rs2741870 11 3690109 G C 0.371 0.243 0.661 1.83 (1.19-2.82) 0.0128 1.0000 CHRNA2 rs2565048 8 27330132 T C 0.901 0.807 0.693 2.18 (1.24-3.86) 0.0140 1.0000
CHRNA5 rs951266 15 78878541 T C 0.394 0.263 0.366 1.82 (1.19-2.79) 0.0115 1.0000
rs7180002 15 78873993 T A 0.385 0.276 0.366 1.64 (1.07-2.49) 0.0368 0.9998 Study by Marshall-Gradisnik and colleagues (2016a) in an Australian cohort of 39 CFS cases and 30 controls.
CHRM1 rs2075748 11 62688269 A G 0.244 0.103 0.230 2.79 (1.05-7.43) 0.0369 1.0000
rs11823728 11 62676802 C T 0.947 0.839 0.970 3.45 (1.03-11.55) 0.0394 0.9471 CHRM3 rs4620530 1 240063821 T G 0.474 0.300 0.418 2.11 (1.04-4.28) 0.0381 1.0000
CHRNA2 rs891398 8 27324822 C T 0.553 0.345 0.512 2.35 (1.17-4.70) 0.0168 1.0000
rs2741343 8 27326127 C T 0.553 0.350 0.514 2.29 (1.15-4.59) 0.0186 1.0000 CHRNA3 rs2869546 15 78907345 T C 0.724 0.533 0.650 2.29 (1.13-4.66) 0.0217 1.0000
rs3743074 15 78909480 T C 0.718 0.550 0.657 2.08 (1.03-4.23) 0.0410 1.0000
rs3743075 15 78909452 G A 0.718 0.550 0.659 2.08 (1.03-4.23) 0.0410 1.0000 rs4243084 15 78911672 G C 0.436 0.267 0.625 2.12 (1.03-4.39) 0.0403 1.0000
rs12914385 15 78898723 T C 0.487 0.283 0.405 2.40 (1.17-4.92) 0.0153 1.0000
CHRNA5 rs951266 15 78878541 T C 0.436 0.259 0.366 2.22 (1.07-4.60) 0.0332 1.0000 rs7180002 15 78873993 T A 0.434 0.267 0.366 2.11 (1.02-4.36) 0.0433 1.0000
CHRNB4 rs12441088 15 78928264 T G 0.821 0.621 0.745 2.79 (1.28-6.09) 0.0090 1.0000
CHRNE rs33970119 17 4804902 G A 0.962 0.867 0.938 3.85 (0.97-15.18) 0.0414 0.9997
Table 7.1 footnotes on page 135.
Chapter 7: Systematic Evaluation of Risk Loci from Candidate Gene and Genome-wide Association Studies of Fatigue 133
Table 7.1. Continued Summary of genes from candidate gene association studies for fatigue traits.
Gene SNP Chr SNP position RA OA Frequency of RA
OR (95% CI) Allelic p-value Powerb Case Control 1000 Genomesa
TRPC2 rs6578398 11 3638061 A G 0.346 0.183 0.285 2.36 (1.06-5.27) 0.0338 1.0000
rs7108612 11 3650086 T G 0.192 0.067 0.118 3.33 (1.04-10.63) 0.0337 1.0000
TRPC4 rs2985167 13 38230542 A G 0.718 0.500 0.610 2.54 (1.26-5.16) 0.0094 1.0000
TRPM3 rs1106948 9 74017174 T C 0.603 0.383 0.555 2.44 (1.22-4.87) 0.0107 1.0000
rs1891301 9 74018496 T C 0.577 0.383 0.536 2.19 (1.10-4.36) 0.0241 1.0000
rs6560200 9 73980222 C T 0.603 0.379 0.534 2.48 (1.24-4.95) 0.0100 1.0000 rs11142822 9 74042243 G T 0.962 0.850 0.917 4.41 (1.14-17.09) 0.0212 1.0000
rs12350232 9 74032148 T G 0.603 0.400 0.552 2.27 (1.14-4.52) 0.0183 1.0000 TRPM8 rs6758653 2 234912799 G A 0.756 0.550 0.643 2.54 (1.23-5.25) 0.0108 1.0000
rs11563204 2 234917377 A G 0.355 0.117 0.206 4.17 (1.67-10.41) 0.0014 1.0000
rs17865678 2 234919314 A G 0.460 0.167 0.279 4.25 (1.89-9.57) 0.0003 1.0000 Study by Marshall-Gradisnik and colleagues (2016b) in an Australian cohort of 11 CFS cases and 11 controls.
CHRM2 rs1424569 7 136569416 A G 0.600 0.333 0.491 3.00 (0.88-10.27) 0.0300 1.0000
CHRM3 rs1134 1 239872172 C T 0.654 0.375 0.576 3.15 (0.92-10.78) 0.0200 1.0000 rs576386 1 239995289 C G 0.519 0.250 0.384 3.24 (0.90-11.62) 0.0400 1.0000
rs619214 1 239958622 T G 0.700 0.389 0.536 3.67 (1.05-12.82) 0.0300 1.0000
rs685550 1 239924408 C T 0.222 0.0417 0.243 6.57 (0.65-66.86) 0.0500 1.0000
rs1019882 1 239898856 A G 0.611 0.375 0.577 2.62 (0.78-8.84) 0.0500 1.0000
rs1155611 1 239897827 C T 0.611 0.375 0.576 2.62 (0.78-8.84) 0.0500 1.0000
rs1155612 1 239897705 A G 0.600 0.333 0.486 3.00 (0.88-10.27) 0.0300 1.0000 rs1416789 1 239901645 A G 0.611 0.375 0.549 2.62 (0.78-8.84) 0.0500 1.0000
rs1544170 1 239908236 G A 0.630 0.375 0.545 2.83 (0.83-9.62) 0.0400 1.0000
rs1867263 1 239807920 G A 0.696 0.458 0.611 2.71 (0.79-9.34) 0.0400 1.0000 rs1867264 1 239845277 T A 0.720 0.364 0.390 4.50 (1.26-16.08) 0.0000 1.0000
rs1867265 1 239840107 G A 0.679 0.417 0.609 2.96 (0.86-10.14) 0.0300 1.0000
rs1899616 1 239818568 G A 0.673 0.333 0.576 4.12 (1.17-14.47) 0.0100 1.0000 rs2083817 1 239833605 T A 0.685 0.417 0.408 3.05 (0.89-10.49) 0.0300 1.0000
rs2163546 1 240057960 G A 0.539 0.273 0.519 3.11 (0.88-10.95) 0.0400 1.0000
rs2165872 1 239826988 C T 0.685 0.417 0.593 3.05 (0.89-10.49) 0.0300 1.0000 rs3738436 1 239872493 C A 0.611 0.375 0.576 2.62 (0.78-8.84) 0.0500 1.0000
rs6429147 1 239794794 G C 0.704 0.458 0.597 2.81 (0.81-9.71) 0.0400 1.0000
rs6684622 1 239877537 G C 0.620 0.318 0.544 3.50 (1.01-12.12) 0.0200 1.0000 rs6688537 1 239850588 C A 0.593 0.333 0.494 2.91 (0.85-9.94) 0.0300 1.0000
rs6694220 1 239883616 A G 0.577 0.333 0.483 2.73 (0.80-9.29) 0.0500 1.0000
rs6700643 1 239798921 T C 0.704 0.458 0.583 2.81 (0.81-9.71) 0.0400 1.0000 rs7511970 1 239883255 G A 0.611 0.375 0.576 2.62 (0.78-8.84) 0.0500 1.0000
rs7513746 1 239862411 A G 0.611 0.375 0.576 2.62 (0.78-8.84) 0.0500 1.0000
rs7551001 1 239844600 A G 0.679 0.417 0.580 2.96 (0.86-10.14) 0.0300 1.0000
Table 7.1 footnotes on page 135.
134 Chapter 7: Systematic Evaluation of Risk Loci from Candidate Gene and Genome-wide Association Studies of Fatigue
Table 7.1. Continued Summary of genes from candidate gene association studies for fatigue traits.
Gene SNP Chr SNP position RA OA Frequency of RA
OR (95% CI) Allelic p-value Powerb Case Control 1000 Genomesa
rs10754677 1 239833100 A G 0.615 0.375 0.498 2.67 (0.79-9.01) 0.0500 1.0000
rs10802795 1 239870775 T C 0.611 0.375 0.546 2.62 (0.78-8.84) 0.0500 1.0000
rs10802802 1 239909942 G A 0.625 0.333 0.577 3.33 (0.97-11.49) 0.0200 1.0000
rs10925941 1 239812538 G A 0.704 0.458 0.582 2.81 (0.81-9.71) 0.0400 1.0000
rs10925964 1 239902514 T A 0.611 0.375 0.574 2.62 (0.78-8.84) 0.0500 1.0000
rs11585281 1 239909651 C T 0.611 0.333 0.577 3.14 (0.92-10.79) 0.0200 1.0000 rs12029701 1 239910601 T C 0.611 0.333 0.576 3.14 (0.92-10.79) 0.0200 1.0000
rs12093821 1 239824248 G A 0.704 0.417 0.564 3.32 (0.96-11.57) 0.0200 1.0000 rs12743042 1 239888304 T C 0.630 0.364 0.577 2.98 (0.87-10.14) 0.0300 1.0000
rs16838637 1 239828350 A G 0.679 0.417 0.563 2.96 (0.86-10.14) 0.0300 1.0000
CHRM5 rs511422 15 34282982 C T 0.446 0.208 0.398 3.06 (0.81-11.57) 0.0400 1.0000 rs603152 15 34294637 A C 0.464 0.208 0.410 3.29 (0.87-12.42) 0.0300 1.0000
rs646950 15 34291660 T C 0.462 0.208 0.409 3.26 (0.86-12.28) 0.0300 1.0000
CHRNA2 rs2741341 8 27330286 C T 0.518 0.208 0.427 4.08 (1.08-15.38) 0.0100 1.0000 CHRNA4 rs11698563 20 61992285 C A 0.712 0.292 0.33 5.99 (1.63-22.02) 0.0000 1.0000
CHRNA9 rs4861065 4 40344395 C T 0.393 0.125 0.276 4.53 (0.98-20.84) 0.0200 1.0000
rs4861323 4 40355815 A G 0.821 0.542 0.777 3.89 (0.98-15.41) 0.0100 1.0000
rs7669882 4 40350651 A G 0.393 0.125 0.217 4.53 (0.98-20.84) 0.0200 1.0000
rs10009228 4 40356422 G A 0.821 0.500 0.775 4.60 (1.16-18.18) 0.0000 1.0000
rs10015231 4 40337566 C T 0.804 0.583 0.754 2.92 (0.76-11.28) 0.0400 1.0000 CHRNB1 rs2302767 17 7350544 T C 0.722 0.500 0.690 2.6 (0.74-9.10) 0.0500 1.0000
rs3829603 17 7347042 C A 0.741 0.500 0.694 2.86 (0.80-10.15) 0.0500 1.0000
rs4151134 17 7347123 T C 0.679 0.375 0.535 3.52 (1.02-12.20) 0.0100 1.0000 CHRNB4 rs12440298 15 78927589 T G 0.982 0.833 0.984 11.00 (0.39-313.06) 0.0100 0.9305
CHRND rs2767 2 233400074 T C 0.643 0.333 0.646 3.60 (1.04-12.49) 0.0100 1.0000
rs2853457 2 233397968 A G 0.500 0.250 0.423 3.00 (0.84-10.75) 0.0400 1.0000 rs3762529 2 233392449 T C 0.654 0.375 0.641 3.15 (0.92-10.78) 0.0200 1.0000
rs3791729 2 233395297 C T 0.643 0.375 0.665 3.00 (0.88-10.24) 0.0300 1.0000
rs3828246 2 233398215 C T 0.827 0.583 0.749 3.41 (0.85-13.73) 0.0200 1.0000 rs4973537 2 233391965 A G 0.643 0.375 0.642 3.00 (0.88-10.24) 0.0300 1.0000
rs11674608 2 233404294 C G 0.660 0.222 0.646 6.79 (1.78-25.88) 0.0000 1.0000
rs12463989 2 233395029 T C 0.643 0.333 0.645 3.60 (1.04-12.49) 0.0100 1.0000 rs12466358 2 233397525 T G 0.827 0.583 0.748 3.41 (0.85-13.73) 0.0200 1.0000
rs13026409 2 233402507 C T 0.821 0.583 0.748 3.29 (0.83-13.08) 0.0200 1.0000
rs67583510 2 233405650 G A 0.815 0.546 0.747 3.67 (0.94-14.34) 0.0200 1.0000 rs112001880 2 233403760 I D 0.643 0.333 0.646 3.60 (1.04-12.49) 0.0100 1.0000
CHRNE rs33970119 17 4804902 G A 0.964 0.833 0.938 5.40 (0.44-66.84) 0.0400 1.0000
CHRNG rs13018423 2 233408283 C T 0.821 0.583 0.736 3.29 (0.83-13.08) 0.0200 1.0000
Table 7.1 footnotes on page 135.
Chapter 7: Systematic Evaluation of Risk Loci from Candidate Gene and Genome-wide Association Studies of Fatigue 135
Table 7.1. Continued Summary of genes from candidate gene association studies for fatigue traits.
Gene SNP Chr SNP position RA OA Frequency of RA
OR (95% CI) Allelic p-value Powerb Case Control 1000 Genomesa
TRPC6 rs11224816 11 101396286 T C 0.519 0.208 0.490 4.10 (1.09-15.46) 0.0100 1.0000
TRPM3 rs1317103 9 73195703 C T 0.352 0.083 0.233 5.97 (1.04-34.27) 0.0100 1.0000
rs3812532 9 73483594 C A 0.630 0.375 0.570 2.83 (0.83-9.62) 0.0400 1.0000
rs4620343 9 73736643 T C 0.411 0.167 0.480 3.48 (0.85-14.22) 0.0300 1.0000
rs10780950 9 73193428 T C 0.289 0.083 0.173 4.46 (0.76-26.21) 0.0500 1.0000
TRPM4 rs11083963 19 49665340 A G 0.717 0.458 0.555 3.00 (0.86-10.48) 0.0300 1.0000 TRPV2 rs3514 17 4801594 G C 0.926 0.750 0.162 4.17 (0.65-26.90) 0.0300 1.0000
rs2075763 17 4802685 C T 0.963 0.833 0.942 5.20 (0.44-62.13) 0.0500 0.9999 rs7222754 17 16329745 T C 0.442 0.208 0.385 3.01 (0.80-11.39) 0.0500 1.0000
rs12602006 17 16337288 A G 0.731 0.500 0.650 2.71 (0.77-9.56) 0.0500 1.0000
rs12942540 17 4804073 G C 0.926 0.750 0.852 4.17 (0.65-26.90) 0.0300 1.0000 rs35400274 17 4803711 G A 0.929 0.750 0.849 4.33 (0.66-28.62) 0.0300 1.0000
TRPV3 rs4790519 17 3456735 C T 0.556 0.250 0.440 3.75 (1.04-13.49) 0.0100 1.0000
Chr: Chromosome; RA: risk allele; OA: other allele; OR: odds ratio; CI: confidence interval. a Prevalence shown is for the 1000 Genomes European population. bPower to detect allelic difference at an alpha level of 0.05 in the fatigue cohort. cPrevalence shown in parentheses is for the 1000 Genomes Japanese population.
136 Chapter 7: Systematic Evaluation of Risk Loci from Candidate Gene and Genome-wide Association Studies of Fatigue
Table 7.2. Summary of SNPs from genome-wide association studies for chronic fatigue syndrome.
SNP Chr SNP position RA Freq OR (95% CI) p-value Genea
Study by Smith and colleagues (2011) in an American population of 40 cases and 40 controls.
rs2105239 1 14857643 A 0.410 3.40 (1.73-6.66) 3.00 × 10-4 -
rs10489599 1 16585818 C 0.117 4.69 (1.85-11.85) 6.00 × 10-4 FBXO42
rs6694861 1 114063592 G 0.526 3.34 (1.66-6.72) 8.00 × 10-4 MAGI3
rs1157819 1 209604734 A 0.181 3.65 (1.72-7.78) 6.00 × 10-4 MIR205
rs6721414 2 18494495 C 0.132 4.10 (1.77-9.46) 8.00 × 10-4 DA335567
rs6710681 2 64413221 C 0.295 3.41 (1.76-6.59) 2.00 × 10-4 BF513477
rs356653 2 101539790 C 0.628 4.02 (1.80-9.02) 7.00 × 10-4 NPAS2
rs10496982 2 146308114 A 0.825 15.91 (2.04-124.27) 3.00 × 10-4 - rs2715898 2 201556388 A 0.444 3.75 (1.87-7.53) 2.00 × 10-4 DA324672
rs10510985 3 69663871 G 0.321 3.79 (1.95-7.35) 2.00 × 10-4 - rs4894505 3 175920884 G 0.150 3.78 (1.77-8.07) 8.00 × 10-4 BG218013
rs1377828 3 176245042 C 0.163 3.78 (1.79-7.96) 6.00 × 10-4 DA709814
rs10516187 4 7778524 G 0.775 7.26 (2.04-25.79) 6.00 × 10-4 AFAP1
rs2389957 4 120695322 G 0.463 2.96 (1.53-5.73) 5.00 × 10-4 -
rs33013 5 80060016 G 0.563 3.37 (1.65-6.89) 6.00 × 10-4 MSH3
rs10514211 5 80482514 G 0.000 NC 6.00 × 10-4 RASGRF2
rs3797302 5 145889123 G 0.771 9.78 (2.15-44.41) 9.00 × 10-4 TCERG1
rs4714468 6 41452996 G 0.756 6.12 (1.98-18.95) 7.00 × 10-4 DQ141194
rs6915865 6 91082281 C 0.350 3.05 (1.58-5.89) 7.00 × 10-4 - rs10498968 6 91083880 G 0.128 4.25 (1.90-9.51) 5.00 × 10-4 -
rs9320409 6 97530846 T 0.363 2.97 (1.56-5.68) 8.00 × 10-4 KLHL32
rs2047179 6 98878812 T 0.554 3.78 (1.78-8.03) 3.00 × 10-4 - rs2247215 6 101966454 A 0.500 3.21 (1.63-6.31) 5.00 × 10-4 GRIK2
rs2247218 6 101966553 T 0.473 4.23 (2.04-8.78) 1.00 × 10-4 GRIK2
rs4245562 7 54403985 C 0.162 3.63 (1.69-7.77) 9.00 × 10-4 BX111274
rs10499740 7 54530581 C 0.474 3.32 (1.66-6.66) 9.00 × 10-4 -
rs723886 7 68159592 G 0.500 3.21 (1.63-6.31) 1.00 × 10-3 -
rs3801293 7 96324499 C 0.788 9.44 (2.10-42.51) 8.00 × 10-4 SHFM1
rs1499646 8 75068112 C 0.090 4.35 (1.75-10.82) 9.00 × 10-4 -
rs543736 8 104012949 A 0.141 3.71 (1.68-8.19) 7.00 × 10-4 LOC100131813
rs4236780 8 107936951 G 0.075 4.68 (1.78-12.29) 6.00 × 10-4 - rs871024 9 21803880 A 0.311 3.22 (1.64-6.3) 6.00 × 10-4 MTAP
rs4978076 9 26524684 C 0.351 3.30 (1.70-6.41) 6.00 × 10-4 -
rs10511961 9 71497485 C 0.289 3.38 (1.72-6.62) 7.00 × 10-4 PIP5K1B
rs10509412 10 89599354 G 0.237 3.40 (1.70-6.80) 3.00 × 10-4 CFLP1
rs1325904 10 90280938 T 0.118 3.79 (1.64-8.75) 8.00 × 10-4 C10orf59
rs726817 10 95459817 T 0.618 3.50 (1.62-7.54) 6.00 × 10-4 FRA10AC1
rs10509958 10 114054601 A 0.516 3.56 (1.67-7.59) 4.00 × 10-4 TECTB
rs734640 11 17613348 C 0.114 4.99 (2.09-11.94) 4.00 × 10-4 OTOG
rs10500964 11 23596570 T 0.013 17.97 (2.30-140.50) 1.00 × 10-4 - rs10500965 11 23596625 T 0.013 20.39 (2.63-157.97) 1.00 × 10-4 -
rs10501068 11 26769636 G 0.203 3.74 (1.83-7.66) 3.00 × 10-4 CN274762
rs10501376 11 58971766 C 0.859 NC 5.00 × 10-4 DTX4
rs10488767 11 110458835 A 0.229 3.38 (1.64-6.96) 9.00 × 10-4 ARHGAP20
rs1881470 11 127333840 G 0.152 3.39 (1.34-8.61) 5.00 × 10-4 -
rs10505778 12 14125564 A 0.410 3.79 (1.95-7.38) 2.00 × 10-4 GRIN2B
rs10506025 12 27726370 G 0.244 3.64 (1.83-7.22) 1.00 × 10-4 PPFIBP1
rs4931109 12 29005992 T 0.050 5.92 (1.91-18.31) 9.00 × 10-4 -
rs167337 12 52182052 T 0.763 5.90 (1.89-18.36) 9.00 × 10-4 SCN8A
rs1144418 12 65293514 C 0.632 3.97 (1.76-8.93) 4.00 × 10-4 FLJ41278
rs7994531 13 42977439 C 0.600 7.11 (2.71-18.64) 1.00 × 10-4 BG220650
rs10507556 13 47970075 A 0.013 14.44 (1.83-114.05) 1.00 × 10-3 - rs1931035 13 79274470 G 0.724 7.25 (2.36-22.33) 1.00 × 10-4 -
rs1359536 13 79275793 T 0.782 10.87 (2.42-48.86) 5.00 × 10-4 -
rs547571 13 97231270 G 0.775 7.45 (2.10-26.46) 2.00 × 10-4 HS6ST3
rs1555589 13 100480664 A 0.613 4.27 (1.79-10.17) 4.00 × 10-4 CLYBL
rs7325773 13 104182490 C 0.663 4.46 (1.88-10.60) 6.00 × 10-4 CA425896
rs3759688 14 60975579 T 0.00 NC 8.00 × 10-4 SIX6
rs2372200 14 83029545 A 0.649 4.88 (1.95-12.17) 4.00 × 10-4 -
rs6503623 17 39503659 G 0.013 18.33 (2.36-142.62) 4.00 × 10-4 KRTHA3A
rs400322 19 55172578 G 0.689 6.04 (2.15-16.99) 1.00 × 10-4 LILRB4
rs2059152 19 56318188 G 0.769 5.70 (1.83-17.74) 1.00 × 10-3 NLRP11
rs10500321 19 56319571 T 0.769 5.70 (1.83-17.74) 8.00 × 10-4 NLRP11
rs382958 19 56439438 T 0.486 3.67 (1.80-7.51) 2.00 × 10-4 NLRP13
rs1399592 21 39053862 T 0.351 3.21 (1.64-6.29) 4.00 × 10-4 KCNJ6
Study by Rajeevan and colleagues (2015) in an American cohort of 50 cases and 121 controls.
rs829370 1 21933193 C 0.880 4.00 (0.77-20.72) 1.80 × 10-3 RAP1GAP
rs2235937 1 29631909 A 0.160 1.73 (0.77-3.87) NS PTPRU
rs10800118 1 165599774 C 0.400 1.55 (0.80-3.01) NS MGST3
rs17591814 1 186846598 C 0.300 1.91 (0.97-3.77) 1.00 × 10-2 PLA2G4A
Footnotes for Table 7.2 on page 143.
Chapter 7: Systematic Evaluation of Risk Loci from Candidate Gene and Genome-wide Association Studies of Fatigue 137
Table 7.2. Continued Summary of SNPs from genome-wide association studies for chronic fatigue
syndrome.
SNP Chr SNP position RA Freq OR (95% CI) p-value Genea
rs1061147b 1 196654324 C 0.280 2.00 (1.01-3.98) 6.50 × 10-3 CFH
rs7529589 1 196658279 C 0.280 2.07 (1.04-4.10) 4.30 × 10-3 CFH
rs3020729 2 87012293 T 0.040 4.47 (1.36-14.64) 2.70 × 10-3 CD8A
rs13010656 2 203297068 T 0.440 1.93 (0.99-3.78) 5.70 × 10-3 BMPR2
rs1048829b 2 203430456 T 0.440 1.93 (0.99-3.78) 5.70 × 10-3 BMPR2
rs7616342 3 19433647 G 0.470 2.09 (1.05-4.12) 2.20 × 10-3 KCNH8
rs2228428 3 32995928 T 0.570 2.01 (0.98-4.12) 4.50 × 10-3 CCR4
rs3774268 3 186954324 A 0.771 2.48 (0.92-6.68) 3.90 × 10-3 MASP1
rs1801058 4 3039150 C 0.276 1.93 (0.97-3.84) 1.07 × 10-2 GRK4
rs8336b 4 95211610 A 0.560 1.40 (0.71-2.76) NS SMARCAD1
rs2016483 4 95229039 A 0.490 1.73 (0.88-3.39) 2.22 × 10-2 SMARCAD1
rs9200 5 41142606 G 0.340 2.08 (1.06-4.06) 2.90 × 10-3 C6
rs2014012 5 58612388 A 0.160 2.89 (1.36-6.15) 2.22 × 10-2 PDE4D
rs353254b 5 148748736 A 0.470 1.48 (0.76-2.87) NS IL17B
PCYOX1L
rs372402 5 148752020 T 0.470 1.60 (0.82-3.12) 4.82 × 10-2 IL17B
PCYOX1L
rs4151667 6 31914024 A 0.900 3.72 (0.64-21.52) 5.90 × 10-3 CFB
rs17500510 6 32712818 A 0.850 2.21 (0.68-7.12) 3.12 × 10-2 HLA-DQA2
rs2071800b 6 32714143 T 0.908 1.78 (0.45-7.00) NS HLA-DQA2
rs2582 6 32974551 A 0.830 1.62 (0.60-4.41) NS HLA-DOA
rs733590 6 36645203 C 0.500 1.75 (0.89-3.44) 1.94 × 10-2 CDKN1A
rs2395655b 6 36645696 G 0.450 1.86 (0.95-3.63) 9.40 × 10-3 CDKN1A
rs4242391 8 23000183 C 0.340 1.78 (0.91-3.49) 1.82 × 10-2 TNFRSF10D
rs11575584 9 34661994 A 0.900 1.80 (0.48-6.80) NS CCL27
rs11257804 10 12496055 A 0.560 2.43 (1.16-5.09) 3.00 × 10-4 CAMK1D
rs549908b 11 112020916 G 0.680 1.57 (0.73-3.35) NS IL18
TEX12
rs11214105 11 112037653 A 0.680 1.91 (0.86-4.21) 1.56 × 10-2 IL18
TEX12
rs3802814 11 126162607 A 0.810 2.04 (0.74-5.66) 2.94 × 10-2 TIRAP
rs8177374b 11 126162843 T 0.810 1.60 (0.62-4.11) NS TIRAP
rs4251545 12 44180295 G 0.040 3.03 (0.85-10.72) 3.60 × 10-2 IRAK4
rs9550987 13 24167505 A 0.670 2.27 (1.00-5.15) 2.10 × 10-3 TNFRSF19
rs3751488 14 23304094 G 0.146 2.48 (1.13-5.46) 4.20 × 10-3 MRPL52
rs10498445 14 52740441 C 0.160 2.31 (1.07-5.02) 5.40 × 10-3 PTGDR
rs6115b 14 95053890 G 0.500 2.14 (1.07-4.29) 1.60 × 10-3 SERPINA5
rs6112 14 95054176 T 0.540 2.10 (1.03-4.26) 2.40 × 10-3 SERPINA5
rs6108 14 95058631 A 0.510 1.94 (0.98-3.87) 5.70 × 10-3 SERPINA5
rs9113b 14 95059076 T 0.510 1.91 (0.96-3.79) 7.10 × 10-3 SERPINA5
rs12439525 15 75087405 C 0.020 3.92 (0.75-20.5) NS LMAN1L
rs3803568b 15 75108636 C 0.010 6.54 (0.78-54.95) 3.84 × 10-2 LMAN1L
rs1051007 17 4636813 C 0.800 3.54 (1.06-11.77) 2.00 × 10-4 MED11 CXCL16
rs11658971b 17 4637698 A 0.810 2.89 (0.92-9.10) 1.90 × 10-3 MED11
CXCL16
rs2277680 17 4638563 A 0.390 1.59 (0.82-3.09) NS CSCL16
rs1050998b 17 4638737 T 0.390 1.56 (0.81-3.04) NS CSCL16
rs280502 19 10491475 T 0.830 2.15 (0.72-6.41) 2.59 × 10-2 TYK2
rs2278831 19 52131119 G 0.890 3.22 (0.67-15.54) 9.10 × 10-3 SIGLEC5
rs4819388 21 45647421 C 0.200 1.83 (0.87-3.86) 3.28 × 10-2 ICOSLG
rs228941b 22 37523721 C 0.190 1.81 (0.85-3.86) 4.07 × 10-2 IL2RB
rs228945 22 37525880 A 0.170 2.19 (1.02-4.71) 8.00 × 10-3 IL2RB
rs4253760 22 46622384 G 0.720 2.27 (0.94-5.48) 3.70 × 10-3 PPARA
Study by Schlauch and colleagues (2016) in a cohort of 42 cases and 38 controls.
rs2981884 1 3002072 C 0.803 NC 5.62 × 10-6 PRDM16
rs349391 1 4422005 C 0.303 1.73 (0.90-3.32) 7.67 × 10-6 -
rs349390 1 4423896 C 0.303 1.73 (0.90-3.32) 7.67 × 10-6 - rs12034948 1 4910371 G 0.053 8.53 (2.82-25.77) 2.00 × 10-5 -
rs17426290 1 5683697 T 0.250 2.60 (1.33-5.10) 1.54 × 10-7 -
rs686190 1 12223839 G 0.053 11.65 (3.88-34.92) 1.11 × 10-9 TNFRSF1B
rs7529216 1 18737780 G 0.039 9.17 (2.63-32.03) 1.52 × 10-6 -
rs3920498 1 22492887 C 0.711 8.15 (2.66-24.97) 2.68 × 10-5 LOC105376850
rs16826918 1 22644857 G 0.079 10.11 (3.96-25.82) 1.13 × 10-9 - rs2025499 1 35090749 G 0.079 5.83 (2.26-15.07) 2.82 × 10-5 LOC105378641
rs3913434 1 37449595 T 0.013 46.15 (6.11-348.47) 1.26 × 10-11 GRIK3
rs547977 1 44105650 G 0.000 NC 1.32 × 10-5 - rs12408925 1 48473192 G 0.276 5.24 (2.66-10.31) 2.61 × 10-7 -
rs12407818 1 52904410 C 0.263 3.23 (1.66-6.29) 8.75 × 10-7 ZCCHC11
Footnotes for Table 7.2 on page 143.
138 Chapter 7: Systematic Evaluation of Risk Loci from Candidate Gene and Genome-wide Association Studies of Fatigue
Table 7.2. Continued Summary of SNPs from genome-wide association studies for chronic fatigue
syndrome.
SNP Chr SNP position RA Freq OR (95% CI) p-value Genea
rs17130776 1 69095403 T 0.197 4.69 (2.31-9.54) 7.17 × 10-7 -
rs11163916 1 84691327 G 0.000 NC 5.44 × 10-6 PRKACB
rs233122 1 85790058 C 0.000 NC 3.27 × 10-7 DDAH1
rs6675622 1 94447733 T 0.395 9.20 (4.28-19.77) 5.94 × 10-9 -
rs3861911 1 106465037 C 0.132 4.27 (1.93-9.47) 2.31 × 10-5 LOC401957
rs7537461 1 113383662 C 0.289 3.79 (1.96-7.35) 8.53 × 10-8 LINC01356
rs7540424 1 116721566 A 0.553 2.43 (1.24-4.74) 2.37 × 10-5 -
rs7552454 1 144938879 A 0.197 3.88 (1.91-7.88) 2.11 × 10-7 PDE4DIP
rs2644567 1 146951082 G 0.724 10.31 (2.93-36.24) 1.41 × 10-5 LINC00624
rs11205084 1 152684306 G 0.303 2.79 (1.45-5.35) 7.25 × 10-6 -
rs12086522 1 153139169 T 0.092 6.07 (2.48-14.82) 2.61 × 10-7 LOC105371447
rs6679280 1 154017616 T 0.066 9.66 (3.53-26.41) 1.72 × 10-9 NUP210L
rs10737169 1 154653704 A 0.224 5.36 (2.68-10.75) 2.51 × 10-8 -
rs822027 1 155745810 A 0.013 35.53 (4.69-269.28) 2.52 × 10-9 GON4L
rs822020 1 155754117 C 0.026 20.56 (4.71-89.74) 5.15 × 10-7 GON4L
rs7549528 1 167485852 C 0.000 NC 1.54 × 10-8 CD247
rs275154 1 168061777 G 0.145 4.22 (1.95-9.14) 1.72 × 10-7 GPR161
rs2421987 1 172100831 A 0.132 1.80 (0.77-4.19) 2.47 × 10-5 DNM3
rs12120556 1 172115154 A 0.132 1.93 (0.83-4.46) 1.16 × 10-5 DNM3
rs17368935 1 172306391 G 0.066 9.66 (3.53-26.41) 1.72 × 10-9 DNM3
rs6662412 1 180237765 G 0.013 30.00 (3.94-228.2) 7.69 × 10-8 LHX4
rs589402 1 182542311 T 0.697 8.68 (2.84-26.52) 3.92 × 10-6 RNASEL
rs6656441 1 186182382 C 0.013 23.44 (3.06-179.51) 4.27 × 10-6 LOC105371654
rs689462 1 186651083 C 0.092 8.54 (3.52-20.76) 2.08 × 10-9 PTGS2
rs2816936 1 199982900 A 0.513 14.99 (5.46-41.14) 4.91 × 10-8 - rs7517843 1 212673179 G 0.000 NC 3.15 × 10-5 -
rs1926721 1 230864830 A 0.000 NC 3.15 × 10-5 LOC105373166
rs1458597 1 234154862 G 0.000 NC 3.15 × 10-5 SLC35F3
rs1367276 2 3662060 A 0.618 1.97 (1.00-3.91) 1.97 × 10-5 COLEC11
rs10207238 2 5615223 C 0.013 20.45 (2.66-157.42) 2.62 × 10-5 -
rs270838 2 7783504 C 0.105 7.73 (3.31-18.05) 3.61 × 10-11 - rs16861920 2 14759552 C 0.053 8.53 (2.82-25.77) 1.76 × 10-6 -
rs1366834 2 16389782 G 0.013 21.92 (2.86-168.29) 1.07 × 10-5 -
rs4099911 2 16454575 A 0.408 1.45 (0.78-2.72) 7.55 × 10-6 - rs654807 2 16457284 C 0.421 1.38 (0.74-2.57) 2.57 × 10-5 -
rs798368 2 16845834 T 0.605 1.84 (0.94-3.59) 1.78 × 10-5 FAM49A
rs16987589 2 20586506 A 0.013 23.44 (3.06-179.51) 4.27 × 10-6 - rs17043470 2 22397424 A 0.816 NC 1.52 × 10-5 -
rs2602803 2 30818644 G 0.303 2.20 (1.15-4.21) 9.83 × 10-6 LCLAT1
rs985257 2 38283228 A 0.237 4.30 (2.17-8.51) 2.15 × 10-5 RMDN2
rs1157185 2 38285735 T 0.237 4.30 (2.17-8.51) 5.84 × 10-6 RMDN2
rs1367696 2 38286914 T 0.237 4.30 (2.17-8.51) 5.84 × 10-6 RMDN2
rs13421497 2 45861591 G 0.053 9.49 (3.15-28.59) 1.57 × 10-5 - rs6757543 2 45977472 G 0.237 3.38 (1.71-6.67) 3.68 × 10-6 PRKCE
rs1007540 2 49209108 G 0.368 3.25 (1.70-6.21) 1.17 × 10-8 FSHR
rs6757577 2 65877598 A 0.066 11.18 (4.10-30.51) 2.77 × 10-10 - rs283825 2 79232491 G 0.342 2.83 (1.49-5.38) 4.41 × 10-7 -
rs13398697 2 103757890 A 0.066 7.49 (2.72-20.60) 1.05 × 10-6 -
rs1486178 2 108681163 T 0.000 NC 5.44 × 10-6 LINC01594
rs17041554 2 111564301 A 0.237 5.24 (2.63-10.42) 5.63 × 10-7 ACOXL
rs11895045 2 127492155 T 0.013 20.45 (2.66-157.42) 2.62 × 10-5 -
rs3732196 2 128744457 A 0.000 NC 8.56 × 10-7 SAP130
rs10928930 2 130498982 T 0.026 14.80 (3.36-65.16) 2.23 × 10-6 LOC105373643
rs16842140 2 140129426 G 0.263 2.94 (1.51-5.72) 9.07 × 10-7 LOC105373643
rs6744124 2 142815895 G 0.000 NC 3.15 × 10-5 LRP1B
rs13393078 2 146393980 C 0.079 7.55 (2.94-19.36) 1.24 × 10-7 -
rs2127978 2 157785902 T 0.092 4.93 (2.00-12.12) 1.97 × 10-5 -
rs6735919 2 166489066 T 0.250 2.73 (1.39-5.35) 3.82 × 10-6 CSRNP3
rs12165212 2 176079816 G 0.737 9.64 (2.73-34.01) 1.85 × 10-5 LOC105373751
rs11679695 2 206733638 G 0.224 2.73 (1.37-5.45) 2.80 × 10-6 -
rs16827966 2 232207158 T 0.013 41.67 (5.51-315.00) 5.32 × 10-11 ARMC9
rs622060 2 234948684 C 0.513 2.12 (1.11-4.03) 2.76 × 10-5 -
rs9683305 3 66866 C 0.447 4.87 (2.42-9.79) 3.28 × 10-5 -
rs2200706 3 3673608 T 0.263 5.91 (2.98-11.74) 5.48 × 10-10 - rs2193766 3 8829321 G 0.763 NC 7.80 × 10-8 OXTR
rs1597474 3 13542671 T 0.053 5.26 (1.70-16.27) 1.24 × 10-5 HDAC11
rs12629385 3 14407232 T 0.066 8.74 (3.19-23.95) 5.27 × 10-7 - rs11917596 3 19131933 C 0.026 13.13 (2.97-58.04) 6.34 × 10-6 -
rs197770 3 37515827 G 0.382 2.77 (1.46-5.26) 1.50 × 10-7 ITGA9
Footnotes for Table 7.2 on page 143.
Chapter 7: Systematic Evaluation of Risk Loci from Candidate Gene and Genome-wide Association Studies of Fatigue 139
Table 7.2. Continued Summary of SNPs from genome-wide association studies for chronic fatigue
syndrome.
SNP Chr SNP position RA Freq OR (95% CI) p-value Genea
rs9844641 3 43476335 A 0.750 27.67 (3.60-212.56) 6.65 × 10-6 ANO10
rs9311374 3 45672925 T 0.816 NC 5.62 × 10-6 LIMD1
rs7426702 3 46095074 T 0.132 4.27 (1.93-9.47) 2.31 × 10-5 -
rs4473594 3 46337356 A 0.092 8.54 (3.52-20.76) 1.81 × 10-8 -
rs6445832 3 56905923 G 0.079 8.75 (3.42-22.38) 4.36 × 10-10 ARHGEF3
rs17060061 3 59031063 G 0.763 NC 7.80 × 10-8 C3orf67
rs11711551 3 66697961 G 0.132 5.20 (2.35-11.48) 2.31 × 10-5 -
rs17047694 3 68475098 T 0.303 3.07 (1.60-5.90) 3.66 × 10-9 FAM19A1
rs4422316 3 69202605 T 0.000 NC 3.27 × 10-7 -
rs12629627 3 73629548 G 0.355 3.10 (1.63-5.92) 1.76 × 10-5 PDZRN3
rs6797416 3 80211780 G 0.000 NC 1.71 × 10-9 - rs17019070 3 81603483 C 0.184 3.66 (1.78-7.53) 6.92 × 10-6 GBE1
rs7613828 3 85851246 A 0.737 NC 7.03 × 10-7 CADM2
rs1523773 3 97019048 T 0.000 NC 4.73 × 10-11 EPHA6
rs41330648 3 100760807 G 0.118 8.19 (3.62-18.54) 1.28 × 10-7 -
rs340170 3 111826736 G 0.000 NC 5.44 × 10-6 C3orf52
rs2733416 3 114082645 G 0.000 NC 1.71 × 10-9 ZBTB20
rs11914436 3 121123223 A 0.368 2.65 (1.40-5.02) 1.93 × 10-7 STXBP5L
rs361236 3 136729811 A 0.276 3.33 (1.72-6.45) 1.07 × 10-8 IL20RB
rs890527 3 140774853 T 0.289 3.12 (1.62-6.01) 4.60 × 10-9 SPSB4
rs2196007 3 144120914 T 0.053 8.07 (2.66-24.44) 1.11 × 10-5 -
rs7610618 3 149157706 T 0.013 26.61 (3.49-203.06) 1.57 × 10-6 - rs4505649 3 156146480 C 0.000 NC 3.15 × 10-5 KCNAB1
rs17780243 3 158712150 T 0.368 2.52 (1.33-4.77) 1.95 × 10-6 -
rs41469844 3 176694604 C 0.342 5.10 (2.60-10.01) 1.86 × 10-5 - rs16844808 4 3717581 T 0.079 4.94 (1.90-12.86) 2.42 × 10-6 -
rs17675581 4 5080187 G 0.276 2.75 (1.42-5.32) 3.05 × 10-5 STK32B
rs10009657 4 9833734 G 0.750 6.67 (2.15-20.65) 5.76 × 10-6 SLC2A9
rs16877795 4 11087682 G 0.171 4.85 (2.33-10.10) 1.49 × 10-7 -
rs1873717 4 11767981 T 0.184 4.87 (2.37-10.02) 1.23 × 10-6 LOC105374484
rs41423649 4 21898055 C 0.711 11.00 (3.14-38.57) 5.73 × 10-6 KCNIP4
rs7672066 4 26728719 G 0.000 NC 8.56 × 10-7 TBC1D19
rs1433429 4 35899687 C 0.197 3.88 (1.91-7.88) 2.71 × 10-5 -
rs10517378 4 36532369 C 0.250 3.15 (1.61-6.17) 1.13 × 10-5 LOC105374400
rs2303409 4 37846939 C 0.132 4.71 (2.13-10.43) 3.19 × 10-5 PGM2
rs13148734 4 63330858 A 0.500 3.67 (1.84-7.30) 2.73 × 10-5 LOC100131441
rs4510466 4 113901071 C 0.000 NC 1.22 × 10-7 ANK2
rs17865437 4 118341034 A 0.671 3.25 (1.47-7.20) 7.39 × 10-7 RPSAP35
rs17861907 4 118355398 G 0.671 3.25 (1.47-7.20) 7.39 × 10-7 LINC01378
rs11934366 4 145541864 A 0.000 NC 1.32 × 10-5 - rs1961484 4 156307533 A 0.066 10.65 (3.90-29.08) 1.21 × 10-7 -
rs2882361 4 161379616 G 0.513 10.44 (4.26-25.54) 3.02 × 10-8 -
rs12331711 4 162419624 G 0.066 8.74 (3.19-23.95) 1.52 × 10-6 FSTL5
rs3792615 4 164532801 T 0.724 15.65 (3.53-69.47) 1.66 × 10-5 MARCH1
rs4692612 4 171537901 T 0.092 5.77 (2.36-14.11) 1.17 × 10-5 -
rs6854376 4 174202313 T 0.053 10.53 (3.50-31.63) 1.71 × 10-8 GALNT7
rs2685850 4 190778060 A 0.316 5.75 (2.91-11.36) 2.70 × 10-5 -
rs16886994 5 20464699 G 0.026 13.95 (3.16-61.54) 1.32 × 10-5 CDH18
rs1428323 5 29912034 A 0.237 2.79 (1.41-5.52) 3.28 × 10-5 - rs6892871 5 36669569 G 0.026 7.40 (1.62-33.74) 1.81 × 10-5 SLC1A3
rs7726463 5 37955073 G 0.355 3.83 (1.99-7.38) 5.64 × 10-7 -
rs6871885 5 53296188 A 0.092 6.38 (2.61-15.57) 1.42 × 10-6 ARL15
rs6450296 5 54464679 A 0.118 4.58 (2.01-10.44) 5.16 × 10-6 CDC20B
rs16888306 5 57920048 C 0.079 5.83 (2.26-15.07) 1.31 × 10-5 RAB3C
rs10056584 5 60401442 A 0.026 13.13 (2.97-58.04) 1.46 × 10-5 NDUFAF2
rs6449669 5 62929018 T 0.289 3.61 (1.87-6.98) 6.54 × 10-6 -
rs6863118 5 71240748 G 0.105 7.37 (3.15-17.22) 6.22 × 10-9 -
rs609539 5 106904997 C 0.0132 26.61 (3.49-203.06) 6.14 × 10-7 EFNA5
rs254577 5 134422204 C 0.264 11.04 (5.28-23.07) 2.35 × 10-11 C5orf66
rs6877860 5 135271527 T 0.145 3.28 (1.51-7.16) 2.78 × 10-5 FBXL21
rs11954603 5 136488381 C 0.237 2.19 (1.10-4.35) 1.30 × 10-5 SPOCK1
rs889083 5 145941549 G 0.224 2.87 (1.44-5.71) 3.14 × 10-6 CTB-99A3.1
rs17722227 5 148407945 A 0.026 8.04 (1.77-36.46) 7.70 × 10-6 SH3TC2
rs7701654 5 149956501 G 0.013 30.00 (3.94-228.20) 7.69 × 10-8 - rs10074876 5 152867311 C 0.026 16.59 (3.78-72.77) 1.07 × 10-7 GRIA1
rs6892217 5 171474120 T 0.224 8.19 (4.01-16.73) 6.61 × 10-10 STK10
rs7768988 6 7830384 T 0.237 6.11 (3.05-12.24) 7.15 × 10-7 BMP6
rs6940702 6 11702281 A 0.000 NC 2.18 × 10-6 LOC340184
Footnotes for Table 7.2 on page 143.
140 Chapter 7: Systematic Evaluation of Risk Loci from Candidate Gene and Genome-wide Association Studies of Fatigue
Table 7.2. Continued Summary of SNPs from genome-wide association studies for chronic fatigue
syndrome.
SNP Chr SNP position RA Freq OR (95% CI) p-value Genea
rs41378447 6 22141745 T 0.092 10.84 (4.46-26.34) 1.06 × 10-11 CASC15
NBAT1
rs4714199 6 38891068 C 0.158 5.59 (2.64-11.85) 3.92 × 10-6 DNAH8
rs2436739 6 40496341 G 0.000 NC 3.15 × 10-5 LRFN2
rs2274515 6 42933526 T 0.316 3.52 (1.83-6.77) 2.19 × 10-5 PEX6
rs2748997 6 52601316 C 0.079 9.18 (3.59-23.48) 2.76 × 10-8 -
rs9283919 6 54114066 G 0.053 11.08 (3.69-33.24) 7.87 × 10-7 MLIP
rs9446695 6 73232658 T 0.026 17.53 (4.00-76.78) 3.46 × 10-8 - rs6927507 6 77792085 G 0.184 3.32 (1.61-6.85) 1.99 × 10-6 LOC105377862
rs16890805 6 80195421 T 0.105 4.48 (1.90-10.59) 3.34 × 10-6 LCA5
rs12055682 6 81189123 G 0.263 4.79 (2.43-9.41) 2.99 × 10-9 - rs7739542 6 84716154 C 0.776 NC 2.38 × 10-7 -
rs9362453 6 88516498 G 0.039 9.17 (2.63-32.03) 6.40 × 10-6 LOC101928911
rs3798405 6 116738674 C 0.000 NC 3.15 × 10-5 DSE
rs606324 6 124855854 A 0.079 6.48 (2.52-16.69) 4.39 × 10-8 NKAIN2
rs7747443 6 136713749 C 0.053 10.00 (3.32-30.08) 2.13 × 10-5 MAP7
rs6923953 6 136726688 C 0.053 10.00 (3.32-30.08) 2.13 × 10-5 MAP7
rs997139 6 136751118 G 0.053 10.00 (3.32-30.08) 2.13 × 10-5 MAP7
rs6926583 6 136752092 C 0.053 10.00 (3.32-30.08) 2.13 × 10-5 MAP7
rs11154872 6 136797757 C 0.053 9.49 (3.15-28.59) 2.64 × 10-5 MAP7
rs3778315 6 136850687 G 0.039 12.17 (3.52-42.07) 1.65 × 10-5 MAP7
rs7742257 6 144987202 T 0.053 8.07 (2.66-24.44) 4.89 × 10-6 UTRN
rs9485028 6 146209670 A 0.066 5.04 (1.80-14.1) 2.53 × 10-5 SHPRH
rs17085519 6 154870688 G 0.026 13.13 (2.97-58.04) 6.34 × 10-6 LOC100129996
rs1859512 7 8535486 T 0.066 6.73 (2.44-18.58) 8.54 × 10-6 NXPH1
rs6973776 7 38319378 T 0.000 NC 3.27 × 10-7 TARP
rs2237406 7 39423305 T 0.026 13.13 (2.97-58.04) 6.34 × 10-6 POU6F2
rs7789233 7 51947068 A 0.000 NC 1.32 × 10-5 - rs11506050 7 52160653 G 0.013 20.45 (2.66-157.42) 2.62 × 10-5 -
rs1195242 7 68686390 C 0.000 NC 1.32 × 10-5 -
rs41456945 7 71149459 C 0.026 18.50 (4.23-80.94) 1.07 × 10-8 WBSCR17
rs1859790 7 75916073 T 0.092 6.70 (2.75-16.34) 1.56 × 10-7 SRRM3
rs17156195 7 81981170 T 0.118 5.32 (2.34-12.07) 9.34 × 10-6 CACNA2D1
rs4623336 7 98088932 T 0.158 4.85 (2.29-10.27) 2.68 × 10-8 - rs41385645 7 98974038 T 0.132 6.29 (2.85-13.88) 1.99 × 10-7 ARPC1B
rs17475512 7 102512488 G 0.382 3.62 (1.88-6.96) 2.56 × 10-6 FBXL13
FAM185A
rs6957524 7 106186766 G 0.013 26.61 (3.49-203.06) 6.14 × 10-7 -
rs213981 7 117254527 G 0.329 4.31 (2.22-8.35) 2.94 × 10-7 CFTR
rs7783582 7 123169504 T 0.039 9.17 (2.63-32.03) 1.62 × 10-5 IQUB
rs1526415 7 125079875 T 0.132 4.06 (1.83-9.02) 3.05 × 10-5 LOC100506664
rs1222400 7 133817981 T 0.039 12.83 (3.72-44.3) 2.89 × 10-8 LRGUK
rs2960770 7 142012291 C 0.539 5.67 (2.60-12.33) 3.01 × 10-5 TRB
rs11972875 7 154647387 G 0.053 7.20 (2.37-21.90) 3.27 × 10-5 DPP6
rs6950641 7 157029359 T 0.092 5.48 (2.23-13.42) 6.49 × 10-6 UBE3C
rs41363145 8 6064108 C 0.079 6.82 (2.65-17.54) 1.32 × 10-6 - rs2916699 8 6214922 A 0.026 13.13 (2.97-58.04) 1.46 × 10-5 -
rs11984468 8 16367555 C 0.184 4.87 (2.37-10.02) 2.04 × 10-5 LOC101929028
rs17643851 8 18437729 G 0.079 8.75 (3.42-22.38) 9.93 × 10-8 PSD3
rs17733133 8 22385061 G 0.171 4.85 (2.33-10.10) 2.60 × 10-5 PPP3CC
rs17052315 8 24517047 A 0.053 10.00 (3.32-30.08) 5.97 × 10-8 -
rs4236924 8 41975979 C 0.079 5.83 (2.26-15.07) 1.31 × 10-5 - rs4738955 8 63260182 A 0.382 1.70 (0.91-3.19) 8.10 × 10-6 NKAIN3
rs1350060 8 63848337 A 0.211 3.10 (1.54-6.23) 2.01 × 10-6 NKAIN3
rs16937494 8 72101749 G 0.039 9.17 (2.63-32.03) 1.62 × 10-5 LOC105375894
rs2033069 8 81079607 A 0.539 5.67 (2.60-12.33) 1.03 × 10-5 TPD52
rs7010471 8 97350955 G 0.053 12.24 (4.09-36.66) 2.49 × 10-10 -
rs16883408 8 113331553 C 0.237 5.51 (2.76-10.98) 1.06 × 10-8 CSMD3
rs7830366 8 114880641 T 0.316 3.52 (1.83-6.77) 4.48 × 10-6 -
rs6470455 8 127678907 G 0.342 4.81 (2.46-9.39) 7.73 × 10-6 -
rs7834482 8 127785745 G 0.461 2.93 (1.52-5.63) 2.83 × 10-5 LOC105375753
rs2648883 8 129076594 G 0.250 4.20 (2.14-8.26) 2.27 × 10-5 PVT1
rs16902672 8 129165859 C 0.211 5.25 (2.60-10.59) 1.77 × 10-8 -
rs7011650 8 134390976 T 0.263 3.08 (1.58-6.00) 5.89 × 10-6 LOC105375771
rs12001751 9 2295279 T 0.000 NC 5.44 × 10-6 LOC105375955
rs12551218 9 9058949 T 0.079 5.23 (2.02-13.57) 3.31 × 10-6 PTPRD
rs2891242 9 25062644 C 0.171 4.85 (2.33-10.10) 9.33 × 10-6 - rs10114442 9 27335837 A 0.000 NC 2.18 × 10-6 MOB3B
rs7847862 9 35933210 G 0.329 3.49 (1.82-6.70) 3.47 × 10-7 -
Footnotes for Table 7.2 on page 143.
Chapter 7: Systematic Evaluation of Risk Loci from Candidate Gene and Genome-wide Association Studies of Fatigue 141
Table 7.2. Continued Summary of SNPs from genome-wide association studies for chronic fatigue
syndrome.
SNP Chr SNP position RA Freq OR (95% CI) p-value Genea
rs7863401 9 35942406 A 0.645 5.23 (2.20-12.46) 2.44 × 10-5 -
rs12684292 9 35944224 T 0.645 5.23 (2.20-12.46) 2.44 × 10-5 - rs7019283 9 35945975 C 0.632 5.54 (2.33-13.16) 8.05 × 10-6 -
rs13285078 9 35946275 C 0.632 5.54 (2.33-13.16) 8.05 × 10-6 -
rs12235235 9 36091133 T 0.079 10.61 (4.15-27.08) 5.76 × 10-16 RECK
rs2988013 9 37062647 C 0.053 8.07 (2.66-24.44) 2.38 × 10-5 -
rs7019328 9 74835176 T 0.645 14.88 (4.29-51.64) 5.66 × 10-6 GDA
rs10121299 9 79397301 C 0.289 4.42 (2.27-8.61) 6.62 × 10-9 PRUNE2
PCA3
rs17085969 9 85640254 T 0.750 27.67 (3.60-212.56) 8.99 × 10-7 RASEF
rs10978470 9 109136192 G 0.158 5.59 (2.64-11.85) 4.32 × 10-9 - rs1610024 9 111614766 A 0.250 4.64 (2.35-9.14) 2.89 × 10-7 -
rs10980229 9 112925977 G 0.289 2.45 (1.28-4.72) 1.41 × 10-5 PALM2-AKAP2
rs10817082 9 113474166 C 0.632 6.42 (2.6-15.83) 2.53 × 10-5 MUSK
rs7849492 9 122619031 C 0.145 5.91 (2.74-12.75) 9.95 × 10-10 -
rs7020077 9 127010874 A 0.184 4.22 (2.05-8.68) 1.87 × 10-5 -
rs7859623 9 127314768 C 0.000 NC 3.15 × 10-5 NR6A1
rs7853174 9 129419990 G 0.513 1.26 (0.68-2.36) 2.17 × 10-5 LMX1B
rs10988052 9 131353253 G 0.079 6.15 (2.39-15.86) 4.71 × 10-6 SPTAN1
rs418216 10 4668586 T 0.000 NC 1.32 × 10-5 - rs4242794 10 5536590 A 0.000 NC 3.27 × 10-7 LOC105376379
rs7095919 10 11267360 G 0.079 5.83 (2.26-15.07) 1.31 × 10-5 CELF2
rs584569 10 30496246 A 0.066 8.74 (3.19-23.95) 2.84 × 10-8 -
rs2490495 10 32726529 G 0.342 2.02 (1.07-3.82) 2.97 × 10-5 LOC101929431
rs12761944 10 32803484 A 0.342 2.02 (1.07-3.82) 2.97 × 10-5 CCDC7
rs1763788 10 32917853 A 0.342 2.02 (1.07-3.82) 2.97 × 10-5 CCDC7
rs1577372 10 32938382 A 0.342 2.02 (1.07-3.82) 2.97 × 10-5 CCDC7
rs1762529 10 32968080 A 0.342 2.02 (1.07-3.82) 2.97 × 10-5 CCDC7
rs2784574 10 32976689 G 0.342 2.02 (1.07-3.82) 2.97 × 10-5 CCDC7
rs11009106 10 33123413 C 0.329 2.24 (1.18-4.26) 1.42 × 10-5 CCDC7
rs2995467 10 33135952 G 0.342 2.02 (1.07-3.82) 2.97 × 10-5 CCDC7
rs11010290 10 36060309 T 0.395 6.04 (2.99-12.21) 1.61 × 10-6 -
rs7895391 10 47006369 T 0.145 5.37 (2.49-11.59) 7.78 × 10-7 -
rs12572431 10 47592472 G 0.250 3.46 (1.77-6.79) 9.90 × 10-6 ANTXRLP1
rs6479969 10 52900003 G 0.382 5.54 (2.78-11.05) 1.28 × 10-7 PRKG1
rs1915603 10 66455644 G 0.039 9.73 (2.79-33.90) 5.15 × 10-8 -
rs16926249 10 71100726 G 0.474 3.13 (1.61-6.08) 1.31 × 10-5 HK1
rs2288374 10 79761041 T 0.132 4.06 (1.83-9.02) 9.09 × 10-6 POLR3A
rs17112444 10 101749974 A 0.026 21.64 (4.96-94.39) 8.02 × 10-10 DNMBP
rs1932556 10 120251528 T 0.605 54.13 (7.15-409.98) 1.63 × 10-9 - rs10788258 10 123936896 T 0.092 7.39 (3.04-17.99) 1.04 × 10-7 TACC2
rs2421122 10 124473379 C 0.000 NC 3.27 × 10-7 -
rs2803453 10 131075207 C 0.000 NC 2.18 × 10-6 - rs1041296 10 132003031 G 0.211 4.76 (2.37-9.59) 6.89 × 10-9 -
rs9419277 10 133834704 G 0.026 14.8 (3.36-65.16) 8.93 × 10-7 -
rs11021876 11 11509885 T 0.013 28.28 (3.71-215.42) 2.21 × 10-7 GALNT18
rs11027583 11 23897934 T 0.092 7.39 (3.04-17.99) 7.03 × 10-9 -
rs11038285 11 45071321 G 0.013 30.00 (3.94-228.20) 7.69 × 10-8 -
rs7121660 11 45358483 A 0.447 4.53 (2.27-9.03) 4.10 × 10-6 - rs1977985 11 59052769 G 0.750 9.00 (2.54-31.85) 2.96 × 10-5 SLC25A47P1
rs3017495 11 70666253 T 0.026 13.95 (3.16-61.54) 2.42 × 10-6 SHANK2
rs7107438 11 75709833 C 0.026 12.33 (2.78-54.66) 1.60 × 10-5 UVRAG
rs17133553 11 99316881 A 0.276 5.53 (2.80-10.91) 4.74 × 10-8 CNTN5
rs10789931 11 112842773 T 0.303 3.74 (1.94-7.23) 3.08 × 10-5 NCAM1
rs12417706 11 116059440 T 0.105 7.02 (3.00-16.42) 1.90 × 10-6 - rs3867246 11 124271743 T 0.132 5.20 (2.35-11.48) 1.88 × 10-9 -
rs4144897 11 128181191 T 0.118 5.06 (2.23-11.51) 5.12 × 10-6 -
rs7119924 11 130004026 C 0.184 3.66 (1.78-7.53) 1.58 × 10-7 APLP2
rs12305678 12 2763539 G 0.224 4.41 (2.21-8.79) 7.87 × 10-9 CACNA1C
rs11062852 12 3936632 C 0.197 4.07 (2.00-8.26) 4.84 × 10-9 PARP11
rs14541 12 8800566 G 0.224 4.20 (2.11-8.37) 1.70 × 10-6 MFAP5
rs11056347 12 15243991 A 0.053 9.49 (3.15-28.59) 5.27 × 10-7 -
rs16927111 12 24207823 G 0.211 3.25 (1.62-6.54) 1.69 × 10-5 SOX5
rs11168709 12 38984042 T 0.013 35.53 (4.69-269.28) 2.52 × 10-9 - rs2062758 12 39045452 T 0.105 7.02 (3.00-16.42) 1.18 × 10-5 -
rs12300888 12 52531500 C 0.079 6.82 (2.65-17.54) 2.65 × 10-6 -
rs17123453 12 60553534 C 0.066 6.37 (2.30-17.62) 2.25 × 10-5 - rs41441747 12 61745898 C 0.197 4.27 (2.10-8.66) 6.72 × 10-7 -
rs7307225 12 71898358 G 0.316 3.19 (1.66-6.11) 2.53 × 10-5 LGR5
Footnotes for Table 7.2 on page 143.
142 Chapter 7: Systematic Evaluation of Risk Loci from Candidate Gene and Genome-wide Association Studies of Fatigue
Table 7.2. Continued Summary of SNPs from genome-wide association studies for chronic fatigue
syndrome.
SNP Chr SNP position RA Freq OR (95% CI) p-value Genea
rs17019561 12 92118436 C 0.013 23.44 (3.06-179.51) 1.02 × 10-5 -
rs12312259 12 92148729 C 0.158 6.46 (3.05-13.69) 3.60 × 10-10 - rs17024760 12 96220100 T 0.211 7.11 (3.49-14.49) 3.22 × 10-7 LOC105369921
rs7301442 12 104026861 T 0.013 30.00 (3.94-228.20) 7.69 × 10-8 STAB2
rs17035358 12 104709745 A 0.066 9.66 (3.53-26.41) 1.72 × 10-9 TXNRD1
rs9668748 12 114387000 C 0.079 8.75 (3.42-22.38) 4.02 × 10-6 RBM19
rs7960674 12 123089209 C 0.000 NC 5.44 × 10-6 KNTC1
rs7306948 12 123345347 G 0.145 4.43 (2.05-9.59) 1.01 × 10-5 HIP1R
rs866781 12 128357089 A 0.000 NC 8.56 × 10-7 -
rs12317807 12 130395910 T 0.132 4.71 (2.13-10.43) 1.47 × 10-9 -
rs1696407 12 130459987 G 0.197 3.70 (1.82-7.51) 1.46 × 10-5 LOC105370076
rs2801659 13 19383274 C 0.237 5.24 (2.63-10.42) 1.32 × 10-5 -
rs9285128 13 21088186 A 0.250 4.00 (2.04-7.86) 2.15 × 10-9 CRYL1
rs7987491 13 21672725 G 0.158 4.20 (1.98-8.91) 1.19 × 10-5 -
rs17079111 13 24218816 G 0.053 9.00 (2.98-27.15) 1.53 × 10-6 TNFRSF19
rs9581771 13 27452351 T 0.092 7.76 (3.19-18.87) 7.96 × 10-9 -
rs17647077 13 46617212 G 0.000 NC 1.32 × 10-5 ZC3H13
rs41464146 13 47764748 C 0.026 18.50 (4.23-80.94) 3.22 × 10-8 -
rs9585049 13 100047159 T 0.039 15.75 (4.58-54.13) 5.25 × 10-10 -
rs10047684 13 105365339 A 0.237 4.51 (2.28-8.94) 7.31 × 10-8 - rs2017563 13 110262110 A 0.013 41.67 (5.51-315.00) 1.55 × 10-7 LOC105370359
rs9301483 13 111530373 A 0.250 5.40 (2.72-10.71) 8.08 × 10-8 - rs7321094 13 111684162 T 0.237 4.30 (2.17-8.51) 1.45 × 10-6 -
rs2204978 14 22518491 A 0.211 3.58 (1.78-7.19) 1.26 × 10-7 TRA
rs17255510 14 22662856 C 0.171 10.23 (4.82-21.71) 6.61 × 10-10 TRA
rs11157573 14 22889777 G 0.158 5.09 (2.40-10.77) 2.97 × 10-10 TRA
rs10144138 14 22933962 T 0.026 27.75 (6.38-120.63) 6.99 × 10-14 TRA
TRD
rs4982735 14 23626745 C 0.000 NC 3.27 × 10-7 SLC7A8
rs17256392 14 23987437 A 0.026 7.40 (1.62-33.74) 1.81 × 10-5 THTPA
rs17781246 14 41939907 G 0.289 2.57 (1.34-4.95) 1.48 × 10-6 - rs2816751 14 49812483 C 0.184 5.11 (2.48-10.51) 5.43 × 10-10 -
rs10146102 14 51850519 T 0.079 7.18 (2.80-18.43) 1.09 × 10-6 -
rs17127809 14 55188357 T 0.053 8.53 (2.82-25.77) 1.76 × 10-6 SAMD4A
rs7154569 14 61945965 T 0.000 NC 2.18 × 10-6 PRKCH
rs17098846 14 62032938 A 0.026 13.95 (3.16-61.54) 2.42 × 10-6 LOC101927780
rs10483750 14 63135397 T 0.079 5.83 (2.26-15.07) 1.31 × 10-5 - rs7153874 14 66308460 G 1.000 NC 1.32 × 10-5 -
rs7143222 14 73071319 T 0.000 NC 1.54 × 10-8 DPF3
rs2079989 14 73244469 C 0.421 1.51 (0.81-2.83) 6.68 × 10-6 DPF3
rs10133617 14 81134332 T 0.039 10.31 (2.97-35.84) 2.36 × 10-6 CEP128
rs17120254 14 85209862 A 0.605 NC 5.20 × 10-13 -
rs10129777 14 85516453 G 0.079 7.93 (3.10-20.32) 6.97 × 10-6 - rs8016502 14 85657757 G 0.013 20.45 (2.66-157.42) 2.62 × 10-5 LOC105370604
rs10137248 14 90320818 G 0.039 7.11 (2.01-25.14) 6.59 × 10-6 EFCAB11
rs2249954 14 92383999 G 0.079 10.11 (3.96-25.82) 5.47 × 10-11 FBLN5
rs10144861 14 92465346 G 0.145 4.65 (2.15-10.05) 2.74 × 10-5 TRIP11
rs7159091 14 95345271 C 0.039 6.64 (1.87-23.56) 1.58 × 10-5 -
rs17092382 14 95920937 A 0.000 NC 1.54 × 10-8 SYNE3
rs12443497 15 52373388 T 0.013 21.92 (2.86-168.29) 1.07 × 10-5 -
rs2869820 15 79217808 T 0.000 NC 4.39 × 10-8 CTSH
rs9920285 15 79488084 A 0.039 10.31 (2.97-35.84) 2.36 × 10-6 ANKRD34C-AS1
rs8029503 15 92488592 T 0.105 8.10 (3.47-18.93) 5.66 × 10-11 SLCO3A1
rs9744291 15 98552768 G 0.276 5.84 (2.95-11.57) 1.08 × 10-5 LOC105371008
rs8050875 16 11223537 G 0.724 15.65 (3.53-69.47) 6.91 × 10-6 CLEC16A
rs16970887 16 21115345 C 0.066 6.37 (2.30-17.62) 2.25 × 10-5 DNAH3
rs16973831 16 24572946 T 0.000 NC 1.22 × 10-7 RBBP6
rs13339179 16 25117049 T 0.039 12.83 (3.72-44.3) 2.89 × 10-8 LCMT1-AS1
rs16975878 16 26454601 G 0.789 22.13 (2.86-171.49) 2.06 × 10-5 -
rs6497951 16 26528582 T 0.000 NC 5.44 × 10-6 -
rs3095598 16 52566862 C 0.039 18.25 (5.32-62.62) 1.02 × 10-10 TOX3
rs41368852 16 65930903 G 0.039 12.83 (3.72-44.30) 3.28 × 10-6 -
rs4843884 16 86030100 G 0.066 8.31 (3.03-22.79) 1.83 × 10-6 -
rs8046503 16 86115955 A 0.316 3.19 (1.66-6.11) 8.41 × 10-6 - rs8057267 16 87541080 G 0.171 4.41 (2.11-9.19) 1.06 × 10-7 LOC101928737
rs7404102 16 88282637 A 0.000 NC 5.44 × 10-6 ZNF469
rs7220341 17 4214705 G 0.276 4.97 (2.53-9.75) 5.41 × 10-10 UBE2G1
rs6502875 17 5612641 G 0.316 3.90 (2.02-7.53) 5.97 × 10-8 -
rs16956158 17 6594844 G 0.447 3.95 (2.01-7.77) 6.93 × 10-7 SLC13A5
Footnotes for Table 7.2 on page 143.
Chapter 7: Systematic Evaluation of Risk Loci from Candidate Gene and Genome-wide Association Studies of Fatigue 143
Table 7.2. Continued Summary of SNPs from genome-wide association studies for chronic fatigue
syndrome.
SNP Chr SNP position RA Freq OR (95% CI) p-value Genea
rs16944757 17 11349689 A 0.053 7.63 (2.51-23.15) 2.80 × 10-5 SHISA6
rs4792493 17 14404572 G 0.000 NC 3.27 × 10-7 - rs3095168 17 16260198 A 0.237 3.22 (1.63-6.36) 1.91 × 10-6 -
rs271662 17 48018370 C 0.132 5.72 (2.59-12.62) 2.60 × 10-6 -
rs2192217 17 54500732 A 0.026 12.33 (2.78-54.66) 1.60 × 10-5 ANKFN1
rs9913705 17 60004928 G 0.013 25.00 (3.27-191.10) 1.65 × 10-6 INTS2
rs8073194 17 63601658 T 0.013 20.45 (2.66-157.42) 2.62 × 10-5 -
rs6504560 17 65959300 T 0.026 17.53 (4.00-76.78) 2.69 × 10-7 BPTF
rs690607 17 72888412 A 0.158 4.85 (2.29-10.27) 1.56 × 10-7 FADS6
rs29110 18 9951994 C 0.013 20.45 (2.66-157.42) 2.62 × 10-5 VAPA
rs496731 18 26491368 T 0.474 3.56 (1.81-6.98) 5.52 × 10-6 - rs1362859 18 32918639 G 0.329 2.86 (1.50-5.45) 8.99 × 10-8 ZNF24
rs948440 18 34820988 C 0.132 6.92 (3.14-15.26) 3.92 × 10-10 CELF4
rs12607783 18 49038486 A 0.132 4.71 (2.13-10.43) 4.31 × 10-8 LINC01630
rs12965947 18 49880082 A 0.053 12.24 (4.09-36.66) 5.82 × 10-7 DCC
rs9964872 18 62176077 A 0.026 17.53 (4.00-76.78) 9.92 × 10-8 -
rs9946817 18 70367007 C 0.158 2.96 (1.38-6.34) 5.35 × 10-7 - rs4892034 18 70399988 A 0.118 3.53 (1.53-8.11) 1.58 × 10-5 -
rs11873202 18 72716230 G 0.000 NC 8.56 × 10-7 ZNF407
rs682564 18 76177100 A 0.026 13.95 (3.16-61.54) 2.42 × 10-6 - rs243391 19 4449808 G 0.276 3.33 (1.72-6.45) 2.96 × 10-6 UBXN6
rs16994314 19 7176974 T 0.118 6.45 (2.85-14.61) 2.42 × 10-7 INSR
rs10402951 19 7555092 C 0.342 2.97 (1.56-5.67) 1.13 × 10-5 PEX11G
rs479448 19 7831061 C 0.026 13.13 (2.97-58.04) 6.34 × 10-6 CLEC4M
rs4808297 19 21935700 T 0.132 4.06 (1.83-9.02) 1.16 × 10-5 ZNF100
rs16970196 19 35794984 A 0.750 6.67 (2.15-20.65) 5.76 × 10-6 MAG
rs6508891 19 40128737 T 0.118 5.58 (2.46-12.67) 1.75 × 10-7 LOC100129935
rs7253295 19 50106308 A 0.829 NC 1.52 × 10-5 PRR12
rs6055456 20 7985376 C 0.105 4.97 (2.11-11.70) 1.83 × 10-5 TMX4
rs6074914 20 15519613 A 0.658 8.22 (2.96-22.80) 3.20 × 10-5 MACROD2
rs7347140 20 40738421 T 0.026 14.80 (3.36-65.16) 2.23 × 10-6 PTPRT
rs6093591 20 40831399 T 0.184 4.03 (1.96-8.28) 6.67 × 10-7 PTPRT
rs7272593 20 44432523 G 0.053 8.07 (2.66-24.44) 4.89 × 10-6 DNTTIP1
rs41493945 20 50957627 A 0.013 48.53 (6.43-366.21) 6.25 × 10-13 LOC105372666
rs2294584 20 51309773 T 0.000 NC 1.32 × 10-5 LOC105372666
rs927651 20 52772896 G 0.461 2.93 (1.52-5.63) 3.21 × 10-5 CYP24A1
rs6098723 20 54259204 G 0.013 20.45 (2.66-157.42) 2.62 × 10-5 LOC105372676
rs4812100 20 58145558 G 0.289 4.20 (2.16-8.16) 2.21 × 10-6 -
rs9977796 21 16868783 G 0.000 NC 8.56 × 10-7 -
rs13052044 21 19655756 T 0.158 5.33 (2.52-11.30) 2.66 × 10-5 TMPRSS15
rs7279994 21 31753700 C 0.211 3.75 (1.87-7.54) 2.22 × 10-5 -
rs9984519 21 34196975 T 0.316 1.97 (1.03-3.76) 8.22 × 10-6 -
rs8130198 21 43630324 C 0.382 2.27 (1.20-4.28) 6.63 × 10-6 ABCG1
rs3788079 21 45348179 C 0.000 NC 3.42 × 10-12 AGPAT3
rs7290437 22 22456739 G 0.250 2.73 (1.39-5.35) 3.52 × 10-8 IGL
rs16985794 22 22561277 C 0.066 6.02 (2.17-16.69) 1.58 × 10-6 IGL
rs16980810 22 26219764 A 0.158 4.62 (2.18-9.80) 6.05 × 10-6 MYO18B
rs12170932 22 27338123 T 0.013 21.92 (2.86-168.29) 1.07 × 10-5 -
rs2015035 22 30771554 T 0.039 12.17 (3.52-42.07) 3.10 × 10-5 CCDC157
rs6008155 22 47749241 G 0.039 9.73 (2.79-33.90) 6.32 × 10-6 LOC339685
rs11090847 22 48899419 T 0.171 3.63 (1.74-7.60) 2.92 × 10-5 FAM19A5
rs5770525 22 49930353 G 0.500 4.25 (2.10-8.61) 1.11 × 10-5 C22orf34
rs9628158 22 50255459 T 0.000 NC 2.18 × 10-6 ZBED4
ALG12 aNearest physical genes were reported. Chr: Chromosome; RA: risk allele; Freq: frequency of risk allele in controls; OR: odds ratio; CI: confidence interval; NS: not significant. NC: Not calculable due to an allele frequency
of 0 in either the cases or controls.
144 Chapter 7: Systematic Evaluation of Risk Loci from Candidate Gene and Genome-wide Association Studies of Fatigue
Table 7.3. Summary of SNPs from genome-wide association study for self-reported tiredness.a
SNP Chromosome SNP position Gene p-value
1:64178756_C_Tb 1 64178756 - 1.36 × 10-11
rs142592148 1 75842193 SLC44A5 5.88 × 10-8
rs7219015 17 2555592 PAFAH1B1 6.86 × 10-8 aRisk alleles and effect sizes were not reported. bAffymetrix ID is reported because
this SNP does not have an rs ID. The study by Deary and colleagues (2017) was
conducted in a UK population. Tiredness was assessed by the question: “Over the past two weeks, how often have you felt tired or had little energy?”. 6,948
individuals responded “nearly every day”, 6,404 individuals responded “more than
half the days”, 44,208 individuals responded “several days”, and 51,416 individuals responded “not at all”.
Table 7.4. Summary of genes from gene-based association analysis of self-reported tiredness.
Gene Chromosome Start Stop p-value
DRD2 11 113280317 113346001 2.94 × 10-7 PRRC2C 1 171454666 171562650 1.43 × 10-6
C3orf84 3 49215069 49229291 1.45 × 10-6
ANO10 3 43407818 43663560 1.52 × 10-6 ASXL3 18 31158541 31327399 2.67 × 10-6
RHOA 3 49396578 49449526 4.07 × 10-6
CTNND1 11 57529234 57586652 4.09 × 10-6 THEM4 1 151843342 151882361 5.44 × 10-6
FBXO21 12 117581585 117628300 5.66 × 10-6 ADARB1 21 46494493 46646478 6.01 × 10-6
NAPA 19 47990891 48018515 6.06 × 10-6
KANSL1L 2 210885435 211036051 6.24 × 10-6 RHCG 15 90014638 90039799 6.90 × 10-6
PLAC8 4 84011201 84035911 6.95 × 10-6
KLF7 2 207945529 208030614 7.40 × 10-6 RPE 2 210867352 210886291 1.00 × 10-5
TMX2 11 57479995 57508445 1.36 × 10-5
SNF8 17 47007458 47022154 1.38 × 10-5
CCDC36 3 49235861 49295537 1.38 × 10-5
SSBP4 19 18530146 18545372 1.87 × 10-5
ISYNA1 19 18545198 18549111 1.93 × 10-5 RELT 11 73087405 73108519 2.37 × 10-5
CSMD3 8 113235157 114449242 2.49 × 10-5
ZDHHC5 11 57435474 57468659 2.66 × 10-5 METTL16 17 2319343 2415200 2.67 × 10-5
SRRM4 12 119419300 119600856 3.03 × 10-5
BSN 3 49591922 49708982 3.20 × 10-5 NRXN1 2 50145643 51259674 3.25 × 10-5
ZNF780A 19 40575059 40596845 3.30 × 10-5
SMC1B 22 45739944 45809500 3.33 × 10-5 TCTA 3 49449639 49453909 3.36 × 10-5
GIP 17 47035918 47045955 3.45 × 10-5
CKMT1A 15 43985084 43991420 4.09 × 10-5 NICN1 3 49459766 49466757 4.18 × 10-5
UBE2Z 17 46985731 47006422 5.11 × 10-5
DAG1 3 49506146 49573048 5.26 × 10-5
ATP11B 3 182511291 182639423 5.28 × 10-5
PSMC4 19 40477073 40487353 5.44 × 10-5
FAM168A 11 73117028 73309228 5.86 × 10-5 CCNT2 2 135676392 135716915 6.25 × 10-5
OPA1 3 193310933 193415600 6.42 × 10-5
CATSPER2 15 43922772 43941039 6.52 × 10-5 ZBTB37 1 173837493 173855774 6.67 × 10-5
ELL 19 18553473 18632937 6.91 × 10-5
SERPING1 11 57365027 57382326 7.49 × 10-5 PLGRKT 9 5357966 5437937 7.89 × 10-5
PRR12 19 50094912 50129696 8.37 × 10-5
UBA7 3 49842638 49851391 8.48 × 10-5 CAMK1D 10 12391583 12871735 9.36 × 10-5
The study by Deary and colleagues (2017) was conducted in a UK
population. Tiredness was assessed by the question: “Over the past two weeks, how often have you felt tired or had little energy?”. 6,948
individuals responded “nearly every day”, 6,404 individuals responded
“more than half the days”, 44,208 individuals responded “several days”,
and 51,416 individuals responded “not at all”.
Chapter 7: Systematic Evaluation of Risk Loci from Candidate Gene and Genome-wide Association Studies of Fatigue 145
Table 7.5. Summary of SNPs from genome-wide association studies of depression phenotypes, in
Europeans.
SNP Chr SNP position RA Freq OR (95% CI) p-value Gene
Study by Hyde and colleagues (2016) in a cohort of 130,620 self-reported and clinically evaluated cases
and 347,620 controls.a
rs10514299 5 87663610 C 0.2406 1.05 (1.04-1.07) 9.99 × 10-16 -
rs1518395 2 58208074 A 0.6130 1.03 (1.02-1.05) 4.32 × 10-12 -
rs2179744 22 41621714 G 0.2817 1.04 (1.02-1.05) 6.03 × 10-11 L3MBTL2 rs11209948 1 72811904 G 0.6389 1.04 (1.02-1.05) 8.38 × 10-11 -
rs454214 5 88003403 T 0.4319 1.03 (1.02-1.05) 1.09 × 10-9 -
rs301806 1 8482078 T 0.4481 1.03 (1.02-1.04) 1.90 × 10-9 RERE rs1475120 6 105389953 G 0.5475 1.03 (1.02-1.04) 4.17 × 10-9 -
rs10786831 10 106614571 G 0.3993 1.03 (1.02-1.04) 8.11 × 10-9 SORCS3
rs12552 13 53625781 G 0.4450 1.05 (1.03-1.06) 8.16 × 10-9 OLFM4 rs6476606 9 rs6476606 G 0.3628 1.03 (1.02-1.04) 1.20 × 10-8 PAX5
rs8025231 15 37648402 A 0.4275 1.04 (1.02-1.05) 1.23 × 10-8 -
rs12065553 1 80793118 A 0.2797 1.03 (1.02-1.05) 1.32 × 10-8 -
rs1656369 3 158280085 A 0.6703 1.04 (1.02-1.05) 1.34 × 10-8 -
rs4543289 5 164484948 T 0.5199 1.03 (1.02-1.04) 1.36 × 10-8 -
rs2125716 12 84941429 G 0.2345 1.04 (1.02-1.05) 3.05 × 10-8 - rs2422321 1 rs2422321 A 0.4410 1.03 (1.02-1.04) 3.18 × 10-8 -
rs7044150 9 2982931 T 0.6155 1.03 (1.02-1.05) 4.31 × 10-8 -
Study by Power and colleagues (2017) in a cohort of 22,158 MDD and recurrent MDD cases (stratified
into age of onset) and 133,749 controls.
rs7647854 3 184876783 G 0.1600 1.16 (1.11-1.21) 5.20 × 10-11 - Study by Okbay and colleagues (2016) in a cohort of 180,866 individuals with depressive symptoms.
rs7973260 12 118375486 A 0.1900 1.03 (1.02-1.04) 1.80 × 10-9 KSR2
rs62100776 18 50754633 T 0.5600 1.03 (1.02-1.03) 8.50 × 10-9 DCC Study by Direk and colleagues (2016) in a cohort of 98,345 individuals with MDD and recurrent MDD or
depressive symptoms.
rs9825823 3 61082153 T 0.4600 NC 8.20 × 10-9 FHIT
Chr: Chromosome; RA: risk allele; Freq: frequency of risk allele in controls; OR: odds ratio; CI: confidence interval; NS: not significant. NC: Not calculable. ap-values reported are from the meta-analyses of the discovery
and replication analyses, while the OR and CI reported are with respect to the discovery cohort as the effect was
not reported for the meta-analysis
7.3.2 Study Cohorts, Genotyping Data and Quality-control
CFS Cohort
The present study was conducted using data from two community cohorts. The CFS
cohort was established at the Menzies Health Institute Queensland and consisted of
47 cases (9 males and 38 females) and 55 controls (17 males and 38 females). All
cases met the 1994 Centres for Disease Control criteria for CFS (Fukuda et al.,
1994). Participants were assessed at the National Centre for Neuroimmunology and
Emerging Diseases and provided signed consent to participate in the study, which
was approved by the Griffith University Human Research Ethics Committee.
Genotyping of the CFS cohort was conducted using the Illumina Human Omni
Express Exome-8v1_B Array. The inclusion threshold for common SNP markers
was a minimum call rate of 95%, MAF greater than 0.01, HWE p-value greater than
1 × 10-6, and a genotyping call (GC) threshold greater than 0.7. While the inclusion
threshold for exome SNP markers was a minimum call rate of 95%, a minor allele
count of three in cases and controls separately, a HWE p-value greater than 1 × 10-6,
and a GC threshold greater than 0.7. Finally, only autosomal chromosome SNP
146 Chapter 7: Systematic Evaluation of Risk Loci from Candidate Gene and Genome-wide Association Studies of Fatigue
markers were included in the analysis. After quality control the CFS genotype dataset
contained 582,261 common SNP markers and 26,957 exome SNP markers (totalling
609,218 SNP markers), of which 276,392 mapped to 16,486 genes. Given the CFS
cohort was densely genotyped for common and exome variants imputation was not
conducted as sufficient coverage was provided for replication of previous CGA and
GWA results.
Fatigue Cohort
The fatigue cohort was obtained as part of the over 50’s (aged) study conducted at
QIMR Berghofer (QIMRB). Within the study 2,281 twin pairs from the Australian
twin registry were invited to complete a mailed Health and Lifestyle Questionnaire
(Bucholz et al., 1998; Mosing et al., 2012). Informed written consent was obtained
from each participant, and the study was approved by the Human Research Ethics
Committee (HREC) of QIMRB. The fatigue classification utilised within this study
was assessed by the Schedule of Fatigue and Anergia (SOFA) (Hickie et al., 1996),
which was originally designed to identify CFS cases. Consequently, the fatigued
state identified by the SOFA should be similar to the fatigue experienced by CFS
patients. Ten questions are contained in the SOFA, however, a shorter eight item
version was included in the Health and Lifestyle Questionnaire due to two questions
being replicated within the General Health Questionnaire (GHQ) (Goldberg &
Blackwell, 1970), that was also administered to participants. Responses to the eight
SOFA and two GHQ items were used to assess fatigue within the cohort, as
previously detailed (Corfield et al., 2016a). Individuals were classified as fatigued if
they reported three or more of the ten fatigue symptoms (muscle pain at rest, post-
exertional muscle pain, post-exertional muscle fatigue, post-exertional fatigue,
hypersomnia, insomnia, poor concentration, speech problems, poor memory, and
headaches), over the past few weeks. The over 50’s study comprised 3,061
individuals with a fatigue classification.
Genotyping data was available for a subset of the over 50’s study, from a larger
genotyping project—which was conducted in numerous waves. In depth explanation
of the genotyping and quality control methods utilised have previously been detailed
(Mbarek et al., 2016; Medland et al., 2009). Briefly, standard quality-control
measures were utilised across the project (minimum call rate of 95%, MAF ≥ 0.01,
HWE p-value < 1 × 10-6, and GC threshold of 0.7) which was conducted within each
Chapter 7: Systematic Evaluation of Risk Loci from Candidate Gene and Genome-wide Association Studies of Fatigue 147
batch. Genotyping data passing QC was merged into a single dataset, which was
imputed up to the haplotype reference consortium (HRC) version r1.1 using the
Michigan Imputation server (Das et al., 2016). The best guess genotypes (reaching
the 95% threshold) for all autosomal chromosome SNP markers were used in the
study. Additionally, ancestry outliers were excluded from the study, whereby,
individuals of non-European decent as determined by principal-components analysis
were removed from the dataset if they were more than six standard deviations from
the mean European population of principle components (i.e., principle component
one and two). Resulting in a final fatigue genotype dataset of 307 cases (85 males
and 222 females) and 744 controls (181 males and 563 females), with 3,197,479 SNP
markers passing quality-control, of which 1,255,622 mapped to 14,129 genes.
7.3.3 Statistical Analysis
GWA analysis within the CFS cohort was conducted in Plink (Purcell et al., 2007).
The analysis was conducted utilising a chi-squared allelic test with one degree of
freedom assuming a log-additive model. Meanwhile, GWA analysis within the
fatigue cohort was conducted in GEMMA (Zhou & Stephens, 2012). The analysis
was conducted utilising univariate linear mixed modelling, including a genetic
relationship matrix, to account for the twin structure of the cohort. Linear mixed
modelling is a powerful technique, which prevents false positive associations by
accounting for underlying structure (namely population stratification or genetic
relatedness) within the data (Yang et al., 2014; Zhou & Stephens, 2012). In 2014,
Eu- Ahsunthornwattana and colleagues (2014) concluded linear-mixed models
effectively controlled for false positives in family-based case-control GWA analyses
after comparison with alternative approaches for analysing binary disease traits. In
particular, GEMMA was shown to robustly control genomic inflation for both binary
and quantitative traits.
Results from the GWA analyses were used to assess the association of
previously implicated SNPs within the study cohorts. Additionally, the level of
linkage disequilibrium within genomic regions of interest from the complete GWA
SNP datasets were further investigated utilising locus zoom plots (Pruim et al.,
2010). Furthermore, the results from the GWA analyses were utilised to conduct a
gene-based analysis, using MAGMA (de Leeuw et al., 2015), with a window size of
0kb (this package and settings were utilised to enable direct comparison with results
148 Chapter 7: Systematic Evaluation of Risk Loci from Candidate Gene and Genome-wide Association Studies of Fatigue
from the gene-based analysis in the UKbiobank tiredness dataset (Deary et al.,
2017)). Results from the gene-based analysis were used to assess the association of
previously implicated genes within the study cohorts.
Power calculations were utilised to determine the level of power we have
within the fatigue cohort to replicate the results of previously implicated CGA SNPs
(Purcell et al., 2003). Power calculations (Table 7.1) were conducted assuming a
multiplicative model (which is equivalent to a log-additive model—which is utilised
in Plink), with disease prevalence for fatigue of 30.7% (Corfield et al., In press), a
significance threshold of 0.05, and the 1000 Genomes Project allele frequencies
within a European population for each CGA SNP was utilised (The 1000 Genomes
Project Consortium, 2015).
7.4 RESULTS
7.4.1 Previously Implicated SNPs and Genes
CFS CGA Studies
Of the 151 previously implicated CFS CGA SNPs, 47 passed QC in the CFS dataset
and 72 were in the fatigue dataset. Two SNPs reached the Bonferroni adjusted p-
value of 0.0011 (0.05/47) in the CFS cohort (rs655207, p = 0.0006 and rs4738202, p
= 0.0009) with four additional SNPs reaching nominal significance (rs6650469, p =
0.0016; rs6429157, p = 0.0114; rs12914385, p = 0.0224; and rs951266 p = 0.0235)—
all six SNPs had effects in the same direction as previously reported (Supplementary
Table 7.2). However, all of these SNPs were previously reported within the studies
by Marshall-Gradisnik and colleagues (2015a; 2016a; 2015b) which included
samples from the same CFS cohort analysed in our study. Power calculations
indicated we have over 90% power, at a p-value threshold of 0.05, to identify all 72
of the previously implicated CFS CGA SNPs within the fatigue dataset. However, no
evidence for an association was observed, even at a p-value threshold of 0.05, for any
of the 72 SNPs within the fatigue cohort (Supplementary Table 7.2).
Meanwhile of the 39 previously implicated CFS CGA genes, 37 were included
in the CFS cohort and 35 were included in the fatigue dataset. Four genes (CHRNA5,
p = 0.0358; TRPC4, p = 0.0360; TRPC6, p = 0.0372; and TRPA1, p = 0.0466)
reached nominal significance within the CFS cohort while only one reached nominal
significance (TRPA1, p = 0.0376) in the fatigue cohort. Notably, TRPA1 reached
Chapter 7: Systematic Evaluation of Risk Loci from Candidate Gene and Genome-wide Association Studies of Fatigue 149
nominal significance in both cohorts (Supplementary Table 7.3) although the
Bonferroni adjusted p-value of 0.0013 in the CFS cohort and 0.0014 in the fatigue
cohort was not reached.
CFS GWA Studies
Of the 524 previously implicated CFS GWA SNPs, 81 passed QC in the CFS dataset
and 110 were in the fatigue dataset. Three of the SNPs within the CFS cohort
reached nominal significance (rs400322, p = 0.0326; rs197770, p = 0.0359; and
rs9200, p = 0.0369) with rs400322 and rs9200 having effects in the same direction
(Supplementary Table 7.4). However, this does not reach the Bonferroni adjusted p-
value of 0.0006 (0.05/81). Similarly, four SNPs within the fatigue dataset had p-
values less than 0.05 (rs7306948, p = 0.0024; rs6721414, p = 0.0034; rs10121299, p
= 0.0072; and rs1157185, p = 0.0405), although rs1157185 was the only SNP with
the same effect direction (Supplementary Table 7.4). However, this does not reach
the Bonferroni adjusted p-value of 0.0005 (0.05/110).
Meanwhile, of the 319 previously implicated CFS GWA genes, 252 were
included in the CFS dataset and 237 were included in the fatigue dataset. Nineteen of
the genes within our CFS cohort reached nominal significance (PLA2G4A, p =
0.0001; MOB3B, p = 0.0003; AFAP1, p = 0.0075; RASEF, p = 0.0108; CELF4, p =
0.0112; ITGA9, p = 0.0119; ARPC1B, p = 0.0132; TXNRD1, p = 0.0183; NR6A1, p =
0.0228; PTGS2, p = 0.0235; FSHR, p = 0.0244; CDH18, p = 0.0264; SLC13A5, p =
0.0267; PDE4D, p = 0.0285; DNMBP, p = 0.0287; FBXL13, p = 0.0383; SLC35F3, p
= 0.0396; FAM185A, p = 0.0452; CLEC16A, p = 0.0477) (Supplementary Table 7.5),
with PLA2G4A reaching the Bonferroni adjusted p-value of 0.0002 (0.05/252).
Similarly, seven genes within our fatigue cohort reached nominal significance
(PRUNE2; p = 0.0003; RNASEL, p = 0.0045; COLEC11, p = 0.0080; TNFRSF10D, p
= 0.0275; HS6ST3, p = 0.0281; ANK2, p = 0.0458; and ARL15, p = 0.0476)
(Supplementary Table 7.5), however, none of the genes reached the Bonferroni
adjusted p-value of 0.0002 (0.05/237).
Tiredness GWA and Gene-Based Analyses
None of the tiredness GWA SNPs were in the CFS or fatigue cohorts. Although
looking at the regions immediately surrounding the SNPs (50kb upstream and
downstream), no peak in association signal was observed in either cohort
(Supplementary Figure 7.1). However, of the 51 previously implicated tiredness
150 Chapter 7: Systematic Evaluation of Risk Loci from Candidate Gene and Genome-wide Association Studies of Fatigue
genes, 50 were included in the CFS cohort and 41 were included in the fatigue
cohort. Comparing the gene-based analysis results from the UKbiobank tiredness
association analysis with results from our cohorts revealed two of the genes within
our CFS cohort reached nominal significance (DAG1, p = 0.0269 and ZBTB37, p =
0.0412) (Supplementary Table 7.6), however, this does not reach the Bonferroni
adjusted p-value of 0.0010 (0.05/50). Similarly, four of the genes within our fatigue
cohort reached nominal significance (PLGRKT, p = 0.0027; KANSL1L, p = 0.0033;
RPE, p = 0.0054; and ZBTB37, p = 0.0457) (Supplementary Table 7.6), however, this
does not reach the Bonferroni adjusted p-value of 0.0012 (0.05/41). Notably,
ZBTB37 reached nominal significance within both cohorts, however, no evidence for
association was observed for the five genes which reached genome-wide significance
in the gene-based analysis of tiredness.
MDD GWA Studies
Of the 21 SNPs associated with MDD 11 passed QC in the CFS cohort and 12 were
in the fatigue dataset. One SNP reached nominal significance and had the same
direction of effect (rs10514299, p = 0.0264) in the CFS cohort (Supplementary Table
7.7), however this did not reach the Bonferroni adjusted significance threshold of
0.0045. Meanwhile, no evidence for association was observed within the fatigue
cohort (Supplementary Table 7.7). Additionally, all eight genes investigated were
included in the CFS and fatigue dataset. However, no evidence for association was
observed in either cohort (Supplementary Table 7.8).
7.4.2 Genome-wide association results
CFS
The GWA analysis conducted for CFS in 47 cases and 55 controls had a genomic
inflation (λ) of 0.99 (Supplementary Figure 7.2). Three SNPs were suggestively
associated (p < 1 × 10-5) with CFS (Table 7.6) in the overall GWA analysis (Figure
7.1) rs12473577 on chromosome 2 (p = 2.65 × 10-6), rs652252 on chromosome 1 (p
= 4.53 × 10-6), and rs1888140 on chromosome 13 (p = 9.66 × 10-6). Regional
association plots for these loci are shown in Supplementary Figures 7.3.
Additionally, three genes were suggestively associated (p < 1 × 10-4) with CFS
(Table 7.7) in the gene-based analysis CAPRIN1 (p = 3.17 × 10-5), EMCN (p = 4.13 ×
10-5), and CDCP2 (p = 5.34 × 10-5).
Chapter 7: Systematic Evaluation of Risk Loci from Candidate Gene and Genome-wide Association Studies of Fatigue 151
Fatigue
The GWA analysis conducted for fatigue in 307 cases and 744 controls had a
genomic inflation (λ) of 1.02 (Supplementary Figure 7.4). Fifty-seven SNPs were
suggestively associated with fatigue in the overall GWA analysis (Figure 7.2).
Although high levels of LD are observed between the SNPs (Supplementary Figures
7.5), as shown in the regional association plots for the 6 genomic locations. The top
SNP from each of the six genomic locations is rs874681 on chromosome 11 (p =
8.09 × 10-7), rs16849948 on chromosome 3 (p = 1.81 × 10-6), rs1701470 on
chromosome 10 (p = 5.20 × 10-6), rs352582 on chromosome 5 (p = 5.25 × 10-6),
rs359477 on chromosome 5 (p = 7.34 × 10-6), and rs4237354 on chromosome 10 (p =
8.89 × 10-6) (Table 7.6). Additionally, two genes were suggestively associated (p < 1
× 10-4) with fatigue (Table 7.7) in the gene-based analysis PLXDC2 (p = 5.80 × 10-6)
and TBCA (p = 6.74 × 10-6).
152 Chapter 7: Systematic Evaluation of Risk Loci from Candidate Gene and Genome-wide Association Studies of Fatigue
Figure 7.1. Manhattan plot of the chronic fatigue syndrome (CFS) cohort genome-wide association raw p-values. The horizontal dashed line corresponds to the genome-wide
significance threshold (p < 5 × 10-8). The three genes suggestively associated (p < 1 × 10-4) with CFS in gene-based analyses are indicated in green (CDCP2), pink (EMCN),
and blue (CAPRIN1).
Chapter 7: Systematic Evaluation of Risk Loci from Candidate Gene and Genome-wide Association Studies of Fatigue 153
Figure 7.2. Manhattan plot of the fatigue cohort genome-wide association raw p-values. The horizontal dashed line corresponds to the genome-wide significance threshold (p
< 5 × 10-8). The two genes suggestively associated (p < 1 × 10-4) with fatigue in gene-based analyses are indicated in pink (TBCA) and blue (PLXDC2).
154 Chapter 7: Systematic Evaluation of Risk Loci from Candidate Gene and Genome-wide Association Studies of Fatigue
Table 7.6. Summary of SNPs reaching suggestive significance thresholds for chronic fatigue syndrome and fatigue.
SNP Chr SNP position RA OA Frequency of RA OR (95% CI) p-value Variant type Gene symbol
Chronic fatigue syndrome
rs12473577 2 53516993 A G 0.373 3.97 (2.21-7.14) 2.65 × 10-6 Intergenic -
rs652252 1 238749403 C T 0.582 4.91 (2.40-10.03) 4.53 × 10-6 Intergenic - rs1888140 13 44817707 A G 0.164 4.13 (2.16-7.90) 9.66 × 10-6 Intergenic Between SMIM2 and SERP2
Fatigue
rs874681 11 20625343 T C 0.242 1.12 (1.07-1.17) 8.09 × 10-7 Intron SLC6A5 rs16849948 3 94498339 A G 0.031 1.33 (1.18-1.49) 1.81 × 10-6 Intergenic -
rs1701470 10 12154833 C T 0.706 1.11 (1.06-1.16) 5.20 × 10-6 Intron DHTKD1
rs352582 5 77033413 G A 0.631 1.10 (1.06-1.15) 5.25 × 10-6 Intron TBCA rs359477 5 173305432 C T 0.463 1.10 (1.05-1.14) 7.34 × 10-6 Intergenic -
rs4237354 10 20459237 G A 0.554 1.10 (1.05-1.14) 8.89 × 10-6 Intron PLXDC2
Chr: Chromosome; RA: risk allele; OA: other allele; OR: odds ratio; CI: confidence interval.
Table 7.7. Summary of genes reaching suggestive significance thresholds from gene-based association analysis for chronic fatigue syndrome and fatigue.
Gene Chromosome Start Stop Number of SNPsa P-value
Chronic fatigue syndrome
CAPRIN1 11 34073230 34124157 6 3.17 × 10-5
EMCN 4 101316498 101439250 24 4.13 × 10-5 CDCP2 1 54604668 54618679 13 5.34 × 10-5
Fatigue
PLXDC2 10 20105118 20575199 549 5.80 × 10-6 TBCA 5 76986995 77072185 114 6.74 × 10-6 aNumber of SNPs found within the start and stop sites.
Chapter 7: Systematic Evaluation of Risk Loci from Candidate Gene and Genome-wide Association Studies of Fatigue 155
7.5 DISCUSSION
To date, CGA and GWA studies investigating CFS have been conducted in very
small cohorts. The CFS cohort within the present study is similar in size to the
previously conducted GWA studies. However, the fatigue cohort was larger than any
of the previously published CFS association studies. Considering the self-report
questionnaire utilised within the fatigue cohort was originally designed to identify
CFS cases, the fatigue symptoms assessed are characteristic of those experienced by
individuals with CFS. Therefore, the fatigue cohort utilised was ideal to investigate if
less severe fatigue phenotypes are associated with similar genetic contributions as
CFS.
The inability to replicate the majority of previous association studies results
within our CFS or fatigue cohort indicates they are likely false positives. Although,
the CGA analysis of DISC1 was conducted in a Japanese cohort. Results from the
present study indicate DISC1 is not associated with fatigue in Europeans. However,
the genetic contribution of fatigue may be population specific, therefore, to
determine if the DISC1 association is a true finding the result needs to be replicated
in a Japanese cohort.
Schlauch and colleagues (2016) stated that 28 of the SNPs suggestively
associated with CFS by Smith and colleagues (2011) were contained in the
genotyping data of their cohort. Of these 28 SNPs, Schlauch et al (2016) indicated
rs10509412 is associated with CFS in their cohort. However, this SNP is not listed in
their Supplementary material (Supplementary Table 7.1) which included all
autosomal SNPs associated with CFS at a threshold of p < 3.3 × 10-5. No evidence of
association was observed with rs10509412 in our CFS cohort (p = 0.0870).
Of the remaining previously implicated SNPs and genes, the only evidence for
replication was observed for rs655207 located in an intron within TRPC4 (p =
0.0006), rs4738202 located in an intron within TRPA1 (p = 0.0009), and PLA2G4A
(p = 0.0001) in the CFS cohort. All three genes are associated with the
immunological system, providing support for the immune system playing a key role
in the pathophysiology of CFS. However, the majority of candidate gene studies
have focused on genes associated with the immune system; therefore it is not
possible to determine if genes related to other systems in the human body are
156 Chapter 7: Systematic Evaluation of Risk Loci from Candidate Gene and Genome-wide Association Studies of Fatigue
associated with CFS from this analysis. Additionally, both SNPs which were
replicated in the CFS cohort were identified by Marshall-Gradisnik and colleagues
(2015a; 2016b; 2015b) and it is unknown if these samples are independent from our
CFS cohort. Therefore, further investigation in larger cohorts are necessary to
determine if these genes contribute to the development of a CFS and other fatigue
phenotypes. Particularly considering the results from the UKbiobank tiredness GWA
and gene-based analysis (which were vastly more powerful) were not replicated in
either cohort. Although the fatigue severity investigated in the current study was
higher—indicating different genes may be associated with varying severities to the
fatigue continuum further follow-up of the novel tiredness associations is warranted.
Results from the GWA and gene-based analyses in our CFS cohort identified
three suggestively associated SNPs (rs12473577, rs652252, and rs1888140) and
three suggestively associated genes (CAPRIN1, EMCN, and CDCP2). However,
considering the small sample size of this cohort and that evidence for an association
was not observed in the fatigue cohort these results require replication. Similarly,
results from the GWA and gene-based analysis in our fatigue cohort identified six
genomic regions of interest (with the top SNPs (rs874681, rs16849948, rs1701470,
rs352582, rs359477, and rs4237354) and two suggestively associated genes
(PLXDC2 and TBCA). Considering the larger cohort size, the SNPs and genes
suggestively associated with fatigue are less likely to be false positives. Furthermore,
considering the clusters of SNPs suggestively associated with fatigue on
chromosome 3, chromosome 5, chromosome 10, and chromosome 11 indicates these
regions warrant further investigation in larger fatigue cohorts.
The genes which are implicated from these genomic locations are SLC6A5,
DHTKD1, TBCA, and PLXDC2. SLC6A5 (solute carrier family 6 member 5), which
encodes a neurotransmitter transporter. Functional characterisation of SLC6A5
indicates the gene plays a crucial role in extracellular glycine clearance during
glycine-mediated neurotransmission. Considering changes in brain morphology and
decreased basal ganglia activation have been implicated in the pathophysiology of
CFS (Barnden et al., 2015; Miller et al., 2014; Tang et al., 2015), this region warrants
further investigation to determine the functional implications and importance of
SLC6A5 involvement in fatigue. DHTKD1 (dehydrogenase E1 and transketolase
domain containing 1), which encodes a mitochondrial protein. Functional
Chapter 7: Systematic Evaluation of Risk Loci from Candidate Gene and Genome-wide Association Studies of Fatigue 157
characterisation of DHTKD1 indicates the gene plays a crucial role in mitochondrial
biogenesis and reactive oxygen species degradation (Xu et al., 2013). Considering
mitochondrial dysfunction has previously been implicated in the pathophysiology of
CFS (Booth et al., 2012; Gorman et al., 2015; Meeus et al., 2013; Myhill et al.,
2013), this region warrants further investigation to determine the functional
implications and importance of DHTKD1 genetic variants in fatigue. TBCA (tublin
folding cofactor A), encodes a protein involved in the correct folding of beta-tublin
and is essential for cell viability (Nolasco et al., 2005; Tian et al., 1996). The
metabolism of proteins is the main pathway related to TBCA. Therefore, there are
numerous roles TBCA could play in the development or maintenance of a fatigued
state. Finally, PLXDC2 (plexin domain containing 2) which is a transmembrane
protein that acts as a cell-surface receptor for the pigment epithelium-derived factor
protein (Cheng et al., 2014). However, little else is known about the functionality of
PLXDC2. Therefore, we are unable to speculate about the functional role of PLXD2
in fatigue.
To our knowledge, this is the first study investigating the molecular genetics of
fatigue. Although, Deary and colleagues (2017) investigated the molecular genetics
of self-reported tiredness in the UKbiobank, our fatigue phenotype is more severe
and the physical symptoms reported are those required for diagnosis with CFS. The
increased phenotypic similarity between our fatigue cohort and individuals with CFS
indicates the underlying genetics are likely comparable. Therefore, enabling us to
replicate our CFS analysis in a larger sample and address the main limitation of this
study, which was the small CFS cohort sample size. Although, the CFS cohort
sample size was comparable to previously published CFS GWA studies and our main
aim was to replicate previous findings. The use of twin data, in the fatigue cohort,
could be considered a limitation. However, Minică and colleagues (2014) have
shown that the type I error rate is not affected by the inclusion of MZ twin pairs
within GWA studies. Finally, our analyses within the fatigue cohort were restricted
to individuals of European origin because the generalizability of the study was
underpowered.
In summary, results presented in the present study indicate that the majority of
previously reported nominally and genome-wide significant SNPs and genes from
CGA and GWA studies of CFS are likely false positives. Similar studies
158 Chapter 7: Systematic Evaluation of Risk Loci from Candidate Gene and Genome-wide Association Studies of Fatigue
investigating the genetic association findings of other common, complex, traits have
comparable results (de Vries et al., 2015; Hirschhorn et al., 2002). However, our
study has identified three SNPs and three genes which warrant further investigation
in larger CFS and fatigue samples. Additionally, six genomic locations of interest
which are potentially associated with fatigue were identified. Future investigations
into the genetic contribution of fatigue and CFS should concentrate on utilising
larger, well characterised cohorts with stringent quality control, to advance our
knowledge of the underlying biology of these traits, while minimising the publication
of false positive association results.
Chapter 8: General Discussion 159
Chapter 8: General Discussion
The findings detailed within this dissertation, evaluated the comorbidity and genetics
of fatigue and depression. Within this chapter, a summary of results will be
presented, followed by an explanation of the strengths and limitations of the analyses
conducted within this thesis, before the future directions and conclusions are
outlined.
8.1 SUMMARY OF FINDINGS
Initially, a phenotypic analysis was conducted, which involved a symptomatic
analysis of fatigue and depression. In Chapter 3 the prevalence and risk of co-
occurring fatigue and depression were reported, in addition to the results of the
symptomatic analysis. Within the study cohort, 6.7% of individuals were fatigued
and depressed. The key finding identified within Chapter 3 was that the overlapping
symptoms of fatigue and depression do not facilitate the association between the
traits. Specifically, a significantly increased risk of depression in fatigued
individuals, compared to non-fatigued individuals and the total population, was
observed independently of the overlapping symptoms. Furthermore, a significantly
increased risk of fatigue in depressed individuals, compared to non-depressed
individuals and the total population was observed independently of the overlapping
symptoms. To our knowledge, this is the first time the risk of co-occurring fatigue
and depression has been assessed independently of their overlapping symptoms.
In Chapter 4 and 5 the familiality and heritability of fatigue and depression
were investigated. In Chapter 4, a familial contribution to fatigue was identified with
a significant additive genetic contribution of 40%. Importantly, sex-specific effects
were not identified. Comparison of these findings with previous studies (Schur et al.,
2007; Sullivan et al., 2005), led to the conclusion that the etiology and heritability of
fatigue may vary across the lifespan. Similarly, within Chapter 5 the heritability of a
broad depression phenotype (major or minor depression) was higher in older adults
(aged 50-92) at 48% compared 40% in younger adults (aged 23-38)—indicating the
etiology and heritability of depression may also vary across the lifespan. This finding
is substantiated by Power and colleagues (2017) recent finding that rs7647854 is
160 Chapter 8: General Discussion
associated with MDD onset in adults aged 27 or older. Further investigation within
Chapter 5 revealed minor and major depression lie on a genetic continuum, with
minor depression having an additive genetic contribution of 37%, in older adults. To
our knowledge, this is the first investigation into the heritability of minor depression.
Although, a genetic continuum between MDD and depressive symptoms or
subthreshold depression phenotypes (such as MiDD or dysthymia) has been
suggested (Ayuso-Mateos et al., 2010; Lobo & Agius, 2012). This is the first study
that substantiates this hypothesis, with Direk and colleagues (2016), identification of
a risk locus associated with a broad depression phenotype (of MDD and depression
symptoms) providing further evidence supporting the clinical relevance of
investigating a broad depression phenotype.
The results of Chapters 4 and 5 were used as the foundations for the genetic
relationship analysis conducted in Chapter 6. A significant additive genetic
correlation of 0.71 was identified between depression and fatigue. The significance
of additive genetic factors common to both depression and fatigue remained
significant independently of the traits overlapping symptoms. Previous studies
provided some evidence that additive genetic factors were common to both
depression and fatigue (Fowler et al., 2006; Hur et al., 2012). However, the role of
the traits overlapping symptoms had never been investigated. Our results provide
further support to the findings of Chapter 3, which indicated the association between
fatigue and depression is independent of the traits overlapping symptoms. Rather, the
association is likely attributable to the non-causal genetic model of inheritance.
Finally, within Chapter 7 the molecular genetics of fatigue and CFS were
investigated. Four genomic regions of interest, potentially associated with fatigue
were identified, which are located downstream of C5orf38 on chromosome 5, within
DHTKD1 on chromosome 10, within SLC6A5 on chromosome 11, and between
TSHZ2 and ZNF217 on chromosome 20. Meanwhile, the inability to replicate
previous CFS CGA and in particular GWA results highlights the need for increased
cohort sizes, which are deeply phenotyped, and collaboration between researchers to
facilitate the identification of robustly associated risk loci for CFS and prevent the
publication of questionable and inaccurate results.
Chapter 8: General Discussion 161
8.2 LIMITATIONS
The use of twin data throughout this dissertation can be viewed as both an advantage
and disadvantage. Potential confounding from health-care seeking behaviour was
removed through the use of a community-based cohort. Although the use of self-
report rather than interview-based fatigue and in particular depression introduces the
potential for misclassification, thereby, reducing power throughout the studies.
However, the criteria for prolonged fatigue and CF classification only requires self-
reported fatigue. Therefore, clinical fatigue data experienced over a comparable
timeframe is not available. Meanwhile, the classification of minor and major
depression was conducted based on the DSM major depressive episode criteria which
are used to establish a clinical diagnosis of MDD. Furthermore, the prevalence of
minor and major depression within the study cohorts utilised throughout this
dissertation were consistent with previously published cohorts with similar age
ranges (Centers for Disease Control and Prevention, 2010).
The analyses conducted within Chapter 4, 5, and 6 would not have been possible
without the utilisation of twin data. The main limitation of twin modelling is that the
data may not be representative of the overall population. However, utilisation of twin
data has numerous advantages including the ability to estimate the relative
contribution of genetic and environmental factors to the variation of a trait. The
calculation of heritability estimates is based on the assumption that MZ twin pairs
share 100% of their genes while DZ twin pairs only share 50%. Although SNP-based
heritability and genetic correlation estimates can be calculated, using LD score
regression (Bulik-Sullivan et al., 2015a; Bulik-Sullivan et al., 2015b), the small
number of individuals with genotyping data (307 cases and 744 controls) was not
powerful enough for these analyses to be conducted within this dissertation. The
limitation of the small fatigue and CFS cohorts was observed within Chapter 7 by the
inability to identify genetic risk loci significantly associated with either phenotype.
Although identification of genomic regions of interest was possible within the fatigue
cohort that warrant further investigation, the small sample size prevented
investigation of gene-gene and gene-environment interactions. Furthermore, the
contribution of genetic pleiotropy (where a single gene confers an increased risk to
multiple traits) was unable to be investigated for fatigue and depression within this
dissertation due to the small sample sizes. However, adequately sized cohorts which
162 Chapter 8: General Discussion
enable the identification of robustly associated gene-gene and gene-environment
interactions are just starting to be published for complex traits. Therefore, the
identification of gene-gene and gene-environment interactions and genetic pleiotropy
should be viewed as possible avenues for future investigation for fatigue and
depression, rather than limitations.
8.3 FUTURE DIRECTIONS
This dissertation increases our understanding of the comorbidity and genetics of
fatigue and depression. However, the results detailed and technological advances
lead to numerous possible avenues for future investigations. Possible analyses to
investigate the molecular genetic overlap between fatigue and depression include the
utilisation of summary statistics from GWA analyses conducted in the psychiatric
genetic consortium, 23&Me, and UK Biobank datasets to determine if polygenic risk
scores for MDD, self-reported major depression, and self-reported tiredness,
respectively, can predict fatigue.
Building on recent findings by Power et al (2017) and Pearson et al (2016),
additional avenues for future investigation within large population cohorts include
the utilisation of additional phenotypic data within genetic analyses to reduce
heterogeneity. Based on the results of Chapter 4 and previous studies we
hypothesised the etiology of fatigue may differ with age. One possible avenue for
future investigation is to conduct a GWA analysis, based on the method utilised by
Power and colleagues (2017), to investigate the genetics of fatigue across the
lifespan. This avenue of investigation could facilitate the elucidation of the molecular
mechanisms of fatigue that differ with age. Furthermore, the investigation should
consider the possibility of genetic effects that are specific to males or females, within
different age groups. Results from these analyses could then be utilised to investigate
changes in SNP-based heritability estimates and SNP effects with age (and sex) for
fatigue. Similar methods could be utilised to investigate potential changes in SNP-
based heritability estimates and SNP effects with age and sex for depression.
The investigation into the SNP-based heritability of depression symptoms
clusters conducted by Pearson et al (2016), can be expanded to identify genetic risk
loci associated with specific depression symptom clusters in larger GWA cohorts.
Similarly, the analysis could be replicated within large fatigue cohorts to determine if
Chapter 8: General Discussion 163
the SNP-based heritability varies within specific fatigue symptom clusters and
identify genetic risk loci associated with the varying symptom domains.
Additionally, the analysis could be expanded to investigate the comorbidity of
fatigue and depression by utilising key distinguishing symptoms to select more
homogeneous subgroups, as suggested within Chapter 3. The molecular mechanisms
underlying the comorbidity could be further investigated through the use of fatigue
symptom clustering within a GWA and SNP-based heritability analysis of
depression. Considering we determined within Chapter 3 and 6 that overlapping
symptoms are not driving the comorbidity between fatigue and depression replicating
this analysis by investigating the depression symptom clustering within GWA and
SNP-based heritability analysis of fatigue may provide insight into the functional
biology associated with varying phenotype presentations. Importantly, minor
depression cases should be utilised within these analyses given the results detailed in
Chapter 5 revealed minor and major depression exist on a single genetic continuum.
Finally, within Chapter 7 we emphasised the importance of utilising large,
well-characterised cohorts with stringent quality control measures. Recent advances
in statistical methods have produced the development of a genome-wide association
study by proxy (GWAX) method, whereby first-degree relatives of cases are used (as
proxy cases) in randomly ascertained cohorts (Liu et al., 2017). Considering the low
prevalence of CFS, the GWAX approach may provide an effective strategy to gain
power by increasing study cohort size, thereby enabling identification and replication
of genetic risk loci robustly associated with the trait.
8.4 CONCLUSIONS
In conclusion, the high levels of comorbidity observed between fatigue and
depression are independent of the traits overlapping symptoms and is likely
attributable to the non-causal genetic relationship which exists between the traits.
Whereby, a significant proportion of the comorbidity between fatigue and depression
is explained by shared genetic factors.
164 Bibliography
Bibliography
Ablashi, D. V., Eastman, H. B., Owen, C. B., Roman, M. M., Friedman, J.,
Zabriskie, J. B., . . . Whitman, J. E. (2000). Frequent HHV-6 reactivation in
multiple sclerosis (MS) and chronic fatigue syndrome (CFS) patients. Journal
of Clinical Virology, 16(3), 179-191. doi:10.1016/S1386-6532(99)00079-7
Addington, A. M., Gallo, J. J., Ford, D. E., & Eaton, W. W. (2001). Epidemiology of
unexplained fatigue and major depression in the community: The Baltimore
ECA follow-up, 1981-1994. Psychological Medicine, 31(6), 1037-1044.
doi:10.1017/S0033291701004214
Afari, N., & Buchwald, D. (2003). Chronic fatigue syndrome: A review. American
Journal of Psychiatry, 160(2), 221-236. doi:10.1176/appi.ajp.160.2.221
Akaike, H. (1973). Information theory and an extension of the maximum likelihood
principle Paper presented at the Second International Symposium on
Information Theory, Tsahkadsor, Armenia, USSR.
Akaike, H. (1974). A new look at the statistical model identification. IEEE
Transactions on Automatic Control, 19(6), 716-723.
doi:10.1109/TAC.1974.1100705
Albright, F., Light, K., Light, A., Bateman, L., & Cannon-Albright, L. A. (2011).
Evidence for a heritable predisposition to chronic fatigue syndrome. BMC
Neurology, 11, 62-67. doi:10.1186/1471-2377-11-62
American Psychiatric Association. (2000). Diagnostic and statistical manual of
mental disorders: DSM-IV-TR (Vol. 4th). Washington, DC: American
Psychiatric Association.
American Psychiatric Association. (2013). Diagnostic and statistical manual of
mental disorders, fifth edition (DSM-V) (Vol. 5th). Washington, D.C:
American Psychiatric Publishing.
Anguelova, M., Benkelfat, C., & Turecki, G. (2003). A systematic review of
association studies investigating genes coding for serotonin receptors and the
serotonin transporter: I. Affective disorders. Molecular Psychiatry, 8(6), 574-
591. doi:10.1038/sj.mp.4001328
Aragam, N., Wang, K. S., & Pan, Y. (2011). Genome-wide association analysis of
gender differences in major depressive disorder in the Netherlands NESDA
and NTR population-based samples. Journal of Affective Disorders, 133(3),
516-521. doi:10.1016/j.jad.2011.04.054
Arashima, Y., Kato, K., Komiya, T., Kumasaka, K., Matsukawa, Y., Murakami, M., .
. . Arakawa, Y. (2004). Improvement of chronic nonspecific symptoms by
long-term minocycline treatment in Japanese patients with Coxiella burnetii
infection considered to have post-Q fever fatigue syndrome. Internal
Medicine, 43(1), 49-54. doi:10.2169/internalmedicine.43.49
Armstrong, C. W., McGregor, N. R., Butt, H. L., & Gooley, P. R. (2014).
Metabolism in chronic fatigue syndrome. Advances in Clinical Chemistry, 66,
121-172. doi:10.1016/B978-0-12-801401-1.00005-0
Bibliography 165
Arnold, L. M. (2008). Understanding fatigue in major depressive disorder and other
medical disorders. Psychosomatics, 49(3), 185-190.
doi:10.1176/appi.psy.49.3.185
Ayuso-Mateos, J. L., Nuevo, R., Verdes, E., Naidoo, N., & Chatterji, S. (2010). From
depressive symptoms to depressive disorders: The relevance of thresholds.
British Journal of Psychiatry, 196(5), 365-371.
doi:10.1192/bjp.bp.109.071191
Bakken, I., Tveito, K., Gunnes, N., Ghaderi, S., Stoltenberg, C., Trogstad, L., . . .
Magnus, P. (2014). Two age peaks in the incidence of chronic fatigue
syndrome/myalgic encephalomyelitis: A population-based registry study
from Norway 2008 inverted question mark2012. BMC Medicine, 12(1), 167-
173. doi:10.1186/preaccept-1843684980131041
Ball, H. A., Sumathipala, A., Siribaddana, S. H., Kovas, Y., Glozier, N., McGuffin,
P., & Hotopf, M. (2009). Genetic and environmental contributions to
depression in Sri Lanka. British Journal of Psychiatry, 195(6), 504-509.
doi:10.1192/bjp.bp.109.063529
Ball, H. A., Siribaddana, S. H., Kovas, Y., Glozier, N., McGuffin, P., Sumathipala,
A., & Hotopf, M. (2010a). Epidemiology and symptomatology of depression
in Sri Lanka: A cross-sectional population-based survey in Colombo District.
Journal of Affective Disorders, 123(1-3), 188-196.
doi:10.1016/j.jad.2009.08.014
Ball, H. A., Sumathipala, A., Siribaddana, S. H., Kovas, Y., Glozier, N., McGuffin,
P., & Hotopf, M. (2010b). Aetiology of fatigue in Sri Lanka and its overlap
with depression. British Journal of Psychiatry, 197(2), 106-113.
doi:10.1192/bjp.bp.109.069674
Ball, H. A., Siribaddana, S. H., Sumathipala, A., Kovas, Y., Glozier, N., Rijsdijk, F.,
. . . Hotopf, M. (2011). Genetic and environmental contributions to the
overlap between psychological, fatigue and somatic symptoms: A twin study
in Sri Lanka. Twin Res Hum Genet, 14(1), 53-63. doi:10.1375/twin.14.1.53
Barnden, L. R., Crouch, B., Kwiatek, R., Burnet, R., Mernone, A., Chryssidis, S., . . .
Del Fante, P. (2011). A brain MRI study of chronic fatigue syndrome:
Evidence of brainstem dysfunction and altered homeostasis. NMR in
Biomedicine, 24(10), 1302-1312. doi:10.1002/nbm.1692
Barnden, L. R., Crouch, B., Kwiatek, R., Burnet, R., & Del Fante, P. (2015).
Evidence in chronic fatigue syndrome for severity-dependent upregulation of
prefrontal myelination that is independent of anxiety and depression. NMR in
Biomedicine, 28(3), 404-413. doi:10.1002/nbm.3261
Bebbington, P. E., Dunn, G., Jenkins, R., Lewis, G., Brugha, T., Farrell, M., &
Meltzer, H. (1998). The influence of age and sex on the prevalence of
depressive conditions: Report from the National Survey of Psychiatric
Morbidity. Psychological Medicine, 28(1), 9-19.
Bedford, A., & Deary, I. J. (1997). The personal disturbance scale (DSSI/sAD):
Development, use and structure. Personality and Individual Differences,
22(4), 493-510. doi:10.1016/s0191-8869(96)00242-5
Bell, K. M., Cookfair, D., Bell, D. S., Reese, P., & Cooper, L. (1991). Risk factors
associated with chronic fatigue syndrome in a cluster of pediatric cases.
Reviews of Infectious Diseases, 13 Suppl 1, S32-S38.
Bennett, B., Goldstein, D., Lloyd, A., Davenport, T., & Hickie, I. (2004). Fatigue and
psychological distress-exploring the relationship in women treated for breast
166 Bibliography
cancer. European Journal of Cancer, 40(11), 1689-1695.
doi:10.1016/j.ejca.2004.03.021
Beqaj, S. H., Lerner, A. M., & Fitzgerald, J. T. (2008). Immunoassay with
cytomegalovirus early antigens from gene products p52 and CM2 (UL44 and
UL57) detects active infection in patients with chronic fatigue syndrome.
Journal of Clinical Pathology, 61(5), 623-626. doi:10.1136/jcp.2007.050633
Bierl, C., Nisenbaum, R., Hoaglin, D. C., Randall, B., Jones, A. B., Unger, E. R., &
Reeves, W. C. (2004). Regional distribution of fatiguing illnesses in the
United States: A pilot study. Popul Health Metr, 2(1), 1-7. doi:10.1186/1478-
7954-2-1
Bierut, L. J., Heath, A. C., Bucholz, K. K., Dinwiddie, S. H., Madden, P. A. F.,
Statham, D. J., . . . Martin, N. G. (1999). Major depressive disorder in a
community-based twin sample: Are there different genetic and environmental
contributions for men and women? Archives of General Psychiatry, 56(6),
557-563. doi:10.1001/archpsyc.56.6.557
Bifulco, A., Moran, P. M., Baines, R., Bunn, A., & Stanford, K. (2002). Exploring
psychological abuse in childhood: II. Association with other abuse and adult
clinical depression. Bulletin of the Menninger Clinic, 66(3), 241-258.
Bijl, R. V., de Graaf, R., Ravelli, A., Smit, F., & Vollebergh, W. A. M. (2002).
Gender and age-specific first incidence of DSM-III-R psychiatric disorders in
the general population. Social Psychiatry and Psychiatric Epidemiology,
37(8), 372-379. doi:10.1007/s00127-002-0566-3
Billing-Ross, P., Germain, A., Ye, K., Keinan, A., Gu, Z., & Hanson, M. R. (2016).
Mitochondrial DNA variants correlate with symptoms in myalgic
encephalomyelitis/chronic fatigue syndrome. Journal of Translational
Medicine, 14(1), 19-30. doi:10.1186/s12967-016-0771-6
Boker, S., Neale, M., Maes, H., Wilde, M., Spiegel, M., Brick, T., . . . Fox, J. (2011).
OpenMx: An open source extended structural equation modeling framework.
Psychometrika, 76(2), 306-317. doi:10.1007/s11336-010-9200-6
Bond, P. A. (1993). A role for herpes simplex virus in the aetiology of chronic
fatigue syndrome and related disorders. Medical Hypotheses, 40(5), 301-308.
doi:10.1016/0306-9877(93)90010-N
Booth, N. E., Myhill, S., & McLaren-Howard, J. (2012). Mitochondrial dysfunction
and the pathophysiology of myalgic encephalomyelitis/chronic fatigue
syndrome (ME/CFS). International Journal of Clinical and Experimental
Medicine, 5(3), 208-220.
Bosker, F. J., Hartman, C. A., Nolte, I. M., Prins, B. P., Terpstra, P., Posthuma, D., . .
. Nolen, W. A. (2011). Poor replication of candidate genes for major
depressive disorder using genome-wide association data. Molecular
Psychiatry, 16(5), 516-532. doi:10.1038/mp.2010.38
Bou-Holaigah, I., Rowe, P. C., Kan, J., & Calkins, H. (1995). The relationship
between neurally-mediated hypotension and the chronic fatigue syndrome.
Jama-Journal of the American Medical Association, 274(12), 961-967.
doi:10.1001/jama.274.12.961
Brown, A. A., Jason, L. A., Evans, M. A., & Flores, S. (2013). Contrasting case
definitions: The ME international consensus criteria vs. The Fukuda et al.
CFS criteria. North American Journal of Psychology, 15(1), 103-120.
Brown, B. I. (2014). Chronic fatigue syndrome: A personalized integrative medicine
approach. Alternative Therapies in Health and Medicine, 20(1), 29-40.
Bibliography 167
Bucholz, K. K., Heath, A. C., Madden, P. A. F., Slutske, W. S., Statham, D. J.,
Dunne, M. P., & Martin, N. G. (1998). Drinking in an older population:
Cross-sectional and longitudinal data from the Australian twin registry. In E.
L. Gomberg, A. M. Hegedus, & R. A. Zucker (Eds.), Alcohol problems and
aging (pp. 41-62). Bethesda, MD: National Institutes of Health.
Buchwald, D., Herrell, R., Ashton, S., Belcourt, M., Schmaling, K., Sullivan, P., . . .
Goldberg, J. (2001). A twin study of chronic fatigue. Psychosomatic
Medicine, 63(6), 936-943. doi:10.1097/00006842-200111000-00012
Bulik-Sullivan, B., Finucane, H. K., Anttila, V., Gusev, A., Day, F. R., Loh, P. R., . .
. Neale, B. M. (2015a). An atlas of genetic correlations across human diseases
and traits. Nature Genetics, 47(11), 1236-1241. doi:10.1038/ng.3406
Bulik-Sullivan, B. K., Loh, P. R., Finucane, H. K., Ripke, S., Yang, J., Schizophrenia
Working Group of the Psychiatric Genomics Consortium, . . . Neale, B. M.
(2015b). LD Score regression distinguishes confounding from polygenicity in
genome-wide association studies. Nature Genetics, 47(3), 291-295.
doi:10.1038/ng.3211
Burri, A., Ogata, S., Livshits, G., & Williams, F. (2015). The association between
chronic widespread musculoskeletal pain, depression and fatigue is
genetically mediated. PloS One, 10(11), e0140289.
doi:10.1371/journal.pone.0140289
Canadian Laboratory Center for Disease Control (LCDC). (1993). Report of the
working group on the possible relationship between hepatitis B vaccination
and the chronic fatigue syndrome. CMAJ: Canadian Medical Association
Journal, 149(3), 314-319.
Capelli, E., Lorusso, L., Ghitti, M., Venturini, L., Cusa, C., & Ricevuti, G. (2015).
Chronic fatigue syndrome: Features of a population of patients from northern
Italy. International Journal of Immunopathology and Pharmacology, 28(1),
53-59. doi:10.1177/0394632015572074
Carlo-Stella, N., Badulli, C., De Silvestri, A., Bazzichi, L., Martinetti, M., Lorusso,
L., . . . Cuccia, M. (2006). A first study of cytokine genomic polymorphisms
in CFS: Positive association of TNF-857 and IFNgamma 874 rare alleles.
Clinical and Experimental Rheumatology, 24(2), 179-182.
Carlo-Stella, N., Bozzini, S., De Silvestri, A., Sbarsi, I., Pizzochero, C., Lorusso, L., .
. . Cuccia, M. (2009). Molecular study of receptor for advanced glycation
endproduct gene promoter and identification of specific HLA haplotypes
possibly involved in chronic fatigue syndrome. International Journal of
Immunopathology and Pharmacology, 22(3), 745-754.
doi:10.1177/039463200902200320
Carruthers, B. M., Jain, A. K., De Meirleir, K., Peterson, D., Klimas, N., Lerner, M.,
. . . Van de Sande, M. I. (2003). Myalgic encephalomyelitis/chronic fatigue
syndrome: Clinical working case definition, diagnostic and treatment
protocols. Journal of Chronic Fatigue Syndrome, 11(1), 7-36.
Carruthers, B. M., van de Sande, M. I., De Meirleir, K. L., Klimas, N. G., Broderick,
G., Mitchell, T., . . . Stevens, S. (2011). Myalgic encephalomyelitis:
International consensus criteria. Journal of Internal Medicine, 270(4), 327-
338. doi:10.1111/j.1365-2796.2011.02428.x
Castro-Marrero, J., Cordero, M. D., Saez-Francas, N., Jimenez-Gutierrez, C.,
Aguilar-Montilla, F. J., Aliste, L., & Alegre-Martin, J. (2013). Could
mitochondrial dysfunction be a differentiating marker between chronic
168 Bibliography
fatigue syndrome and fibromyalgia? Antioxid Redox Signal, 19(15), 1855-
1860. doi:10.1089/ars.2013.5346
Cathébras, P. J., Robbins, J. M., Kirmayer, L. J., & Hayton, B. C. (1992). Fatigue in
primary care: Prevalence, psychiatric comorbidity, illness behavior, and
outcome. Journal of General Internal Medicine, 7(3), 276-286.
doi:10.1007/BF02598083
Cella, M., White, P. D., Sharpe, M., & Chalder, T. (2013). Cognitions, behaviours
and co-morbid psychiatric diagnoses in patients with chronic fatigue
syndrome. Psychological Medicine, 43(2), 375-380.
doi:10.1017/s0033291712000979
Centers for Disease Control and Prevention. (2010). Current depression among
adults - United States, 2006 and 2008. Morbidity and Mortality Weekly
Report, 59(38), 1229-1235.
Chalder, T., Berelowitz, G., Pawlikowska, T., Watts, L., Wessely, S., Wright, D., &
Wallace, E. P. (1993). Development of a fatigue scale. Journal of
Psychosomatic Research, 37(2), 147-153. doi:10.1016/0022-3999(93)90081-
P
Chapenko, S., Krumina, A., Kozireva, S., Nora, Z., Sultanova, A., Viksna, L., &
Murovska, M. (2006). Activation of human herpesviruses 6 and 7 in patients
with chronic fatigue syndrome. Journal of Clinical Virology, 37 Suppl 1,
S47-S51. doi:10.1016/s1386-6532(06)70011-7
Chen, L. S., Eaton, W. W., Gallo, J. J., Nestadt, G., & Crum, R. M. (2000). Empirical
examination of current depression categories in a population-based study:
Symptoms, course, and risk factors. American Journal of Psychiatry, 157(4),
573-580. doi:10.1176/appi.ajp.157.4.573
Chen, M. K. (1986). The epidemiology of self-perceived fatigue among adults.
Preventive Medicine, 15(1), 74-81. doi:10.1016/0091-7435(86)90037-X
Cheng, G., Zhong, M., Kawaguchi, R., Kassai, M., Al-Ubaidi, M., Deng, J., . . . Sun,
H. (2014). Identification of PLXDC1 and PLXDC2 as the transmembrane
receptors for the multifunctional factor PEDF. Elife, 3, e05401.
doi:10.7554/eLife.05401
Cheverud, J. M. (2001). A simple correction for multiple comparisons in interval
mapping genome scans. Heredity, 87, 52-58. doi:10.1046/j.1365-
2540.2001.00901.x
Chia, J. K., & Chia, L. Y. (1999). Chronic Chlamydia pneumoniae infection: A
treatable cause of chronic fatigue syndrome. Clinical Infectious Diseases,
29(2), 452-453. doi:10.1086/520239
Chia, J. K., & Chia, A. (2003). Diverse etiologies for chronic fatigue syndrome.
Clinical Infectious Diseases, 36(5), 671-672. doi:10.1086/367666
Chia, J. K. (2005). The role of enterovirus in chronic fatigue syndrome. Journal of
Clinical Pathology, 58(11), 1126-1132. doi:10.1136/jcp.2004.020255
Cho, H. J., Menezes, P. R., Hotopf, M., Bhugra, D., & Wessely, S. (2009).
Comparative epidemiology of chronic fatigue syndrome in Brazilian and
British primary care: Prevalence and recognition. British Journal of
Psychiatry, 194(2), 117-122. doi:10.1192/bjp.bp.108.051813
Choppa, P. C., Vojdani, A., Tagle, C., Andrin, R., & Magtoto, L. (1998). Multiplex
PCR for the detection of Mycoplasma fermentans, M. hominis and M.
penetrans in cell cultures and blood samples of patients with chronic fatigue
syndrome. Molecular and Cellular Probes, 12(5), 301-308.
doi:10.1006/mcpr.1998.0186
Bibliography 169
Chung, K. F., Yu, Y. M., & Yeung, W. F. (2015). Correlates of residual fatigue in
patients with major depressive disorder: The role of psychotropic medication.
Journal of Affective Disorders, 186, 192-197. doi:10.1016/j.jad.2015.07.026
Chung, Y., Lee, S. Y., Elston, R. C., & Park, T. (2007). Odds ratio based multifactor-
dimensionality reduction method for detecting gene-gene interactions.
Bioinformatics, 23(1), 71-76. doi:10.1093/bioinformatics/btl557
Clarke, H., Flint, J., Attwood, A. S., & Munafo, M. R. (2010). Association of the 5-
HTTLPR genotype and unipolar depression: A meta-analysis. Psychological
Medicine, 40(11), 1767-1778. doi:10.1017/s0033291710000516
Clayton, E. W. (2015). Beyond myalgic encephalomyelitis/chronic fatigue
syndrome: an IOM report on redefining an illness. JAMA, 313(11), 1101-
1102. doi:10.1001/jama.2015.1346
Cleare, A. J. (2004). The HPA axis and the genesis of chronic fatigue syndrome.
Trends in Endocrinology and Metabolism, 15(2), 55-59.
doi:10.1016/j.tem.2003.12.002
Clements, G. B., McGarry, F., Nairn, C., & Galbraith, D. N. (1995). Detection of
enterovirus-specific RNA in serum: The relationship to chronic fatigue.
Journal of Medical Virology, 45(2), 156-161. doi:10.1002/jmv.1890450208
Cohen-Woods, S., Craig, I. W., & McGuffin, P. (2013). The current state of play on
the molecular genetics of depression. Psychological Medicine, 43(4), 673-
687. doi:10.1017/s0033291712001286
Collin, S. M., Bakken, I. J., Nazareth, I., Crawley, E., & White, P. D. (2017). Trends
in the incidence of chronic fatigue syndrome and fibromyalgia in the UK,
2001-2013: a Clinical Practice Research Datalink study. Journal of the Royal
Society of Medicine, 141076817702530. doi:10.1177/0141076817702530
CONVERGE Consortium. (2015). Sparse whole-genome sequencing identifies two
loci for major depressive disorder. Nature, 523(7562), 588-591.
doi:10.1038/nature14659
Corfield, E. C., Martin, N. G., & Nyholt, D. R. (2016a). Co-occurrence and
symptomatology of fatigue and depression. Comprehensive Psychiatry, 71, 1-
10. doi:10.1016/j.comppsych.2016.08.004
Corfield, E. C., Martin, N. G., & Nyholt, D. R. (2016b). Shared Genetic Factors in
the Co-Occurrence of Depression and Fatigue. Twin Research and Human
Genetics, 1-9. doi:10.1017/thg.2016.79
Costello, E. J., & Angold, A. (1988). Scales to assess child and adolescent
depression: Checklists, screens, and nets. Journal of the American Academy
of Child and Adolescent Psychiatry, 27(6), 726-737. doi:10.1097/00004583-
198811000-00011
Cross-Disorder Group of the Psychiatric Genomics Consortium. (2013). Genetic
relationship between five psychiatric disorders estimated from genome-wide
SNPs. Nature Genetics, 45(9), 984-994. doi:10.1038/ng.2711
Cuijpers, P., de Graaf, R., & van Dorsselaer, S. (2004). Minor depression: Risk
profiles, functional disability, health care use and risk of developing major
depression. Journal of Affective Disorders, 79(1–3), 71-79.
doi:10.1016/S0165-0327(02)00348-8
Cuijpers, P., & Smit, F. (2004). Subthreshold depression as a risk indicator for major
depressive disorder: A systematic review of prospective studies. Acta
Psychiatrica Scandinavica, 109(5), 325-331. doi:10.1111/j.1600-
0447.2004.00301.x
170 Bibliography
Cuijpers, P., Smit, F., Oostenbrink, J., de Graaf, R., Ten Have, M., & Beekman, A.
(2007). Economic costs of minor depression: A population-based study. Acta
Psychiatrica Scandinavica, 115(3), 229-236. doi:10.1111/j.1600-
0447.2006.00851.x
Cullen, W., Kearney, Y., & Bury, G. (2002). Prevalence of fatigue in general
practice. Irish Journal of Medical Science, 171(1), 10-12.
Dalakas, M. C. (2003). Enteroviruses in chronic fatigue syndrome: "now you see
them, now you don't". Journal of Neurology, Neurosurgery and Psychiatry,
74(10), 1361-1362. doi:10.1136/jnnp.74.10.1361
Das, S., Forer, L., Schonherr, S., Sidore, C., Locke, A. E., Kwong, A., . . .
Fuchsberger, C. (2016). Next-generation genotype imputation service and
methods. Nature Genetics, 48(10), 1284-1287. doi:10.1038/ng.3656
David, A., Pelosi, A., McDonald, E., Stephens, D., Ledger, D., Rathbone, R., &
Mann, A. (1990). Tired, weak, or in need of rest: Fatigue among general
practice attenders. BMJ, 301(6762), 1199-1202.
doi:10.1136/bmj.301.6762.1199
de Lange, F. P., Kalkman, J. S., Bleijenberg, G., Hagoort, P., van der Meer, J. W., &
Toni, I. (2005). Gray matter volume reduction in the chronic fatigue
syndrome. Neuroimage, 26(3), 777-781.
doi:10.1016/j.neuroimage.2005.02.037
de Leeuw, C. A., Mooij, J. M., Heskes, T., & Posthuma, D. (2015). MAGMA:
Generalized gene-set analysis of GWAS data. PLoS Computational Biology,
11(4), e1004219. doi:10.1371/journal.pcbi.1004219
de Vries, B., Anttila, V., Freilinger, T., Wessman, M., Kaunisto, M. A., Kallela, M., .
. . van den Maagdenberg, A. M. (2015). Systematic re-evaluation of genes
from candidate gene association studies in migraine using a large genome-
wide association data set. Cephalalgia. doi:10.1177/0333102414566820
Deary, V., Hagenaars, S. P., Harris, S. E., Hill, W. D., Davies, G., Liewald, D. C., . .
. Deary, I. J. (2017). Genetic contributions to self-reported tiredness.
Molecular Psychiatry. doi:10.1038/mp.2017.5
Deckx, L., van den Akker, M., Vergeer, D., van Abbema, D., van den Berkmortel, F.,
Linsen, L., . . . Buntinx, F. (2015). The value of fatigue severity to rule out
depression in older adult patients with cancer. Oncology Nursing Forum,
42(4), E302-E309. doi:10.1188/15.onf.e302-e309
Devanur, L. D., & Kerr, J. R. (2006). Chronic fatigue syndrome. Journal of Clinical
Virology, 37(3), 139-150. doi:10.1016/j.jcv.2006.08.013
Di Luca, D., Zorzenon, M., Mirandola, P., Colle, R., Botta, G. A., & Cassai, E.
(1995). Human herpesvirus 6 and human herpesvirus 7 in chronic fatigue
syndrome. Journal of Clinical Microbiology, 33(6), 1660-1661.
Direk, N., Williams, S., Smith, J. A., Ripke, S., Air, T., Amare, A. T., . . . Sullivan,
P. F. (2016). An analysis of two genome-wide association meta-analyses
identifies a new locus for broad depression phenotype. Biological Psychiatry.
doi:10.1016/j.biopsych.2016.11.013
Duncan, A. E., Munn-Chernoff, M. A., Hudson, D. L., Eschenbacher, M. A.,
Agrawal, A., Grant, J. D., . . . Heath, A. C. (2014). Genetic and
environmental risk for major depression in African-American and European-
American women. Twin Res Hum Genet, 17(4), 244-253.
doi:10.1017/thg.2014.28
Bibliography 171
Endresen, G. K. (2003). Mycoplasma blood infection in chronic fatigue and
fibromyalgia syndromes. Rheumatology International, 23(5), 211-215.
doi:10.1007/s00296-003-0355-7
Erlwein, O., Kaye, S., McClure, M. O., Weber, J., Wills, G., Collier, D., . . . Cleare,
A. (2010). Failure to detect the novel retrovirus XMRV in chronic fatigue
syndrome. PloS One, 5(1), e8519. doi:10.1371/journal.pone.0008519
Eu-Ahsunthornwattana, J., Miller, E. N., Fakiola, M., Jeronimo, S. M., Blackwell, J.
M., & Cordell, H. J. (2014). Comparison of methods to account for
relatedness in genome-wide association studies with family-based data. PLoS
Genet, 10(7), e1004445. doi:10.1371/journal.pgen.1004445
EvengÅRd, B., Jacks, A., Pedersen, N. L., & Sullivan, P. F. (2005). The
epidemiology of chronic fatigue in the Swedish Twin Registry. Psychological
Medicine, 35(9), 1317-1326. doi:10.1017/S0033291705005052
Faravelli, C., Alessandra Scarpato, M., Castellini, G., & Lo Sauro, C. (2013). Gender
differences in depression and anxiety: The role of age. Psychiatry Research,
210(3), 1301-1303. doi:10.1016/j.psychres.2013.09.027
Farmer, A., Scourfield, J., Martin, N., Cardno, A., & McGuffin, P. (1999). Is
disabling fatigue in childhood influenced by genes? Psychological Medicine,
29(2), 279-282. doi:10.1017/s0033291798008095
Fava, M., Ball, S., Nelson, J. C., Sparks, J., Konechnik, T., Classi, P., . . . Thase, M.
E. (2014). Clinical relevance of fatigue as a residual symptom in major
depressive disorder. Depression and Anxiety, 31(3), 250-257.
doi:10.1002/da.22199
Ferentinos, P., Koukounari, A., Power, R., Rivera, M., Uher, R., Craddock, N., . . .
Lewis, C. M. (2015). Familiality and SNP heritability of age at onset and
episodicity in major depressive disorder. Psychological Medicine, 45(10),
2215-2225. doi:10.1017/s0033291715000215
Fernandez-Pujals, A. M., Adams, M. J., Thomson, P., McKechanie, A. G.,
Blackwood, D. H. R., Smith, B. H., . . . McIntosh, A. M. (2015).
Epidemiology and Heritability of Major Depressive Disorder, Stratified by
Age of Onset, Sex, and Illness Course in Generation Scotland: Scottish
Family Health Study (GS:SFHS). PloS One, 10(11), e0142197.
doi:10.1371/journal.pone.0142197
Fils, J. M., Penick, E. C., Nickel, E. J., Othmer, E., Desouza, C., Gabrielli, W. F., &
Hunter, E. E. (2010). Minor versus major depression: A comparative clinical
study. Primary Care Companion to the Journal of Clinical Psychiatry, 12(1),
PCC.08m00752. doi:10.4088/PCC.08m00752blu
Flint, J., & Kendler, K. S. (2014). The genetics of major depression. Neuron, 81(3),
484-503. doi:10.1016/j.neuron.2014.01.027
Fowler, T. A., Rice, F., Thapar, A., & Farmer, A. (2006). Relationship between
disabling fatigue and depression in children: Genetic study. British Journal of
Psychiatry, 189, 247-253. doi:10.1192/bjp.bp.105.011379
Friedberg, F., Tintle, N., Clark, J., & Bromet, E. J. (2015). Prolonged fatigue in
Ukraine and the United States: Prevalence and risk factors. Fatigue, 3(1), 33-
46. doi:10.1080/21641846.2014.993829
Fukuda, K., Straus, S. E., Hickie, I., Sharpe, M. C., Dobbins, J. G., & Komaroff, A.
(1994). The chronic fatigue syndrome: A comprehensive approach to its
definition and study. Annals of Internal Medicine, 121(12), 953-959.
doi:10.7326/0003-4819-121-12-199412150-00009
172 Bibliography
Fukuda, S., Hashimoto, R., Ohi, K., Yamaguti, K., Nakatomi, Y., Yasuda, Y., . . .
Watanabe, Y. (2010). A functional polymorphism in the disrupted-in
schizophrenia 1 gene is associated with chronic fatigue syndrome. Life
Sciences, 86(19–20), 722-725. doi:10.1016/j.lfs.2010.03.007
Furlong, R. A., Ho, L., Walsh, C., Rubinsztein, J. S., Jain, S., Paykel, E. S., . . .
Rubinsztein, D. C. (1998). Analysis and meta-analysis of two serotonin
transporter gene polymorphisms in bipolar and unipolar affective disorders.
American Journal of Medical Genetics, 81(1), 58-63.
doi:10.1002/(SICI)1096-8628(19980207)81:1<58::AID-AJMG11>3.0.CO;2-
V
Garcia-Borreguero, D., Dale, J. K., Rosenthal, N. E., Chiara, A., O'Fallon, A.,
Bartko, J. J., & Straus, S. E. (1998). Lack of seasonal variation of symptoms
in patients with chronic fatigue syndrome. Psychiatry Research, 77(2), 71-77.
doi:10.1016/S0165-1781(97)00141-8
Gaysina, D., Cohen, S., Craddock, N., Farmer, A., Hoda, F., Korszun, A., . . .
McGuffin, P. (2008). No association with the 5,10-methylenetetrahydrofolate
reductase gene and major depressive disorder: Results of the depression case
control (DeCC) study and a meta-analysis. American Journal of Medical
Genetics. Part B: Neuropsychiatric Genetics, 147b(6), 699-706.
doi:10.1002/ajmg.b.30665
Gershon, E. S., Hamovit, J., Guroff, J. J., Dibble, E., Leckman, J. F., Sceery, W., . . .
Bunney, W. E., Jr. (1982). A family study of schizoaffective, bipolar I,
bipolar II, unipolar, and normal control probands. Archives of General
Psychiatry, 39(10), 1157-1167. doi:10.1001/archpsyc.1982.04290100031006
Gilbody, S., Lewis, S., & Lightfoot, T. (2007). Methylenetetrahydrofolate reductase
(MTHFR) genetic polymorphisms and psychiatric disorders: A HuGE review.
American Journal of Epidemiology, 165(1), 1-13. doi:10.1093/aje/kwj347
Glaser, R., Padgett, D. A., Litsky, M. L., Baiocchi, R. A., Yang, E. V., Chen, M., . . .
Williams, M. V. (2005). Stress-associated changes in the steady-state
expression of latent Epstein-Barr virus: Implications for chronic fatigue
syndrome and cancer. Brain, Behavior, and Immunity, 19(2), 91-103.
doi:10.1016/j.bbi.2004.09.001
Glowinski, A. L., Madden, P. A., Bucholz, K. K., Lynskey, M. T., & Heath, A. C.
(2003). Genetic epidemiology of self-reported lifetime DSM-IV major
depressive disorder in a population-based twin sample of female adolescents.
Journal of Child Psychology and Psychiatry and Allied Disciplines, 44(7),
988-996.
Goel, N., Terman, M., & Terman, J. S. (2002). Depressive symptomatology
differentiates subgroups of patients with seasonal affective disorder.
Depression and Anxiety, 15(1), 34-41. doi:10.1002/da.1083
Goertzel, B. N., Pennachin, C., de Souza Coelho, L., Gurbaxani, B., Maloney, E. M.,
& Jones, J. F. (2006). Combinations of single nucleotide polymorphisms in
neuroendocrine effector and receptor genes predict chronic fatigue syndrome.
Pharmacogenomics, 7(3), 475-483. doi:10.2217/14622416.7.3.475
Goldberg, D. P., & Blackwell, B. (1970). Psychiatric illness in general practice. A
detailed study using a new method of case identification. British Medical
Journal, 1(5707), 439-443.
Gorman, G. S., Elson, J. L., Newman, J., Payne, B., McFarland, R., Newton, J. L., &
Turnbull, D. M. (2015). Perceived fatigue is highly prevalent and debilitating
Bibliography 173
in patients with mitochondrial disease. Neuromuscular Disorders, 25(7), 563-
566. doi:10.1016/j.nmd.2015.03.001
Gow, J. W., Behan, W. M., Clements, G. B., Woodall, C., Riding, M., & Behan, P.
O. (1991). Enteroviral RNA sequences detected by polymerase chain reaction
in muscle of patients with postviral fatigue syndrome. BMJ, 302(6778), 692-
696.
Grans, H., Nilsson, P., & Evengard, B. (2005). Gene expression profiling in the
chronic fatigue syndrome. Journal of Internal Medicine, 258(4), 388-390.
doi:10.1111/j.1365-2796.2005.01548.x
Greenberg, P. E., Fournier, A. A., Sisitsky, T., Pike, C. T., & Kessler, R. C. (2015).
The economic burden of adults with major depressive disorder in the United
States (2005 and 2010). Journal of Clinical Psychiatry, 76(2), 155-162.
doi:10.4088/JCP.14m09298
Griffith, J. P., & Zarrouf, F. A. (2008). A systematic review of chronic fatigue
syndrome: Don't assume it's depression. Primary Care Companion to the
Journal of Clinical Psychiatry, 10(2), 120-128. doi:10.4088/PCC.v10n0206
Hadzi-Pavlovic, D., Hickie, I. B., Wilson, A. J., Davenport, T. A., Lloyd, A. R., &
Wakefield, D. (2000). Screening for prolonged fatigue syndromes: Validation
of the SOFA scale. Social Psychiatry and Psychiatric Epidemiology, 35(10),
471-479. doi:10.1007/s001270050266
Hamaguchi, M., Kawahito, Y., Takeda, N., Kato, T., & Kojima, T. (2011).
Characteristics of chronic fatigue syndrome in a Japanese community
population chronic fatigue syndrome in Japan. Clinical Rheumatology, 30(7),
895-906. doi:10.1007/s10067-011-1702-9
Harvey, S. B., Wessely, S., Kuh, D., & Hotopf, M. (2009). The relationship between
fatigue and psychiatric disorders: Evidence for the concept of neurasthenia.
Journal of Psychosomatic Research, 66(5), 445-454.
doi:10.1016/j.jpsychores.2008.12.007
Heath, A. C., Howells, W., Kirk, K. M., Madden, P. A., Bucholz, K. K., Nelson, E.
C., . . . Martin, N. G. (2001). Predictors of non-response to a questionnaire
survey of a volunteer twin panel: Findings from the Australian 1989 twin
cohort. Twin Research, 4(2), 73-80.
Hek, K., Demirkan, A., Lahti, J., Terracciano, A., Teumer, A., Cornelis, M. C., . . .
Murabito, J. (2013). A genome-wide association study of depressive
symptoms. Biological Psychiatry, 73(7), 667-678.
doi:10.1016/j.biopsych.2012.09.033
Hickie, I., Bennett, B., Lloyd, A., Heath, A., & Martin, N. (1999a). Complex genetic
and environmental relationships between psychological distress, fatigue and
immune functioning: A twin study. Psychological Medicine, 29(2), 269-277.
doi:10.1017/s0033291798007922
Hickie, I., Kirk, K., & Martin, N. (1999b). Unique genetic and environmental
determinants of prolonged fatigue: A twin study. Psychological Medicine,
29(2), 259-268. doi:10.1017/s0033291798007934
Hickie, I., Koschera, A., Hadzi-Pavlovic, D., Bennett, B., & Lloyd, A. (1999c). The
temporal stability and co-morbidity of prolonged fatigue: A longitudinal
study in primary care. Psychological Medicine, 29(4), 855-861.
Hickie, I., Davenport, T., Wakefield, D., Vollmer-Conna, U., Cameron, B., Vernon,
S. D., . . . Lloyd, A. (2006). Post-infective and chronic fatigue syndromes
precipitated by viral and non-viral pathogens: Prospective cohort study. BMJ,
333(7568), 575-580. doi:10.1136/bmj.38933.585764.AE
174 Bibliography
Hickie, I. B., Hooker, A. W., HadziPavlovic, D., Bennett, B. K., & Wilson, A. J.
(1996). Fatigue in selected primary care settings: Sociodemographic and
psychiatric correlates. Medical Journal of Australia, 164(10), 585-588.
Hickie, I. B., Bansal, A. S., Kirk, K. M., Lloyd, A. R., & Martin, N. G. (2001). A
twin study of the etiology of prolonged fatigue and immune activation. Twin
Research, 4(2), 94-102. doi:10.1375/1369052012209
Hirschhorn, J. N., Lohmueller, K., Byrne, E., & Hirschhorn, K. (2002). A
comprehensive review of genetic association studies. Genetics in Medicine,
4(2), 45-61. doi:10.1097/00125817-200203000-00002
Hoad, A., Spickett, G., Elliott, J., & Newton, J. (2008). Postural orthostatic
tachycardia syndrome is an under-recognized condition in chronic fatigue
syndrome. Qjm-an International Journal of Medicine, 101(12), 961-965.
doi:10.1093/qjmed/hcn123
Holmes, G. P., Kaplan, J. E., Stewart, J. A., Hunt, B., Pinsky, P. F., & Schonberger,
L. B. (1987). A cluster of patients with a chronic mononucleosis-like
syndrome. Is Epstein-Barr virus the cause? JAMA, 257(17), 2297-2302.
doi:10.1001/jama.1987.03390170053027
Holmes, G. P., Kaplan, J. E., Gantz, N. M., Komaroff, A. L., Schonberger, L. B.,
Straus, S. E., . . . et al. (1988). Chronic fatigue syndrome: A working case
definition. Annals of Internal Medicine, 108(3), 387-389. doi:10.7326/0003-
4819-108-3-387
Hur, Y.-M., Burri, A., & Spector, T. D. (2012). The genetic and environmental
structure of the covariation among the symptoms of insomnia, fatigue, and
depression in adult females. Twin Research and Human Genetics, 15(06),
720-726. doi:10.1017/thg.2012.60
Hybels, C. F., Landerman, L. R., & Blazer, D. G. (2012). Age differences in
symptom expression in patients with major depression. International Journal
of Geriatric Psychiatry, 27(6), 601-611. doi:10.1002/gps.2759
Hyde, C. L., Nagle, M. W., Tian, C., Chen, X., Paciga, S. A., Wendland, J. R., . . .
Winslow, A. R. (2016). Identification of 15 genetic loci associated with risk
of major depression in individuals of European descent. Nature Genetics,
48(9), 1031-1036. doi:10.1038/ng.3623
Ikuta, K., Yamada, T., Shimomura, T., Kuratsune, H., Kawahara, R., Ikawa, S., . . .
Sairenji, T. (2003). Diagnostic evaluation of 2', 5'-oligoadenylate synthetase
activities and antibodies against Epstein-Barr virus and Coxiella burnetii in
patients with chronic fatigue syndrome in Japan. Microbes Infect, 5(12),
1096-1102. doi:10.1016/j.micinf.2003.07.002
Institute of Medicine (IOM). (2015). Beyond myalgic encephalomyelitis/chronic
fatigue syndrome: Redefining an Illness. Washington, DC: The National
Academies Press.
Jacobson, S. K., Daly, J. S., Thorne, G. M., & McIntosh, K. (1997). Chronic
parvovirus B19 infection resulting in chronic fatigue syndrome: Case history
and review. Clinical Infectious Diseases, 24(6), 1048-1051.
doi:10.1086/513627
Janssens, K. A., Zijlema, W. L., Joustra, M. L., & Rosmalen, J. G. (2015). Mood and
Anxiety Disorders in Chronic Fatigue Syndrome, Fibromyalgia, and Irritable
Bowel Syndrome: Results From the LifeLines Cohort Study. Psychosomatic
Medicine, 77(4), 449-457. doi:10.1097/PSY.0000000000000161
Jason, L. A., Taylor, R., Wagner, L., Holden, J., Ferrari, J. R., Plioplys, A. V., . . .
Papernik, M. (1995). Estimating rates of chronic fatigue syndrome from a
Bibliography 175
community-based sample: A pilot study. American Journal of Community
Psychology, 23(4), 557-568. doi:10.1007/bf02506968
Jason, L. A., Richman, J. A., Rademaker, A. W., Jordan, K. M., Plioplys, A. V.,
Taylor, R. R., . . . Plioplys, S. (1999). A community-based study of chronic
fatigue syndrome. Archives of Internal Medicine, 159(18), 2129-2137.
doi:10.1001/archinte.159.18.2129
Jason, L. A., Taylor, R. R., & Carrico, A. W. (2001). A community-based study of
seasonal variation in the onset of chronic fatigue syndrome and idiopathic
chronic fatigue. Chronobiology International, 18(2), 315-319.
doi:10.1081/CBI-100103194
Jason, L. A., Corradi, K., Torres-Harding, S., Taylor, R. R., & King, C. (2005).
Chronic fatigue syndrome: The need for subtypes. Neuropsychology Review,
15(1), 29-58. doi:10.1007/s11065-005-3588-2
Jason, L. A., Benton, M. C., Valentine, L., Johnson, A., & Torres-Harding, S. (2008).
The eonomic impact of ME/CFS: Individual and societal costs. Dynamic
Medicine, 7(1), 6. doi:10.1186/1476-5918-7-6
Jason, L. A., Brown, A., Clyne, E., Bartgis, L., Evans, M., & Brown, M. (2012).
Contrasting case definitions for chronic fatigue syndrome, myalgic
encephalomyelitis/chronic fatigue syndrome and myalgic encephalomyelitis.
Evaluation and the Health Professions, 35(3), 280-304.
doi:10.1177/0163278711424281
Jason, L. A., Brown, A., Evans, M., Sunnquist, M., & Newton, J. L. (2013).
Contrasting chronic fatigue syndrome versus myalgic
encephalomyelitis/chronic fatigue syndrome. Fatigue, 1(3), 168-183.
doi:10.1080/21641846.2013.774556
Johnston, S., Brenu, E. W., Staines, D., & Marshall-Gradisnik, S. (2013). The
prevalence of chronic fatigue syndrome/ myalgic encephalomyelitis: A meta-
analysis. Clinical Epidemiology, 5, 105-110. doi:10.2147/clep.s39876
Johnston, S. C., Brenu, E. W., Hardcastle, S. L., Huth, T. K., Staines, D. R., &
Marshall-Gradisnik, S. M. (2014). A comparison of health status in patients
meeting alternative definitions for chronic fatigue syndrome/myalgic
encephalomyelitis. Health Qual Life Outcomes, 12, 64-70. doi:10.1186/1477-
7525-12-64
Jones, J. F., Streib, J., Baker, S., & Herberger, M. (1991). Chronic fatigue syndrome:
I. Epstein-Barr virus immune response and molecular epidemiology. Journal
of Medical Virology, 33(3), 151-158. doi:10.1002/jmv.1890330303
Judd, L. L., Akiskal, H. S., & Paulus, M. P. (1997). The role and clinical significance
of subsyndromal depressive symptoms (SSD) in unipolar major depressive
disorder. Journal of Affective Disorders, 45(1–2), 5-18. doi:10.1016/S0165-
0327(97)00055-4
Kato, K., Sullivan, P. F., Evengard, B., & Pedersen, N. L. (2009). A population-
based twin study of functional somatic syndromes. Psychological Medicine,
39(3), 497-505. doi:10.1017/s0033291708003784
Katon, W. J., Buchwald, D. S., Simon, G. E., Russo, J. E., & Mease, P. J. (1991).
Psychiatric illness in patients with chronic fatigue and those with rheumatoid
arthritis. Journal of General Internal Medicine, 6(4), 277-285.
doi:10.1007/BF02597420
Katz, B. Z. (2002). Update on chronic fatigue syndrome and Epstein-Barr virus.
Pediatric Annals, 31(11), 741-744. doi:10.3928/0090-4481-20021101-10
176 Bibliography
Kaushik, N., Fear, D., Richards, S. C., McDermott, C. R., Nuwaysir, E. F., Kellam,
P., . . . Kerr, J. R. (2005). Gene expression in peripheral blood mononuclear
cells from patients with chronic fatigue syndrome. Journal of Clinical
Pathology, 58(8), 826-832. doi:10.1136/jcp.2005.025718
Kawai, K., & Kawai, A. (1992). Studies on the relationship between chronic fatigue
syndrome and Epstein-Barr virus in Japan. Internal Medicine, 31(3), 313-318.
doi:10.2169/internalmedicine.31.313
Kawakami, N., Iwata, N., Fujihara, S., & Kitamura, T. (1998). Prevalence of chronic
fatigue syndrome in a community population in Japan. Tohoku Journal of
Experimental Medicine, 186(1), 33-41. doi:10.1620/tjem.186.33
Kendler, K. S., Neale, M. C., Kessler, R. C., Heath, A. C., & Eaves, L. J. (1992). A
population-based twin study of major depression in women. The impact of
varying definitions of illness. Archives of General Psychiatry, 49(4), 257-
266.
Kendler, K. S. (1993). Twin studies of psychiatric illness. Current status and future
directions. Archives of General Psychiatry, 50(11), 905-915.
Kendler, K. S., Neale, M. C., MacLean, C. J., Heath, A. C., Eaves, L. J., & Kessler,
R. C. (1993). Smoking and major depression. A causal analysis. Archives of
General Psychiatry, 50(1), 36-43.
doi:10.1001/archpsyc.1993.01820130038007
Kendler, K. S., Pedersen, N. L., Neale, M. C., & Mathe, A. A. (1995). A pilot
Swedish twin study of affective illness including hospital- and population-
ascertained subsamples: Results of model fitting. Behavior Genetics, 25(3),
217-232. doi:10.1001/archpsyc.1993.01820210033004
Kendler, K. S., Karkowski, L. M., & Prescott, C. A. (1999). Causal relationship
between stressful life events and the onset of major depression. American
Journal of Psychiatry, 156(6), 837-841. doi:10.1176/ajp.156.6.837
Kendler, K. S., & Prescott, C. A. (1999). A population-based twin study of lifetime
major depression in men and women. Archives of General Psychiatry, 56(1),
39-44. doi:10.1001/archpsyc.56.1.39
Kendler, K. S., Thornton, L. M., & Gardner, C. O. (2000). Stressful life events and
previous episodes in the etiology of major depression in women: An
evaluation of the "kindling" hypothesis. American Journal of Psychiatry,
157(8), 1243-1251.
Kendler, K. S., & Aggen, S. H. (2001). Time, memory and the heritability of major
depression. Psychological Medicine, 31(5), 923-928.
Kendler, K. S., Gardner, C. O., Neale, M. C., & Prescott, C. A. (2001). Genetic risk
factors for major depression in men and women: Similar or different
heritabilities and same or partly distinct genes? Psychological Medicine,
31(4), 605-616. doi:10.1017/S0033291701003907
Kendler, K. S., Gatz, M., Gardner, C. O., & Pedersen, N. L. (2006). A Swedish
national twin study of lifetime major depression. American Journal of
Psychiatry, 163(1), 109-114. doi:10.1176/appi.ajp.163.1.109
Kendler, K. S., Aggen, S. H., & Neale, M. C. (2013). Evidence for multiple genetic
factors underlying DSM-IV criteria for major depression. JAMA Psychiatry,
70(6), 599-607. doi:10.1001/jamapsychiatry.2013.751
Kennedy, G., Spence, V. A., McLaren, M., Hill, A., Underwood, C., & Belch, J. J.
(2005). Oxidative stress levels are raised in chronic fatigue syndrome and are
associated with clinical symptoms. Free Radical Biology and Medicine,
39(5), 584-589. doi:10.1016/j.freeradbiomed.2005.04.020
Bibliography 177
Kerr, J. R., Barah, F., Mattey, D. L., Laing, I., Hopkins, S. J., Hutchinson, I. V., &
Tyrrell, D. A. (2001). Circulating tumour necrosis factor-alpha and
interferon-gamma are detectable during acute and convalescent parvovirus
B19 infection and are associated with prolonged and chronic fatigue. Journal
of General Virology, 82(Pt 12), 3011-3019. doi:10.1099/0022-1317-82-12-
3011
Kerr, J. R., Bracewell, J., Laing, I., Mattey, D. L., Bernstein, R. M., Bruce, I. N., &
Tyrrell, D. A. (2002). Chronic fatigue syndrome and arthralgia following
parvovirus B19 infection. Journal of Rheumatology, 29(3), 595-602.
Kerr, J. R., & Mattey, D. L. (2008). Preexisting psychological stress predicts acute
and chronic fatigue and arthritis following symptomatic parvovirus B19
infection. Clinical Infectious Diseases, 46(9), e83-e87. doi:10.1086/533471
Kessler, R. C., Berglund, P., Demler, O., Jin, R., Koretz, D., Merikangas, K. R., . . .
Wang, P. S. (2003). The epidemiology of major depressive disorder: Results
from the National Comorbidity Survey Replication (NCS-R). JAMA, 289(23),
3095-3105. doi:10.1001/jama.289.23.3095
Kim, C. H., Shin, H. C., & Won, C. W. (2005). Prevalence of chronic fatigue and
chronic fatigue syndrome in Korea: Community-based primary care study.
Journal of Korean Medical Science, 20(4), 529-534.
doi:10.3346/jkms.2005.20.4.529
Kirk, K. M., Eaves, L. J., & Martin, N. G. (1999a). Self-transcendence as a measure
of spirituality in a sample of older Australian twins. Twin Research, 2(2), 81-
87. doi:10.1375/136905299320565942
Kirk, K. M., Hickie, I. B., & Martin, N. G. (1999b). Fatigue as related to anxiety and
depression in a community-based sample of twins aged over 50. Social
Psychiatry and Psychiatric Epidemiology, 34(2), 85-90.
doi:10.1007/s001270050116
Kishi, T., Tsunoka, T., Ikeda, M., Kawashima, K., Okochi, T., Kitajima, T., . . .
Iwata, N. (2009). Serotonin 1A receptor gene and major depressive disorder:
An association study and meta-analysis. Journal of Human Genetics, 54(11),
629-633. doi:10.1038/jhg.2009.84
Kishi, T., Yoshimura, R., Fukuo, Y., Okochi, T., Matsunaga, S., Umene-Nakano, W.,
. . . Iwata, N. (2013). The serotonin 1A receptor gene confer susceptibility to
mood disorders: Results from an extended meta-analysis of patients with
major depression and bipolar disorder. European Archives of Psychiatry and
Clinical Neuroscience, 263(2), 105-118. doi:10.1007/s00406-012-0337-4
Knox, K., Carrigan, D., Simmons, G., Teque, F., Zhou, Y., Hackett, J., Jr., . . . Levy,
J. A. (2011). No evidence of murine-like gammaretroviruses in CFS patients
previously identified as XMRV-infected. Science, 333(6038), 94-97.
doi:10.1126/science.1204963
Kohli, M. A., Lucae, S., Saemann, P. G., Schmidt, M. V., Demirkan, A., Hek, K., . . .
Binder, E. B. (2011). The neuronal transporter gene SLC6A15 confers risk to
major depression. Neuron, 70(2), 252-265. doi:10.1016/j.neuron.2011.04.005
Komaroff, A. L. (2006). Is human herpesvirus-6 a trigger for chronic fatigue
syndrome? Journal of Clinical Virology, 37 Suppl 1, S39-S46.
doi:10.1016/s1386-6532(06)70010-5
Koo, D. (1989). Chronic fatigue syndrome. A critical appraisal of the role of Epstein-
Barr virus. Western Journal of Medicine, 150(5), 590-596.
178 Bibliography
Kroenke, K., Wood, D. R., Mangelsdorff, A., Meier, N. J., & Powell, J. B. (1988).
Chronic fatigue in primary care: Prevalence, patient characteristics, and
outcome. JAMA, 260(7), 929-934. doi:10.1001/jama.1988.03410070057028
Kroenke, K., Spitzer, R. L., Williams, J. B., Linzer, M., Hahn, S. R., deGruy, F. V.,
3rd, & Brody, D. (1994). Physical symptoms in primary care. Predictors of
psychiatric disorders and functional impairment. Archives of Family
Medicine, 3(9), 774-779.
Kroenke, K. (2003). Patients presenting with somatic complaints: Epidemiology,
psychiatric comorbidity and management. International Journal of Methods
in Psychiatric Research, 12(1), 34-43. doi:10.1002/mpr.140
Krueger, R. F., & Bezdjian, S. (2009). Enhancing research and treatment of mental
disorders with dimensional concepts: Toward DSM-V and ICD-11. World
Psychiatry, 8(1), 3-6.
Kuehner, C. (2003). Gender differences in unipolar depression: An update of
epidemiological findings and possible explanations. Acta Psychiatrica
Scandinavica, 108(3), 163-174. doi:10.1034/j.1600-0447.2003.00204.x
Landmark-Hoyvik, H., Reinertsen, K. V., Loge, J. H., Kristensen, V. N., Dumeaux,
V., Fossa, S. D., . . . Edvardsen, H. (2010). The genetics and epigenetics of
fatigue. Pm r, 2(5), 456-465. doi:10.1016/j.pmrj.2010.04.003
Lane, R. J., Soteriou, B. A., Zhang, H., & Archard, L. C. (2003). Enterovirus related
metabolic myopathy: A postviral fatigue syndrome. Journal of Neurology,
Neurosurgery and Psychiatry, 74(10), 1382-1386.
doi:10.1136/jnnp.74.10.1382
Lasky-Su, J. A., Faraone, S. V., Glatt, S. J., & Tsuang, M. T. (2005). Meta-analysis
of the association between two polymorphisms in the serotonin transporter
gene and affective disorders. American Journal of Medical Genetics. Part B:
Neuropsychiatric Genetics, 133b(1), 110-115. doi:10.1002/ajmg.b.30104
Ledina, D., Bradaric, N., Milas, I., Ivic, I., Brncic, N., & Kuzmicic, N. (2007).
Chronic fatigue syndrome after Q fever. Medical Science Monitor, 13(7),
Cs88-Cs92.
Lerner, A. M., Beqaj, S. H., Deeter, R. G., & Fitzgerald, J. T. (2004). IgM serum
antibodies to Epstein-Barr virus are uniquely present in a subset of patients
with the chronic fatigue syndrome. In Vivo, 18(2), 101-106.
Levine, P. H., Whiteside, T. L., Friberg, D., Bryant, J., Colclough, G., & Herberman,
R. B. (1998). Dysfunction of natural killer activity in a family with chronic
fatigue syndrome. Clinical Immunology and Immunopathology, 88(1), 96-
104. doi:10.1006/clin.1998.4554
Lewinsohn, P. M., Klein, D. N., Durbin, E. C., Seeley, J. R., & Rohde, P. (2003).
Family study of subthreshold depressive symptoms: Risk factor for MDD?
Journal of Affective Disorders, 77(2), 149-157. doi:10.1016/S0165-
0327(02)00106-4
Lewis, C. M., Ng, M. Y., Butler, A. W., Cohen-Woods, S., Uher, R., Pirlo, K., . . .
McGuffin, P. (2010). Genome-wide association study of major recurrent
depression in the U.K. population. American Journal of Psychiatry, 167(8),
949-957. doi:10.1176/appi.ajp.2010.09091380
Lewis, G., & Wessely, S. (1992). The epidemiology of fatigue: More questions than
answers. Journal of Epidemiology and Community Health, 46(2), 92-97.
doi:10.1136/jech.46.2.92
Bibliography 179
Li, J., & Ji, L. (2005). Adjusting multiple testing in multilocus analyses using the
eigenvalues of a correlation matrix. Heredity, 95(3), 221-227.
doi:10.1038/sj.hdy.6800717
Ligthart, L., Nyholt, D. R., Penninx, B. W. J. H., & Boomsma, D. I. (2010). The
shared genetics of migraine and anxious depression. Headache: The Journal
of Head and Face Pain, 50(10), 1549-1560. doi:10.1111/j.1526-
4610.2010.01705.x
Lin, E., & Huang, L. C. (2008). Identification of significant genes in genomics using
Bayesian variable selection methods. Advances and Applications in
Bioinformatics and Chemistry, 1, 13-18.
Lin, E., & Hsu, S. Y. (2009). A Bayesian approach to gene-gene and gene-
environment interactions in chronic fatigue syndrome. Pharmacogenomics,
10(1), 35-42. doi:10.2217/14622416.10.1.35
Lindal, E., Stefansson, J. G., & Bergmann, S. (2002). The prevalence of chronic
fatigue syndrome in Iceland - A national comparison by gender drawing on
four different criteria. Nordic Journal of Psychiatry, 56(4), 273-277.
doi:10.1080/08039480260242769
Liu, J. Z., Erlich, Y., & Pickrell, J. K. (2017). Case-control association mapping by
proxy using family history of disease. Nature Genetics, 49(3), 325-331.
doi:10.1038/ng.3766
Lobo, D. M., & Agius, M. (2012). The mental illness spectrum. Psychiatr Danub, 24
Suppl 1, S157-S160.
Loge, J. H., Ekeberg, O., & Kaasa, S. (1998). Fatigue in the general Norwegian
population: Normative data and associations. Journal of Psychosomatic
Research, 45(1), 53-65. doi:10.1016/S0022-3999(97)00291-2
Lohoff, F. W. (2010). Overview of the genetics of major depressive disorder. Curr
Psychiatry Rep, 12(6), 539-546. doi:10.1007/s11920-010-0150-6
Lombardi, V. C., Ruscetti, F. W., Das Gupta, J., Pfost, M. A., Hagen, K. S., Peterson,
D. L., . . . Mikovits, J. A. (2009). Detection of an infectious retrovirus,
XMRV, in blood cells of patients with chronic fatigue syndrome. Science,
326(5952), 585-589. doi:10.1126/science.1179052
Lopez-Leon, S., Janssens, A. C., Gonzalez-Zuloeta Ladd, A. M., Del-Favero, J.,
Claes, S. J., Oostra, B. A., & van Duijn, C. M. (2008). Meta-analyses of
genetic studies on major depressive disorder. Molecular Psychiatry, 13(8),
772-785. doi:10.1038/sj.mp.4002088
Lopez Leon, S., Croes, E. A., Sayed-Tabatabaei, F. A., Claes, S., Van Broeckhoven,
C., & van Duijn, C. M. (2005). The dopamine D4 receptor gene 48-base-pair-
repeat polymorphism and mood disorders: A meta-analysis. Biological
Psychiatry, 57(9), 999-1003. doi:10.1016/j.biopsych.2005.01.030
Lubke, G. H., Hottenga, J. J., Walters, R., Laurin, C., de Geus, E. J., Willemsen, G., .
. . Boomsma, D. I. (2012). Estimating the genetic variance of major
depressive disorder due to all single nucleotide polymorphisms. Biological
Psychiatry, 72(8), 707-709. doi:10.1016/j.biopsych.2012.03.011
Luciano, M., Huffman, J. E., Arias-Vasquez, A., Vinkhuyzen, A. A., Middeldorp, C.
M., Giegling, I., . . . Deary, I. J. (2012). Genome-wide association uncovers
shared genetic effects among personality traits and mood states. American
Journal of Medical Genetics. Part B: Neuropsychiatric Genetics, 159b(6),
684-695. doi:10.1002/ajmg.b.32072
180 Bibliography
Lum, E., Medveczky, M. M., & Medveczky, P. G. (2014). Is inherited human
herpesvirus 6 the perpetrator behind some cases of chronic fatigue syndrome?
Future Microbiology, 9(4), 433-436. doi:10.2217/fmb.14.11
Lyall, M., Peakman, M., & Wessely, S. (2003). A systematic review and critical
evaluation of the immunology of chronic fatigue syndrome. Journal of
Psychosomatic Research, 55(2), 79-90. doi:10.1016/S0022-3999(02)00515-9
Lyon, D., McCain, N., Elswick, R. K., Sturgill, J., Ameringer, S., Jallo, N., . . . Grap,
M. J. (2014). Biobehavioral examination of fatigue across populations:
Report from a P30 Center of Excellence. Nursing Outlook, 62(5), 322-331.
doi:10.1016/j.outlook.2014.06.008
Lyons, M. J., Eisen, S. A., Goldberg, J., True, W., Lin, N., Meyer, J. M., . . . Tsuang,
M. T. (1998). A registry-based twin study of depression in men. Archives of
General Psychiatry, 55(5), 468-472. doi:10.1001/archpsyc.55.5.468
Maier, W., Lichtermann, D., Minges, J., Hallmayer, J., Heun, R., Benkert, O., &
Levinson, D. F. (1993). Continuity and discontinuity of affective disorders
and schizophrenia. Results of a controlled family study. Archives of General
Psychiatry, 50(11), 871-883.
Major Depressive Disorder Working Group of the Psychiatric GWAS Consortium,
Ripke, S., Wray, N. R., Lewis, C. M., Hamilton, S. P., Weissman, M. M., . . .
Sullivan, P. F. (2013). A mega-analysis of genome-wide association studies
for major depressive disorder. Molecular Psychiatry, 18(4), 497-511.
doi:10.1038/mp.2012.21
Marshall-Gradisnik, S., Smith, P., Nilius, B., & Staines, D. R. (2015a). Examination
of single nucleotide polymorphisms in acetylcholine receptors in chronic
fatigue syndrome patients. Immunology and Immunogenetics Insights, 7, 7-
20. doi:10.4137/III.S25105
Marshall-Gradisnik, S., Huth, T., Chacko, A., Johnston, S., Smith, P., & Staines, D.
(2016a). Natural killer cells and single nucleotide polymorphisms of specific
ion channels and receptor genes in myalgic encephalomyelitis/chronic fatigue
syndrome. Appl Clin Genet, 9, 39-47. doi:10.2147/tacg.s99405
Marshall-Gradisnik, S., Johnston, S., Chacko, A., Nguyen, T., Smith, P., & Staines,
D. (2016b). Single nucleotide polymorphisms and genotypes of transient
receptor potential ion channel and acetylcholine receptor genes from isolated
B lymphocytes in myalgic encephalomyelitis/chronic fatigue syndrome
patients. Journal of International Medical Research, 44(6), 1381-1394.
doi:10.1177/0300060516671622
Marshall-Gradisnik, S. M., Smith, P., Brenu, E. W., Nilius, B., Ramos, S. B., &
Staines, D. R. (2015b). Examination of single nucleotide polymorphisms
(SNPs) in transient receptor potential (TRP) ion channels in chronic fatigue
syndrome patients. Immunology and Immunogenetics Insights, 7, 1-6.
doi:10.4137/III.S25147
Matano, S., Kinoshita, H., Tanigawa, K., Terahata, S., & Sugimoto, T. (2003). Acute
parvovirus B19 infection mimicking chronic fatigue syndrome. Internal
Medicine, 42(9), 903-905. doi:10.2169/internalmedicine.42.903
Mbarek, H., Steinberg, S., Nyholt, D. R., Gordon, S. D., Miller, M. B., McRae, A. F.,
. . . Boomsma, D. I. (2016). Identification of Common Genetic Variants
Influencing Spontaneous Dizygotic Twinning and Female Fertility. American
Journal of Human Genetics, 98(5), 898-908. doi:10.1016/j.ajhg.2016.03.008
Bibliography 181
McGue, M., Osler, M., & Christensen, K. (2010). Causal inference and observational
research: The utility of twins. Perspectives on Psychological Science, 5(5),
546-556. doi:10.1177/1745691610383511
McGuffin, P., Katz, R., Watkins, S., & Rutherford, J. (1996). A hospital-based twin
register of the heritability of DSM-IV unipolar depression. Archives of
General Psychiatry, 53(2), 129-136.
doi:10.1001/archpsyc.1996.01830020047006
Medland, S. E., Nyholt, D. R., Painter, J. N., McEvoy, B. P., McRae, A. F., Zhu, G., .
. . Martin, N. G. (2009). Common variants in the trichohyalin gene are
associated with straight hair in Europeans. American Journal of Human
Genetics, 85(5), 750-755. doi:10.1016/j.ajhg.2009.10.009
Meeus, M., Nijs, J., Hermans, L., Goubert, D., & Calders, P. (2013). The role of
mitochondrial dysfunctions due to oxidative and nitrosative stress in the
chronic pain or chronic fatigue syndromes and fibromyalgia patients:
Peripheral and central mechanisms as therapeutic targets? Expert Opinion on
Therapeutic Targets, 17(9), 1081-1089. doi:10.1517/14728222.2013.818657
Metzger, K., Fremont, M., Roelant, C., & De Meirleir, K. (2008). Lower frequency
of IL-17F sequence variant (His161Arg) in chronic fatigue syndrome
patients. Biochemical and Biophysical Research Communications, 376(1),
231-233. doi:10.1016/j.bbrc.2008.08.135
Mikovits, J. A., Lombardi, V. C., Pfost, M. A., Hagen, K. S., & Ruscetti, F. W.
(2010). Detection of an infectious retrovirus, XMRV, in blood cells of
patients with chronic fatigue syndrome. Virulence, 1(5), 386-390.
doi:10.4161/viru.1.5.12486
Miller, A. H., Jones, J. F., Drake, D. F., Tian, H., Unger, E. R., & Pagnoni, G.
(2014). Decreased basal ganglia activation in subjects with chronic fatigue
syndrome: Association with symptoms of fatigue. PloS One, 9(5), e98156.
doi:10.1371/journal.pone.0098156
Minică, C. C., Boomsma, D. I., Vink, J. M., & Dolan, C. V. (2014). MZ twin pairs or
MZ singletons in population family-based GWAS? More power in pairs.
Molecular Psychiatry, 19(11), 1154-1155. doi:10.1038/mp.2014.121
Miwa, K., & Fujita, M. (2008). Smatt heart syndrome in patients with chronic fatigue
syndrome. Clinical Cardiology, 31(7), 328-333. doi:10.1002/clc.20227
Miwa, K., & Fujita, M. (2009a). Cardiac function fluctuates during exacerbation and
remission in young adults with chronic fatigue syndrome and "small heart".
Journal of Cardiology, 54(1), 29-35. doi:10.1016/j.jjcc.2009.02.008
Miwa, K., & Fujita, M. (2009b). Cardiovascular dysfunction with low cardiac output
due to a small heart in patients with chronic fatigue syndrome. Internal
Medicine, 48(21), 1849-1854. doi:10.2169/internalmedicine.48.2347
Miwa, K., & Fujita, M. (2011). Small heart with low cardiac output for orthostatic
intolerance in patients with chronic fatigue syndrome. Clinical Cardiology,
34(12), 782-786. doi:10.1002/clc.20962
Morris, A. P., Lindgren, C. M., Zeggini, E., Timpson, N. J., Frayling, T. M.,
Hattersley, A. T., & McCarthy, M. I. (2010). A powerful approach to sub-
phenotype analysis in population-based genetic association studies. Genetic
Epidemiology, 34(4), 335-343. doi:10.1002/gepi.20486
Mosing, M. A., Medland, S. E., McRae, A., Landers, J. G., Wright, M. J., & Martin,
N. G. (2012). Genetic influences on life span and its relationship to
personality: A 16-year follow-up study of a sample of aging twins.
Psychosomatic Medicine, 74(1), 16-22. doi:10.1097/PSY.0b013e3182385784
182 Bibliography
Muglia, P., Tozzi, F., Galwey, N. W., Francks, C., Upmanyu, R., Kong, X. Q., . . .
Roses, A. D. (2010). Genome-wide association study of recurrent major
depressive disorder in two European case-control cohorts. Molecular
Psychiatry, 15(6), 589-601. doi:10.1038/mp.2008.131
Myhill, S., Booth, N. E., & McLaren-Howard, J. (2009). Chronic fatigue syndrome
and mitochondrial dysfunction. International Journal of Clinical and
Experimental Medicine, 2(1), 1-16.
Myhill, S., Booth, N. E., & McLaren-Howard, J. (2013). Targeting mitochondrial
dysfunction in the treatment of myalgic encephalomyelitis/chronic fatigue
syndrome (ME/CFS) - A clinical audit. International Journal of Clinical and
Experimental Medicine, 6(1), 1.
Nacul, L. C., Lacerda, E. M., Pheby, D., Campion, P., Molokhia, M., Fayyaz, S., . . .
Drachler, M. L. (2011). Prevalence of myalgic encephalomyelitis/chronic
fatigue syndrome (ME/CFS) in three regions of England: A repeated cross-
sectional study in primary care. BMC Medicine, 9, 91-102. doi:10.1186/1741-
7015-9-91
Nancy, A. L., & Shoenfeld, Y. (2008). Chronic fatigue syndrome with
autoantibodies—The result of an augmented adjuvant effect of hepatitis-B
vaccine and silicone implant. Autoimmun Rev, 8(1), 52-55.
doi:10.1016/j.autrev.2008.07.026
Narita, M., Nishigami, N., Narita, N., Yamaguti, K., Okado, N., Watanabe, Y., &
Kuratsune, H. (2003). Association between serotonin transporter gene
polymorphism and chronic fatigue syndrome. Biochemical and Biophysical
Research Communications, 311(2), 264-266. doi:10.1016/j.bbrc.2003.09.207
Narusyte, J., Ropponen, A., Alexanderson, K., & Svedberg, P. (2016). Genetic and
environmental influences on disability pension due to mental diagnoses:
Limited importance of major depression, generalized anxiety, and chronic
fatigue. Twin Research and Human Genetics, 19(1), 10-16.
doi:10.1017/thg.2015.86
Natelson, B. H., Cohen, J. M., Brassloff, I., & Lee, H. J. (1993). A controlled study
of brain magnetic resonance imaging in patients with the chronic fatigue
syndrome. Journal of the Neurological Sciences, 120(2), 213-217.
doi:10.1016/0022-510X(93)90276-5
Natelson, B. H., Haghighi, M. H., & Ponzio, N. M. (2002). Evidence for the presence
of immune dysfunction in chronic fatigue syndrome. Clinical and Diagnostic
Laboratory Immunology, 9(4), 747-752. doi:10.1128/CDLI.9.4.747-752.2002
Natelson, B. H., Intriligator, R., Cherniack, N. S., Chandler, H. K., & Stewart, J. M.
(2007). Hypocapnia is a biological marker for orthostatic intolerance in some
patients with chronic fatigue syndrome. Dynamic Medicine, 6, 2-2.
doi:10.1186/1476-5918-6-2
National Research Council (US), & Institute of Medicine (US) Committee on
Depression, Parenting Practices, and the Healthy Development of Children.
(2009). 3, The etiology of depression. In M. J. England & L. J. Sim (Eds.),
Depression in parents, parenting, and children: Opportunities to improve
identification, treatment, and prevention (pp. 73-118). Washington, DC:
National Academies Press (US).
Neale, M., Cardon, L., & Division, N. A. T. O. S. A. (1992). Methodology for
genetic studies of twins and families: Springer.
Nicolson, G. L., Gan, R., & Haier, J. (2003). Multiple co-infections (Mycoplasma,
Chlamydia, human herpes virus-6) in blood of chronic fatigue syndrome
Bibliography 183
patients: Association with signs and symptoms. APMIS, 111(5), 557-566.
doi:10.1034/j.1600-0463.2003.1110504.x
Nijs, J., Nicolson, G. L., De Becker, P., Coomans, D., & De Meirleir, K. (2002).
High prevalence of mycoplasma infections among European chronic fatigue
syndrome patients. Examination of four mycoplasma species in blood of
chronic fatigue syndrome patients. FEMS Immunology and Medical
Microbiology, 34(3), 209-214. doi:10.1111/j.1574-695X.2002.tb00626.x
Nisenbaum, R., Jones, J. F., Unger, E. R., Reyes, M., & Reeves, W. C. (2003). A
population-based study of the clinical course of chronic fatigue syndrome.
Health Qual Life Outcomes, 1, 49-57. doi:10.1186/1477-7525-1-49
Njoku, M. G. C., Leonard, A. J., & Torres-Harding, S. R. (2007). The prevalence of
chronic fatigue syndrome in Nigeria. Journal of Health Psychology, 12(3),
461-474. doi:10.1177/1359105307076233
Nolasco, S., Bellido, J., Goncalves, J., Zabala, J. C., & Soares, H. (2005). Tubulin
cofactor A gene silencing in mammalian cells induces changes in microtubule
cytoskeleton, cell cycle arrest and cell death. FEBS Letters, 579(17), 3515-
3524. doi:10.1016/j.febslet.2005.05.022
Nyholt, D. R. (2004). A simple correction for multiple testing for single-nucleotide
polymorphisms in linkage disequilibrium with each other. American Journal
of Human Genetics, 74(4), 765-769. doi:10.1086/383251
Nyholt, D. R., Gillespie, N. G., Heath, A. C., Merikangas, K. R., Duffy, D. L., &
Martin, N. G. (2004). Latent class and genetic analysis does not support
migraine with aura and migraine without aura as separate entities. Genetic
Epidemiology, 26(3), 231-244. doi:10.1002/gepi.10311
Okada, T., Tanaka, M., Kuratsune, H., Watanabe, Y., & Sadato, N. (2004).
Mechanisms underlying fatigue: A voxel-based morphometric study of
chronic fatigue syndrome. BMC Neurology, 4(1), 14-19. doi:10.1186/1471-
2377-4-14
Okbay, A., Baselmans, B. M. L., De Neve, J.-E., Turley, P., Nivard, M. G., Fontana,
M. A., . . . Cesarini, D. (2016). Genetic variants associated with subjective
well-being, depressive symptoms, and neuroticism identified through
genome-wide analyses. Nature Genetics, 48(6), 624-633.
doi:10.1038/ng.3552
Ortega-Hernandez, O. D., Cuccia, M., Bozzini, S., Bassi, N., Moscavitch, S., Diaz-
Gallo, L. M., . . . Shoenfeld, Y. (2009). Autoantibodies, polymorphisms in the
serotonin pathway, and human leukocyte antigen class II alleles in chronic
fatigue syndrome: Are they associated with age at onset and specific
symptoms? Annals of the New York Academy of Sciences, 1173, 589-599.
doi:10.1111/j.1749-6632.2009.04802.x
Ortega-Hernandez, O. D., & Shoenfeld, Y. (2009). Infection, vaccination, and
autoantibodies in chronic fatigue syndrome, cause or coincidence? Annals of
the New York Academy of Sciences, 1173, 600-609. doi:10.1111/j.1749-
6632.2009.04799.x
Østergaard, S. D., Jensen, S. O. W., & Bech, P. (2011). The heterogeneity of the
depressive syndrome: When numbers get serious. Acta Psychiatrica
Scandinavica, 124(6), 495-496. doi:10.1111/j.1600-0447.2011.01744.x
Patel, V., Kirkwood, B., Weiss, H., Pednekar, S., Fernandes, J., Pereira, B., . . .
Mabey, D. (2005). Chronic fatigue in developing countries: Population based
survey of women in India. BMJ: British Medical Journal, 330(7501), 1190-
1193. doi:10.2307/25459726
184 Bibliography
Pearson, R., Palmer, R. H., Brick, L. A., McGeary, J. E., Knopik, V. S., & Beevers,
C. G. (2016). Additive genetic contribution to symptom dimensions in major
depressive disorder. Journal of Abnormal Psychology, 125(4), 495-501.
doi:10.1037/abn0000161
Peerbooms, O. L., van Os, J., Drukker, M., Kenis, G., Hoogveld, L., de Hert, M., . . .
Rutten, B. P. (2011). Meta-analysis of MTHFR gene variants in
schizophrenia, bipolar disorder and unipolar depressive disorder: Evidence
for a common genetic vulnerability? Brain, Behavior, and Immunity, 25(8),
1530-1543. doi:10.1016/j.bbi.2010.12.006
Powell, R., Ren, J., Lewith, G., Barclay, W., Holgate, S., & Almond, J. (2003).
Identification of novel expressed sequences, up-regulated in the leucocytes of
chronic fatigue syndrome patients. Clinical and Experimental Allergy,
33(10), 1450-1456. doi:10.1046/j.1365-2222.2003.01745.x
Power, R. A., Keers, R., Ng, M. Y., Butler, A. W., Uher, R., Cohen-Woods, S., . . .
Lewis, C. M. (2012). Dissecting the genetic heterogeneity of depression
through age at onset. American Journal of Medical Genetics. Part B:
Neuropsychiatric Genetics, 159b(7), 859-868. doi:10.1002/ajmg.b.32093
Power, R. A., Cohen-Woods, S., Ng, M. Y., Butler, A. W., Craddock, N., Korszun,
A., . . . Uher, R. (2013). Genome-wide association analysis accounting for
environmental factors through propensity-score matching: Application to
stressful live events in major depressive disorder. American Journal of
Medical Genetics. Part B: Neuropsychiatric Genetics, 162b(6), 521-529.
doi:10.1002/ajmg.b.32180
Power, R. A., Keller, M. C., Ripke, S., Abdellaoui, A., Wray, N. R., Sullivan, P. F.,
& Breen, G. (2014). A recessive genetic model and runs of homozygosity in
major depressive disorder. American Journal of Medical Genetics. Part B:
Neuropsychiatric Genetics, 165b(2), 157-166. doi:10.1002/ajmg.b.32217
Power, R. A., Tansey, K. E., Buttenschon, H. N., Cohen-Woods, S., Bigdeli, T., Hall,
L. S., . . . Lewis, C. M. (2017). Genome-wide association for major
depression through age at onset stratification: Major Depressive Disorder
Working Group of the Psychiatric Genomics Consortium. Biological
Psychiatry, 81(4), 325-335. doi:10.1016/j.biopsych.2016.05.010
Prins, J. B., van der Meer, J. W., & Bleijenberg, G. (2006). Chronic fatigue
syndrome. Lancet, 367(9507), 346-355. doi:10.1016/s0140-6736(06)68073-2
Pruim, R. J., Welch, R. P., Sanna, S., Teslovich, T. M., Chines, P. S., Gliedt, T. P., . .
. Willer, C. J. (2010). LocusZoom: Regional visualization of genome-wide
association scan results. Bioinformatics, 26(18), 2336-2337.
doi:10.1093/bioinformatics/btq419
Purandare, N., Burns, A., Craig, S., Faragher, B., & Scott, K. (2001). Depressive
symptoms in patients with Alzheimer's disease. International Journal of
Geriatric Psychiatry, 16(10), 960-964. doi:10.1002/gps.449
Purcell, S., Cherny, S. S., & Sham, P. C. (2003). Genetic power calculator: Design of
linkage and association genetic mapping studies of complex traits.
Bioinformatics, 19(1), 149-150.
Purcell, S., Neale, B., Todd-Brown, K., Thomas, L., Ferreira, M. A. R., Bender, D., .
. . Sham, P. C. (2007). PLINK: A tool set for whole-genome association and
population-based linkage analyses. American Journal of Human Genetics,
81(3), 559-575. doi:10.1086/519795
Bibliography 185
R Core Team. (2014). R: A language and environment for statistical computing.
Vienna, Austria: R Foundation for Statistical Computing. Retrieved from
http://www.R-project.org/
R Development Core Team. (2003). R: A language and environment for statistical
computing. Vienna: R Foundation for Statistical Computing. Retrieved from
http://www.R-project.org
Rajeevan, M. S., Smith, A. K., Dimulescu, I., Unger, E. R., Vernon, S. D., Heim, C.,
& Reeves, W. C. (2007). Glucocorticoid receptor polymorphisms and
haplotypes associated with chronic fatigue syndrome. Genes Brain Behav,
6(2), 167-176. doi:10.1111/j.1601-183X.2006.00244.x
Rajeevan, M. S., Dimulescu, I., Murray, J., Falkenberg, V. R., & Unger, E. R.
(2015). Pathway-focused genetic evaluation of immune and inflammation
related genes with chronic fatigue syndrome. Human Immunology, 76(8),
553-560. doi:10.1016/j.humimm.2015.06.014
Rapaport, M. H., Judd, L. L., Schettler, P. J., Yonkers, K. A., Thase, M. E., Kupfer,
D. J., . . . Rush, A. J. (2002). A descriptive analysis of minor depression.
American Journal of Psychiatry, 159(4), 637-643.
doi:10.1176/appi.ajp.159.4.637
Reeves, W. C., Wagner, D., Nisenbaum, R., Jones, J. F., Gurbaxani, B., Solomon, L.,
. . . Heim, C. (2005). Chronic fatigue syndrome - A clinically empirical
approach to its definition and study. BMC Medicine, 3, 19-27.
doi:10.1186/1741-7015-3-19
Reich, T., James, J. W., & Morris, C. A. (1972). The use of multiple thresholds in
determining the mode of transmission of semi-continuous traits. Annals of
Human Genetics, 36(2), 163-184. doi:10.1111/j.1469-1809.1972.tb00767.x
Reyes, M., Nisenbaum, R., Hoaglin, D. C., Unger, E. R., Emmons, C., Randall, B., . .
. Reeves, W. C. (2003). Prevalence and incidence of chronic fatigue
syndrome in Wichita, Kansas. Archives of Internal Medicine, 163(13), 1530-
1536. doi:10.1001/archinte.163.13.1530
Rietschel, M., Mattheisen, M., Frank, J., Treutlein, J., Degenhardt, F., Breuer, R., . . .
Cichon, S. (2010). Genome-wide association-, replication-, and neuroimaging
study implicates HOMER1 in the etiology of major depression. Biological
Psychiatry, 68(6), 578-585. doi:10.1016/j.biopsych.2010.05.038
Robinson, R. L., Stephenson, J. J., Dennehy, E. B., Grabner, M., Faries, D., Palli, S.
R., & Swindle, R. W. (2015). The importance of unresolved fatigue in
depression: Costs and comorbidities. Psychosomatics, 56(3), 274-285.
doi:10.1016/j.psym.2014.08.003
Rosenthal, T. C., Majeroni, B. A., Pretorius, R., & Malik, K. (2008). Fatigue: An
overview. American Family Physician, 78(10), 1173-1180.
Schlauch, K. A., Khaiboullina, S. F., De Meirleir, K. L., Rawat, S., Petereit, J.,
Rizvanov, A. A., . . . Lombardi, V. C. (2016). Genome-wide association
analysis identifies genetic variations in subjects with myalgic
encephalomyelitis/chronic fatigue syndrome. Transl Psychiatry, 6, e730.
doi:10.1038/tp.2015.208
Schoeman, E. M., Van Der Westhuizen, F. H., Erasmus, E., van Dyk, E., Knowles,
C. V., Al-Ali, S., . . . Elson, J. L. (2017). Clinically proven mtDNA mutations
are not common in those with chronic fatigue syndrome. BMC Medical
Genetics, 18(1), 29-32. doi:10.1186/s12881-017-0387-6
Schuch, J. J., Roest, A. M., Nolen, W. A., Penninx, B. W., & de Jonge, P. (2014).
Gender differences in major depressive disorder: Results from the
186 Bibliography
Netherlands study of depression and anxiety. Journal of Affective Disorders,
156, 156-163. doi:10.1016/j.jad.2013.12.011
Schur, E., Afari, N., Goldberg, J., Buchwald, D., & Sullivan, P. F. (2007). Twin
analyses of fatigue. Twin Res Hum Genet, 10(5), 729-733.
doi:10.1375/twin.10.5.729
Seet, R. C., Quek, A. M., & Lim, E. C. (2007). Post-infectious fatigue syndrome in
dengue infection. Journal of Clinical Virology, 38(1), 1-6.
doi:10.1016/j.jcv.2006.10.011
Seishima, M., Mizutani, Y., Shibuya, Y., & Arakawa, C. (2008). Chronic fatigue
syndrome after human parvovirus B19 infection without persistent viremia.
Dermatology, 216(4), 341-346. doi:10.1159/000116723
Shapero, B. G., Black, S. K., Liu, R. T., Klugman, J., Bender, R. E., Abramson, L.
Y., & Alloy, L. B. (2014). Stressful life events and depression symptoms:
The effect of childhood emotional abuse on stress reactivity. Journal of
Clinical Psychology, 70(3), 209-223. doi:10.1002/jclp.22011
Shapiro, J. S. (2009). Does varicella-zoster virus infection of the peripheral ganglia
cause chronic fatigue syndrome? Medical Hypotheses, 73(5), 728-734.
doi:10.1016/j.mehy.2009.04.043
Sharpe, M., & Wilks, D. (2002). ABC of psychological medicine - Fatigue. British
Medical Journal, 325(7362), 480-483. doi:10.1136/bmj.325.7362.480
Shi, J., Potash, J. B., Knowles, J. A., Weissman, M. M., Coryell, W., Scheftner, W.
A., . . . Levinson, D. F. (2011). Genome-wide association study of recurrent
early-onset major depressive disorder. Molecular Psychiatry, 16(2), 193-201.
doi:10.1038/mp.2009.124
Shimosako, N., & Kerr, J. R. (2014). Use of single-nucleotide polymorphisms
(SNPs) to distinguish gene expression subtypes of chronic fatigue
syndrome/myalgic encephalomyelitis (CFS/ME). Journal of Clinical
Pathology, 67(12), 1078-1083. doi:10.1136/jclinpath-2014-202597
Shyn, S. I., Shi, J., Kraft, J. B., Potash, J. B., Knowles, J. A., Weissman, M. M., . . .
Hamilton, S. P. (2011). Novel loci for major depression identified by
genome-wide association study of sequenced treatment alternatives to relieve
depression and meta-analysis of three studies. Molecular Psychiatry, 16(2),
202-215. doi:10.1038/mp.2009.125
Sigmon, S. T., Pells, J. J., Boulard, N. E., Whitcomb-Smith, S., Edenfield, T. M.,
Hermann, B. A., . . . Kubik, E. (2005). Gender differences in self-reports of
depression: The response bias hypothesis revisited. Sex Roles, 53(5), 401-
411. doi:10.1007/s11199-005-6762-3
Skapinakis, P., Lewis, G., & Mavreas, V. (2004). Temporal relations between
unexplained fatigue and depression: Longitudinal data from an international
study in primary care. Psychosomatic Medicine, 66(3), 330-335.
doi:10.1097/01.psy.0000124757.10167.b1
Smith, A. K., White, P. D., Aslakson, E., Vollmer-Conna, U., & Rajeevan, M. S.
(2006). Polymorphisms in genes regulating the HPA axis associated with
empirically delineated classes of unexplained chronic fatigue.
Pharmacogenomics, 7(3), 387-394. doi:10.2217/14622416.7.3.387
Smith, A. K., Dimulescu, I., Falkenberg, V. R., Narasimhan, S., Heim, C., Vernon, S.
D., & Rajeevan, M. S. (2008). Genetic evaluation of the serotonergic system
in chronic fatigue syndrome. Psychoneuroendocrinology, 33(2), 188-197.
doi:10.1016/j.psyneuen.2007.11.001
Bibliography 187
Smith, A. K., Maloney, E. M., Falkenberg, V. R., Dimulescu, I., & Rajeevan, M. S.
(2009). An angiotensin-1 converting enzyme polymorphism is associated
with allostatic load mediated by C-reactive protein, interleukin-6 and cortisol.
Psychoneuroendocrinology, 34(4), 597-606.
doi:10.1016/j.psyneuen.2008.10.022
Smith, A. K., Fang, H., Whistler, T., Unger, E. R., & Rajeevan, M. S. (2011).
Convergent genomic studies identify association of GRIK2 and NPAS2 with
chronic fatigue syndrome. Neuropsychobiology, 64(4), 183-194.
doi:10.1159/000326692
Smith, J., Fritz, E. L., Kerr, J. R., Cleare, A. J., Wessely, S., & Mattey, D. L. (2005).
Association of chronic fatigue syndrome with human leucocyte antigen class
II alleles. Journal of Clinical Pathology, 58(8), 860-863.
doi:10.1136/jcp.2004.022681
Sommerfeldt, L., Portilla, H., Jacobsen, L., Gjerstad, J., & Wyller, V. B. (2011).
Polymorphisms of adrenergic cardiovascular control genes are associated
with adolescent chronic fatigue syndrome. Acta Paediatrica, 100(2), 293-
298. doi:10.1111/j.1651-2227.2010.02072.x
Sorenson, M. R., & Jason, L. (2013). Dysregulation of the HPA axis in chronic
fatigue syndrome. Advances in Neuroimmune Biology, 4(4), 275-279.
doi:10.3233/NIB-130071
Souza-Teodoro, L. H., de Oliveira, C., Walters, K., & Carvalho, L. A. (2016). Higher
serum dehydroepiandrosterone sulfate protects against the onset of depression
in the elderly: Findings from the English Longitudinal Study of Aging
(ELSA). Psychoneuroendocrinology, 64, 40-46.
doi:10.1016/j.psyneuen.2015.11.005
Steele, L., Dobbins, J. G., Fukuda, K., Reyes, M., Randall, B., Koppelman, M., &
Reeves, W. C. (1998). The epidemiology of chronic fatigue in San Francisco.
American Journal of Medicine, 105(3 PART A), 83S-90S.
doi:10.1016/s0002-9343(98)00158-2
Strauss, B., Loschau, M., Seidel, T., Stallmach, A., & Thomas, A. (2012). Are
fatigue symptoms and chronic fatigue syndrome following Q fever infection
related to psychosocial variables? Journal of Psychosomatic Research, 72(4),
300-304. doi:10.1016/j.jpsychores.2012.01.010
Sullivan, P. (2012). Don't give up on GWAS. Molecular Psychiatry, 17(1), 2-3.
doi:10.1038/mp.2011.94
Sullivan, P. F., Neale, M. C., & Kendler, K. S. (2000). Genetic epidemiology of
major depression: Review and meta-analysis. American Journal of
Psychiatry, 157(10), 1552-1562. doi:10.1176/appi.ajp.157.10.1552
Sullivan, P. F., Kovalenko, P., York, T. P., Prescott, C. A., & Kendler, K. S. (2003).
Fatigue in a community sample of twins. Psychological Medicine, 33(2), 263-
281. doi:10.1017/S0033291702007031
Sullivan, P. F., Evengard, B., Jacks, A., & Pedersen, N. L. (2005). Twin analyses of
chronic fatigue in a Swedish national sample. Psychological Medicine, 35(9),
1327-1336. doi:10.1017/s0033291705005222
Sullivan, P. F., de Geus, E. J., Willemsen, G., James, M. R., Smit, J. H., Zandbelt, T.,
. . . Penninx, B. W. (2009). Genome-wide association for major depressive
disorder: A possible role for the presynaptic protein piccolo. Molecular
Psychiatry, 14(4), 359-375. doi:10.1038/mp.2008.125
Tang, L. W., Zheng, H., Chen, L., Zhou, S. Y., Huang, W. J., Li, Y., & Wu, X.
(2015). Gray matter volumes in patients with chronic fatigue syndrome.
188 Bibliography
Evidence-Based Complementary and Alternative Medicine, 2015, 380615.
doi:10.1155/2015/380615
Terracciano, A., Tanaka, T., Sutin, A. R., Sanna, S., Deiana, B., Lai, S., . . . Costa, P.
T., Jr. (2010). Genome-wide association scan of trait depression. Biological
Psychiatry, 68(9), 811-817. doi:10.1016/j.biopsych.2010.06.030
Thase, M. E., Hahn, C., & Berton, O. (2014). 10, Neurobiological aspects of
depression. In I. H. Gotlib & C. L. Hammen (Eds.), Handbook of Depression
(3 ed., pp. 182-201). New York, US: Guilford Publications.
The 1000 Genomes Project Consortium. (2015). A global reference for human
genetic variation. Nature, 526(7571), 68-74. doi:10.1038/nature15393
Tian, G., Huang, Y., Rommelaere, H., Vandekerckhove, J., Ampe, C., & Cowan, N.
J. (1996). Pathway leading to correctly folded beta-tubulin. Cell, 86(2), 287-
296.
Tomas, C., Newton, J., & Watson, S. (2013). A review of hypothalamic-pituitary-
adrenal axis function in chronic fatigue syndrome. ISRN Neurosci, 2013,
784520. doi:10.1155/2013/784520
Treadway, M. T., & Pizzagalli, D. A. (2014). Imaging the pathophysiology of major
depressive disorder - from localist models to circuit-based analysis. Biol
Mood Anxiety Disord, 4(1), 5. doi:10.1186/2045-5380-4-5
Tritt, S. M., Ryder, A. G., Ring, A. J., & Pincus, A. L. (2010). Pathological
narcissism and the depressive temperament. Journal of Affective Disorders,
122(3), 280-284. doi:10.1016/j.jad.2009.09.006
Tsuang, M. T., Winokur, G., & Crowe, R. R. (1980). Morbidity risks of
schizophrenia and affective disorders among first degree relatives of patients
with schizophrenia, mania, depression and surgical conditions. British
Journal of Psychiatry, 137, 497-504. doi:10.1192/bjp.137.6.497
Uebersax J.S. (2007). User Guide for POLYCORR 1.1. Retrieved from http://john-
uebersax.com/stat/xpc.htm
van de Putte, E. M., van Doornen, L. J., Engelbert, R. H., Kuis, W., Kimpen, J. L., &
Uiterwaal, C. S. (2006). Mirrored symptoms in mother and child with chronic
fatigue syndrome. Pediatrics, 117(6), 2074-2079. doi:10.1542/peds.2005-
2307
Van Der Linden, G., Chalder, T., Hickie, I., Koschera, A., Sham, P., & Wessely, S.
(1999). Fatigue and psychiatric disorder: Different or the same?
Psychological Medicine, 29(4), 863-868. doi:10.1017/S0033291799008697
Verduijn, J., Milaneschi, Y., Schoevers, R. A., van Hemert, A. M., Beekman, A. T.
F., & Penninx, B. W. J. H. (2015). Pathophysiology of major depressive
disorder: mechanisms involved in etiology are not associated with clinical
progression. Transl Psychiatry, 5, e649. doi:10.1038/tp.2015.137
Vernon, S. D., Shukla, S. K., & Reeves, W. C. (2003). Absence of mycoplasma
species DNA in chronic fatigue syndrome. Journal of Medical Microbiology,
52(Pt 11), 1027-1028. doi:10.1099/jmm.0.05316-0
Vernon, S. D., Whistler, T., Cameron, B., Hickie, I. B., Reeves, W. C., & Lloyd, A.
(2006). Preliminary evidence of mitochondrial dysfunction associated with
post-infective fatigue after acute infection with Epstein Barr virus. BMC
Infectious Diseases, 6, 15-21. doi:10.1186/1471-2334-6-15
Vincent, A., Brimmer, D. J., Whipple, M. O., Jones, J. F., Boneva, R., Lahr, B. D., . .
. Reeves, W. C. (2012). Prevalence, incidence, and classification of chronic
fatigue syndrome in Olmsted County, Minnesota, as estimated using the
Bibliography 189
Rochester Epidemiology Project. Mayo Clinic Proceedings, 87(12), 1145-
1152. doi:10.1016/j.mayocp.2012.08.015
Vink, J. M., Bartels, M., van Beijsterveldt, T. C. E. M., van Dongen, J., van Beek, J.
H. D. A., Distel, M. A., . . . Boomsma, D. I. (2012). Sex differences in
genetic architecture of complex phenotypes? PloS One, 7(12), e47371.
doi:10.1371/journal.pone.0047371
Visscher, P. M., Brown, M. A., McCarthy, M. I., & Yang, J. (2012). Five years of
GWAS discovery. American Journal of Human Genetics, 90(1), 7-24.
doi:10.1016/j.ajhg.2011.11.029
Vladutiu, G. D., & Natelson, B. H. (2004). Association of medically unexplained
fatigue with ACE insertion/deletion polymorphism in Gulf War veterans.
Muscle and Nerve, 30(1), 38-43. doi:10.1002/mus.20055
Vojdani, A., Choppa, P. C., Tagle, C., Andrin, R., Samimi, B., & Lapp, C. W.
(1998). Detection of Mycoplasma genus and Mycoplasma fermentans by
PCR in patients with Chronic Fatigue Syndrome. FEMS Immunology and
Medical Microbiology, 22(4), 355-365. doi:10.1111/j.1574-
695X.1998.tb01226.x
Walker, E. A., Katon, W. J., & Jemelka, R. P. (1993). Psychiatric disorders and
medical care utilization among people in the general population who report
fatigue. Journal of General Internal Medicine, 8(8), 436-440.
doi:10.1007/BF02599621
Walsh, C. M., Zainal, N. Z., Middleton, S. J., & Paykel, E. S. (2001). A family
history study of chronic fatigue syndrome. Psychiatric Genetics, 11(3), 123-
128. doi:10.1097/00041444-200109000-00003
Weissman, M. M., Gershon, E. S., Kidd, K. K., Prusoff, B. A., Leckman, J. F.,
Dibble, E., . . . Guroff, J. J. (1984). Psychiatric disorders in the relatives of
probands with affective disorders. The Yale University—National Institute of
Mental Health Collaborative Study. Archives of General Psychiatry, 41(1),
13-21. doi:10.1001/archpsyc.1984.01790120015003
Weissman, M. M., Wickramaratne, P., Adams, P. B., Lish, J. D., Horwath, E.,
Charney, D., . . . Frosch, E. (1993). The relationship between panic disorder
and major depression. A new family study. Archives of General Psychiatry,
50(10), 767-780. doi:10.1001/archpsyc.1993.01820220017003
Wessely, S., Chalder, T., Hirsch, S., Pawlikowska, T., Wallace, P., & Wright, D. J.
M. (1995). Postinfectious fatigue - Prospective cohort study in primary-care.
Lancet, 345(8961), 1333-1338. doi:10.1016/s0140-6736(95)92537-6
Wessely, S., Chalder, T., Hirsch, S., Wallace, P., & Wright, D. (1997). The
prevalence and morbidity of chronic fatigue and chronic fatigue syndrome: A
prospective primary care study. American Journal of Public Health, 87(9),
1449-1455.
White, P. D., Thomas, J. M., Amess, J., Crawford, D. H., Grover, S. A., Kangro, H.
O., & Clare, A. W. (1998). Incidence, risk and prognosis of acute and chronic
fatigue syndromes and psychiatric disorders after glandular fever. British
Journal of Psychiatry, 173, 475-481. doi:10.1192/bjp.173.6.475
White, P. D., Thomas, J. M., Kangro, H. O., Bruce-Jones, W. D., Amess, J.,
Crawford, D. H., . . . Clare, A. W. (2001). Predictions and associations of
fatigue syndromes and mood disorders that occur after infectious
mononucleosis. Lancet, 358(9297), 1946-1954. doi:10.1016/s0140-
6736(01)06961-6
190 Bibliography
Wildman, M. J., Smith, E. G., Groves, J., Beattie, J. M., Caul, E. O., & Ayres, J. G.
(2002). Chronic fatigue following infection by Coxiella burnetii (Q fever):
Ten-year follow-up of the 1989 UK outbreak cohort. QJM, 95(8), 527-538.
doi:10.1093/qjmed/95.8.527
Wilkowska-Chmielewska, J., Szelenberger, W., & Wojnar, M. (2013). Age-
dependent symptomatology of depression in hospitalized patients and its
implications for DSM-5. Journal of Affective Disorders, 150(1), 142-145.
doi:10.1016/j.jad.2012.12.012
Williamson, R. J., Purcell, S., Sterne, A., Wessely, S., Hotopf, M., Farmer, A., &
Sham, P. C. (2005). The relationship of fatigue to mental and physical health
in a community sample. Social Psychiatry and Psychiatric Epidemiology,
40(2), 126-132. doi:10.1007/s00127-005-0858-5
Wong, W. S., & Fielding, R. (2010). Prevalence of chronic fatigue among Chinese
adults in Hong Kong: A population-based study. Journal of Affective
Disorders, 127(1-3), 248-256. doi:10.1016/j.jad.2010.04.029
World Health Organization. (1990). Composite international diagnostic interview
(CIDI): World Health Organization Division of Mental Health.
Wray, N. R., Pergadia, M. L., Blackwood, D. H., Penninx, B. W., Gordon, S. D.,
Nyholt, D. R., . . . Sullivan, P. F. (2012). Genome-wide association study of
major depressive disorder: New results, meta-analysis, and lessons learned.
Molecular Psychiatry, 17(1), 36-48. doi:10.1038/mp.2010.109
Xu, W., Zhu, H., Gu, M., Luo, Q., Ding, J., Yao, Y., . . . Wang, Z. (2013). DHTKD1
is essential for mitochondrial biogenesis and function maintenance. FEBS
Letters, 587(21), 3587-3592. doi:10.1016/j.febslet.2013.08.047
Yang, J., Zaitlen, N. A., Goddard, M. E., Visscher, P. M., & Price, A. L. (2014).
Advantages and pitfalls in the application of mixed-model association
methods. Nature Genetics, 46(2), 100-106. doi:10.1038/ng.2876
Yang, Y., Zhao, H., Heath, A. C., Madden, P. A., Martin, N. G., & Nyholt, D. R.
(2016). Shared genetic factors underlie migraine and depression. Twin Res
Hum Genet, 19(4), 341-350. doi:10.1017/thg.2016.46
Zeng, Y., Navarro, P., Xia, C., Amador, C., Fernandez-Pujals, A. M., Thomson, P.
A., . . . McIntosh, A. M. (2016). Shared genetics and couple-associated
environment are major contributors to the risk of both clinical and self-
declared depression. EBioMedicine, 14, 161-167.
doi:10.1016/j.ebiom.2016.11.003
Zhang, Q. W., Natelson, B. H., Ottenweller, J. E., Servatius, R. J., Nelson, J. J., De
Luca, J., . . . Lange, G. (2000). Chronic fatigue syndrome beginning suddenly
occurs seasonally over the year. Chronobiology International, 17(1), 95-99.
doi:10.1081/CBI-100101035
Zhou, X., & Stephens, M. (2012). Genome-wide efficient mixed-model analysis for
association studies. Nature Genetics, 44(7), 821-824. doi:10.1038/ng.2310
Zimmerman, M., Ellison, W., Young, D., Chelminski, I., & Dalrymple, K. (2015).
How many different ways do patients meet the diagnostic criteria for major
depressive disorder? Comprehensive Psychiatry, 56, 29-34.
doi:10.1016/j.comppsych.2014.09.007
Zintzaras, E. (2006). C677T and A1298C methylenetetrahydrofolate reductase gene
polymorphisms in schizophrenia, bipolar disorder and depression: A meta-
analysis of genetic association studies. Psychiatric Genetics, 16(3), 105-115.
doi:10.1097/01.ypg.0000199444.77291.e2
Appendices 191
Appendices
Appendix A
Chapter 3: Supplementary Data
Supplementary Table 3.1. Fatigue and depression symptom counts in the phenotypic subgroups.
Symptom
Non-depressed,
non-fatigued
(N = 1,750)
Non-depressed,
fatigued
(N = 590)
MiDD,
non-fatigued
(N = 83)
MDD,
non-fatigued
(N = 16)
MiDD,
fatigued
(N = 142)
MDD,
fatigued
(N = 34)
Complete cohort
(N = 2,615)
Fatigue symptoms Muscle pain at rest 40 (2.3) 194 (32.9) 5 (6.0) 0 (0.0) 48 (33.8) 12 (35.3) 299 (11.4)
Post-exertional muscle pain 213 (12.2) 392 (66.4) 8 (9.6) 1 (6.3) 95 (66.9) 19 (55.9) 728 (27.8)
Post-exertional muscle fatigue 211 (12.1) 486 (82.4) 11 (13.3) 1 (6.3) 102 (71.8) 27 (79.4) 838 (32.0) Post-exertional fatigue 75 (4.3) 338 (57.3) 9 (10.8) 1 (6.3) 85 (59.9) 22 (64.7) 530 (20.3)
Hypersomnia 116 (6.6) 242 (41.0) 4 (4.8) 0 (0.0) 59 (41.5) 18 (52.9) 439 (16.8)
Insomnia 53 (3.0) 77 (13.1) 22 (26.5) 11 (68.8) 53 (37.3) 26 (76.5) 242 (9.3) Poor concentration 37 (2.1) 85 (14.4) 12 (14.5) 7 (43.8) 68 (47.9) 23 (67.6) 232 (8.9)
Speech problems 112 (6.4) 195 (33.1) 2 (2.4) 0 (0.0) 48 (33.8) 14 (41.2) 371 (14.2)
Poor memory 115 (6.6) 224 (38.0) 6 (7.2) 0 (0.0) 58 (40.8) 15 (44.1) 418 (16.0) Headaches 191 (10.9) 234 (39.7) 13 (15.7) 1 (6.3) 75 (52.8) 16 (47.1) 530 (20.3)
Depression symptoms Depressed mood 38 (2.2) 17 (2.9) 59 (71.1) 16 (100.0) 104 (73.2) 32 (94.1) 266 (10.2)
Anhedonia 51 (2.9) 44 (7.5) 61 (73.5) 16 (100.0) 95 (66.9) 33 (97.1) 300 (11.2)
Insomnia 8 (0.5) 17 (2.9) 11 (13.3) 14 (87.5) 35 (24.6) 30 (88.2) 115 (4.4)
Psychomotor agitation 2 (0.1) 6 (1.0) 3 (3.6) 6 (37.5) 18 (12.7) 18 (52.9) 53 (2.0)
Loss of energy 3 (0.2) 2 (0.3) 5 (6.0) 7 (43.8) 8 (5.6) 17 (50.0) 42 (1.6)
Feeling worthless 55 (3.1) 50 (8.5) 51 (61.4) 16 (100.0) 94 (66.2) 33 (97.1) 299 (11.4) Inability to concentrate 19 (1.1) 22 (3.7) 22 (26.5) 9 (56.3) 37 (26.1) 23 (67.6) 132 (5.0)
Suicidal thoughts 1 (0.1) 1 (0.2) 4 (4.8) 7 (43.8) 11 (7.7) 17 (50.0) 41 (1.6)
For each of the seven phenotypic subgroups, the number of individuals and proportion of individuals (in brackets) reporting a specific symptom.
192 Appendices
Appendix B
Chapter 6: Supplementary Data
Supplementary Table 6.1. Cross-tabulationa of depression and fatigue status within twin pairs.
Non-depressed MiDD MDD Total
Complete twin pairs
Non-fatigued 784 61 14.5 859.5
Fatigued 301.5 48.5 10.5 360.5
Total 1085.5 109.5 25 1220
MZ
Total
Non-fatigued 411 29.5 5.5 446
Fatigued 157.5 31.5 8 197 Total 568.5 61 13.5 643
MZ
female
Non-fatigued 307.5 19 5 331.5
Fatigued 124 28.5 7 159.5
Total 431.5 47.5 12 491
MZ
male
Non-fatigued 103.5 10.5 0.5 114.5
Fatigued 33.5 3 1 37.5 Total 137 13.5 1.5 152
DZss
Total
Non-fatigued 212 16.5 6 234.5
Fatigued 90.5 10.5 0.5 101.5 Total 302.5 27 6.5 336
DZss
female
Non-fatigued 164.5 15 4 183.5
Fatigued 70 9 0.5 79.5 Total 234.5 24 4.5 263
DZss
male
Non-fatigued 47.5 1.5 2 51
Fatigued 20.5 1.5 0 22 Total 68 3 2 73
DZos
female-male
Non-fatigued 163 11 5 179
Fatigued 54 7 1 62 Total 217 18 6 241
DZos
male-female
Non-fatigued 159 19 1 179
Fatigued 53 6 3 62
Total 212 25 4 241 aTables were made symmetrical in same-sex twin pairs by averaging over using either twin 1 or twin 2 as proband. For
example, within the complete twin pairs there was 298 twin pairs where twin 1 was fatigued and twin 2 was non-depressed and 305 twin pairs where twin 2 was fatigued and twin 1 was non-depressed. Therefore, the cross-tabulation averaging over
twin 1 or twin 2 as proband is (298+305)/2=301.5.
Appendices 193
Supplementary Table 6.2. Relative riska of depression and fatigue in males and females.
Proband–co-twin Depressed–non-fatigued Depressed–fatigued Fatigued–non-depressed Fatigued–depressed
Complete cohort (N = 1220) 0.78 (0.67-0.91) 1.58 (1.28-1.96) 0.92 (0.87-0.96) 1.86 (1.36-2.56)
MZ total (N = 643) 0.65 (0.51-0.83) 1.91 (1.49-2.46) 0.87 (0.80-0.94) 2.56 (1.67-3.90)
MZ female (N = 491) 0.57 (0.41-0.78) 2.08 (1.61-2.68) 0.84 (0.77-0.92) 3.07 (1.90-4.98) MZ male (N = 152) 0.97 (0.71-1.34) 1.09 (0.45-2.65) 0.99 (0.87-1.12) 1.11 (0.38-3.28)
DZss total (N = 336) 0.96 (0.75-1.23) 1.10 (0.66-1.84) 0.99 (0.91-1.07) 1.13 (0.57-2.23)
DZss female (N = 262) 0.95 (0.72-1.25) 1.12 (0.64-1.95) 0.98 (0.89-1.08) 1.15 (0.55-2.40) DZss male (N = 73) 1.00 (0.55-1.82) 1.00 (0.25-3.98) 1.00 (0.87-1.15) 0.99 (0.16-6.30)
DZos female-male (N = 241) 0.88 (0.67-1.16) 1.37 (0.77-2.46) 0.96 (0.86-1.06) 0.96 (0.86-1.08)
DZos male-female (N = 241) 0.92 (0.71-1.19) 1.34 (0.73-2.47) 1.44 (0.65-3.21) 1.30 (0.62-2.70) aRelative risks were calculated with respect to non-depressed or non-fatigued status in twin 1.
Supplementary Table 6.3. Relative riska of depression and fatigue within monozygotic (MZ), same-sex dizygotic (DZss), and opposite-sex dizygotic (DZos) twin pairs.
Proband–co-twin MZ (N = 643) DZss (N = 336) DZos (N = 241)
(Female-Male)
DZos (N = 241)
(Male-Female)
MiDD–non-fatigued 0.67 (0.52-0.87) 0.87 (0.64-1.18) 1.01 (0.80-1.28) 0.81 (0.56-1.19)
MiDD–fatigued 1.86 (1.41-2.45) 1.31 (0.79-2.16) 0.96 (0.46-2.00) 1.56 (0.84-2.92) MDD–non-fatigued 0.56 (0.30-1.07) 1.22 (0.89-1.66) 0.33 (0.06-1.82) 1.11 (0.77-1.60)
MDD–fatigued 2.14 (1.35-3.39) 0.48 (0.08-2.98) 3.00 (1.63-5.53) 0.67 (0.11-4.07)
Fatigued–MiDD 2.41 (1.51-3.86) 1.48 (0.71-3.09) 1.84 (0.75-4.53) 0.91 (0.38-2.18) Fatigued–MDD 3.29 (1.12-9.61) 0.39 (0.05-3.17) 0.58 (0.07-4.85) 8.66 (0.92-81.74) aRelative risks were calculated with respect to non-depressed or non-fatigued status in twin 1.
194 Appendices
Supplementary Table 6.4. Polychoric correlations with their 95% confidence intervals for fatigue
and depression in twin pairs according to zygosity.
Twin 1 Twin 2
Minor depression (non-depressed, MiDD)
MiDD Fatigue MiDD Fatigue Monozygotic twin pairs (N = 643 pairs)
Twin 1 MiDD 1.00
Fatigue 0.49 (0.35-0.64)a 1.00 Twin 2 MiDD 0.37 (0.17-0.56)b 0.32 (0.16-0.47)c 1.00
Fatigue 0.32 (0.16-0.49)c 0.43 (0.31-0.54)b 0.42 (0.28-0.57)a 1.00
Dizygotic twin pairs (N = 577 pairs) Twin 1 MiDD 1.00
Fatigue 0.49 (0.34-0.64)a 1.00
Twin 2 MiDD 0.21 (-0.04-0.45)b 0.18 (-0.01-0.37)c 1.00 Fatigue 0.03 (-0.17-0.22)c 0.14 (0.001-0.28)b 0.54 (0.40-0.69)a 1.00
Major depression (non-depressed, MDD)
MDD Fatigue MDD Fatigue
Monozygotic twin pairs (N = 643 pairs)
Twin 1 MDD 1.00
Fatigue 0.36 (0.12-0.59)a 1.00 Twin 2 MDD 0.46 (-0.01-0.93)b 0.43 (0.15-0.70)c 1.00
Fatigue 0.28 (0.03-0.52)c 0.43 (0.31-0.54)b -a 1.00
Dizygotic twin pairs (N = 577 pairs) Twin 1 MDD 1.00
Fatigue 0.47 (0.20-0.74)a 1.00 Twin 2 MDD -b -0.32 (-0.67-0.02)c 1.00
Fatigue 0.14 (-0.18-0.46)c 0.14 (0.001-0.28)b 0.56 (0.33-0.79)a 1.00
Three-category depression (non-depressed, MiDD, MDD)
Depression Fatigue Depression Fatigue
Monozygotic twin pairs (N = 643 pairs)
Twin 1 Depression 1.00 Fatigue 0.46 (0.33-0.59)a 1.00
Twin 2 Depression 0.48 (0.33-0.62)b 0.36 (0.22-0.50)c 1.00
Fatigue 0.32 (0.18-0.47)c 0.43 (0.31-0.54)b 0.51 (0.39-0.64)a 1.00 Dizygotic twin pairs (N = 577 pairs)
Twin 1 Depression 1.00
Fatigue 0.50 (0.36-0.64)a 1.00 Twin 2 Depression 0.24 (0.04-0.43)b 0.05 (-0.13-0.23)c 1.00
Fatigue 0.06 (-0.12-0.24)c 0.14 (0.001-0.28)b 0.57 (0.45-0.70)a 1.00 aPhenotypic correlation between depression and fatigue. bTwin correlation. cCross-twin cross-trait correlation
Supplementary Table 6.5. Bivariate heritability model fits.
Model Minus two log-
likelihood χ2 Δ df p-value AIC
MiDD
ACE 4228.32 -5401.68
AE 4228.71 0.39 3 0.94 -5407.29
CE 4239.54 11.21 3 0.01 -5396.46
E 4295.28 66.96 6 1.71 × 10-12 -5346.72
ADE 4227.72 -0.61 0 1.00 -5402.28 MDD
ACE 3330.32 -5961.68
AE 3330.62 0.30 3 0.96 -5967.38
CE 3342.63 12.32 3 0.01 -5955.37
E 3381.46 51.14 6 2.78 × 10-9 -5922.50
ADE 3327.68 -2.64 0 1.00 -5964.33 Three-category depression
ACE 4647.74 -5080.26
AE 4649.28 1.54 3 0.67 -5084.72
CE 4661.68 13.94 3 3.00 × 10-3 -5072.32
E 4724.88 77.14 6 1.39 × 10-14 -5015.12
ADE 4647.42 -0.32 0 1.00 -5080.58
Note: Fit statistics are compared to ACE model and best-fitting models are indicated in bold. χ2: likelihood-ratio chi-squared
test; Δ df: difference in degrees of freedom.
Appendices 195
Supplementary Table 6.6. Co-twin control of minor depression and fatigue.
Sample
Risk factor: MiDD
Outcome: fatigue
Risk factor: fatigue
Outcome: MiDD
General Population 7.39 (4.33-12.61) [1,247] 7.39 (4.33-12.61) [1,247]
Discordant DZ 5.75 (2.89-11.48) [78] 5.47 (2.58-11.59) [201] Discordant MZ 1.77 (0.96-3.25) [85] 1.83 (0.97-3.46) [192]
MiDD: Minor depressive disorder. aWithin the General Population sample N is the number of individuals, while within the
Discordant MZ and DZ samples N is the number of discordant twin pairs.
Supplementary Table 6.7. Cross-tabulationa of depression and fatigue status within twin pairs
independent of overlapping symptoms.
Non-depressed MiDD MDD Total
MZ Total
Non-fatigued 264.5 4.5 0 269
Fatigued 34 2 0 36
Total 298.5 6.5 0 305
DZss Total
Non-fatigued 126.5 2.5 0 129
Fatigued 24 1 0 25
Total 150.5 3.5 0 154
DZos female-male
Non-fatigued 90 2 0 92
Fatigued 10 1 0 11
Total 100 3 0 103
DZos male-female
Non-fatigued 84 4 0 88
Fatigued 15 0 0 15
Total 99 4 0 103 aTables were made symmetrical in same-sex twin pairs by averaging over using either twin 1 or twin 2 as proband.
Supplementary Table 6.8. Relative riska of depression and fatigue estimated independently of
overlapping symptoms within monozygotic (MZ), same-sex dizygotic (DZss), and opposite-sex
dizygotic (DZos) twin pairs.
Proband - co-twin MZ (N = 319) DZss (N = 160) DZos (N = 119)
(Female-Male)
DZos (N = 119)
(Male-Female)
Depressed -non-fatigued 0.78 (0.47-1.31) 0.85 (0.44-1.65) 1.18 (1.08-1.28) 0.74 (0.33-1.65) Depressed - fatigued 2.70 (0.82-8.93) 1.79 (0.33-9.77) 0 3.33 (0.61-18.34)
Fatigued - non-depressed 0.96 (0.89-1.04) 0.98 (0.90-1.06) 0.93 (0.41-42.45) 1.05 (1.00-1.10)
Fatigued - depressed 3.32 (0.65-16.93) 2.06 (0.21-20.16) 4.18 (0.41-42.45) 0 aRelative risks were calculated with respect to non-depressed or non-fatigued status in twin 1.
Supplementary Table 6.9. Polychoric correlations with their 95% confidence intervals for fatigue
and depression independent of overlapping symptoms in twin pairs according to zygosity.
Twin 1 Twin 2
MiDD Fatigue MiDD Fatigue
Monozygotic twin pairs (N = 643 pairs)
Twin 1 MiDD 1.00 Fatigue 0.38 (0.01-0.74)a 1.00
Twin 2 MiDD -b 0.37 (-0.05-0.80)c 1.00
Fatigue 0.23 (-0.18-0.65)c 0.20 (-0.07-0.48)b -0.77 (-1.00-1.00)a 1.00 Dizygotic twin pairs (N = 577 pairs)
Twin 1 MiDD 1.00
Fatigue 0.50 (0.19-0.80)a 1.00 Twin 2 MiDD -b 0.16 (-0.38-0.71)c 1.00
Fatigue 0.08 (-0.33-0.49)c 0.10 (-0.18-0.38)b 0.40 (-0.04-0.85)a 1.00 aPhenotypic correlation between depression and fatigue. bTwin correlation. cCross-twin cross-trait correlation.
196 Appendices
Supplementary Figure 6.1. Path diagram of the bivariate Cholesky model variance estimates (with
their 95% confidence intervals) for minor depressive disorder (MiDD) and fatigue. The observed traits
are shown in the rectangles. Similarly, the latent variables (additive genetic factors: A, and unique
environmental factors: E) are depicted by circles. The arrows depict the relationship between the
variables. The genetic and environmental correlations between MiDD and fatigue were 0.76 (0.52-
1.00) and 0.29 (0.11-0.46), respectively.
Supplementary Figure 6.2. Path diagram of the bivariate Cholesky model variance estimates (with
their 95% confidence intervals) for major depressive disorder (MDD) and fatigue. The observed traits
are shown in the rectangles. Similarly, the latent variables (additive genetic factors: A, and unique
environmental factors: E) are depicted by circles. The arrows depict the relationship between the
variables. The genetic and environmental correlations between MDD and fatigue were 0.57 (0.21-
1.00) and 0.46 (0.09-0.52), respectively.
Appendices 197
Supplementary Figure 6.3. Path diagram of the bivariate Cholesky model variance estimates (with
their 95% confidence intervals) for three-category depression (non-depressed, MiDD, MDD) and
fatigue. The observed traits are shown in the rectangles. Similarly, the latent variables (additive
genetic factors: A, and unique environmental factors: E) are depicted by circles. The arrows depict the
relationship between the variables. The genetic and environmental correlations between MiDD and
fatigue were 0.71 (0.51-0.93) and 0.35 (0.18-0.51), respectively.
Supplementary Figure 6.4. Path diagram of the bivariate Cholesky model variance estimates (with
their 95% confidence intervals) for depression and fatigue independent of their overlapping
symptomology. The observed traits are shown in the rectangles. Similarly, the latent variables
(additive genetic factors: A, and unique environmental factors: E) are depicted by circles. The arrows
depict the relationship between the variables. The genetic and environmental correlations between
MiDD and fatigue were 1.00 (0.43-1.00) and 0.11 (0.00-0.43), respectively.
198 Appendices
Appendix C
Chapter 7: Supplementary Data
Supplementary Table 7.1. Calculation of Hardy-Weinberg equilibrium (HWE) p-value and allele frequencies for 442 SNPs reported by Schlauch and colleagues (2016) as
being associated with chronic fatigue syndrome. SNP ID’s highlighted in red do not meet the reported MAF (< 0.05) or HWE (p < 8 E-4) thresholds in cases, controls, or the
total cohort separately.
SNP Reported GWA p-value RA Freq RA
OA Freq OA HWE p-value
Case Control Total Case Control Total Case Control Total
rs12235235 5.76 × 10-16 T 0.476 0.079 0.288 C 0.524 0.921 0.713 1.30 × 10-7 8.40 × 10-5 0.0486
rs10144138 6.99 × 10-14 T 0.429 0.026 0.238 C 0.571 0.974 0.763 1.20 × 10-6 0.8677 0.0053
rs17120254 5.20 × 10-13 A 1.000 0.605 0.813 T 0.000 0.395 0.188 NC 0.0473 0.8906
rs41493945 6.25 × 10-13 A 0.393 0.013 0.213 G 0.607 0.987 0.788 2.70 × 10-5 0.9345 0.0158
rs3788079 3.42 × 10-12 C 0.345 0.000 0.181 A 0.655 1.000 0.819 0.0006 NC 0.0477
rs41378447 1.06 × 10-11 T 0.524 0.092 0.319 C 0.476 0.908 0.681 0.0006 0.1887 0.5613
rs3913434 1.26 × 10-11 T 0.381 0.013 0.206 C 0.619 0.987 0.794 0.0009 0.9345 0.1008
rs5967529 1.69 × 10-11 A 0.833 0.211 0.538 G 0.167 0.789 0.463 0.0016 0.7579 9.80 × 10-7
rs254577 2.35 × 10-11 C 0.798 0.263 0.544 T 0.202 0.737 0.456 0.1001 0.7579 0.0501
rs270838 3.61 × 10-11 C 0.476 0.105 0.300 A 0.524 0.895 0.700 1.30 × 10-7 0.4683 0.0010
rs1523773 4.73 × 10-11 T 0.321 0.000 0.169 A 0.679 1.000 0.831 0.0021 NC 0.0694
rs16827966 5.32 × 10-11 T 0.357 0.013 0.194 C 0.643 0.987 0.806 0.0003 0.9345 0.0316
rs2249954 5.47 × 10-11 G 0.464 0.079 0.281 A 0.536 0.921 0.719 1.20 × 10-5 0.5972 0.0167
rs8029503 5.66 × 10-11 T 0.488 0.105 0.306 C 0.512 0.895 0.694 1.50 × 10-5 0.3187 0.0653
rs3095598 1.02 × 10-10 C 0.429 0.039 0.244 T 0.571 0.961 0.756 0.0030 0.8000 0.2876
rs7010471 2.49 × 10-10 G 0.405 0.053 0.238 A 0.595 0.947 0.763 1.00 × 10-5 0.7320 0.0053
rs6757577 2.77 × 10-10 A 0.440 0.066 0.263 G 0.560 0.934 0.738 0.0001 0.6642 0.0425
rs11157573 2.97 × 10-10 G 0.488 0.158 0.331 A 0.512 0.842 0.669 1.50 × 10-5 0.0123 0.1609
rs16987633 3.46 × 10-10 A 0.774 0.355 0.575 G 0.226 0.645 0.425 1.90 × 10-5 0.0477 0.0111
rs12312259 3.60 × 10-10 C 0.548 0.158 0.363 T 0.452 0.842 0.638 4.00 × 10-5 0.2478 0.0290
rs948440 3.92 × 10-10 C 0.512 0.132 0.331 T 0.488 0.868 0.669 1.50 × 10-5 0.3503 0.0159
Footnotes for Supplementary Table 7.1 are on page 214.
Appendices 199
Supplementary Table 7.1. Continued Calculation of Hardy-Weinberg equilibrium (HWE) p-value and allele frequencies for 442 SNPs reported by Schlauch and colleagues
(2016) as being associated with chronic fatigue syndrome. SNP ID’s highlighted in red do not meet the reported MAF (< 0.05) or HWE (p < 8 E-4) thresholds in cases,
controls, or the total cohort separately.
SNP Reported GWA p-value RA Freq RA
OA Freq OA HWE p-value
Case Control Total Case Control Total Case Control Total
rs6445832 4.36 × 10-10 G 0.429 0.079 0.263 A 0.571 0.921 0.738 1.20 × 10-6 0.5972 0.0015
rs9585049 5.25 × 10-10 T 0.393 0.039 0.225 A 0.607 0.961 0.775 0.0004 0.8000 0.0505
rs7220341 5.41 × 10-10 G 0.655 0.276 0.475 A 0.345 0.724 0.525 0.0006 0.0186 0.0236
rs2816751 5.43 × 10-10 C 0.536 0.184 0.369 T 0.464 0.816 0.631 5.80 × 10-7 0.7547 0.0048
rs2200706 5.48 × 10-10 T 0.679 0.263 0.481 C 0.321 0.737 0.519 0.0021 0.1723 0.2575
rs17255510 6.61 × 10-10 C 0.679 0.171 0.438 T 0.321 0.829 0.563 0.3434 0.0308 0.0098
rs6892217 6.61 × 10-10 T 0.702 0.224 0.475 C 0.298 0.776 0.525 0.0447 0.3047 0.1860
rs17112444 8.02 × 10-10 A 0.369 0.026 0.206 G 0.631 0.974 0.794 0.0018 0.8677 0.1008
rs7849492 9.95 × 10-10 C 0.500 0.145 0.331 T 0.500 0.855 0.669 3.70 × 10-6 0.7893 0.0159
rs686190 1.11 × 10-9 G 0.393 0.053 0.231 A 0.607 0.947 0.769 2.70 × 10-5 0.7320 0.0071
rs16826918 1.13 × 10-9 G 0.464 0.079 0.281 A 0.536 0.921 0.719 0.0017 0.5972 0.1979
rs12317807 1.47 × 10-9 T 0.417 0.132 0.281 C 0.583 0.868 0.719 3.70 × 10-6 0.0567 0.0167
rs5974598 1.55 × 10-9 C 0.762 0.303 0.544 T 0.238 0.697 0.456 0.0428 0.6895 0.2902
rs1932556 1.63 × 10-9 T 0.988 0.605 0.806 G 0.012 0.395 0.194 0.9378 0.9573 0.0320
rs6797416 1.71 × 10-9 G 0.286 0.000 0.150 A 0.714 1.000 0.850 0.0095 NC 0.1145
rs2733416 1.71 × 10-9 G 0.286 0.000 0.150 A 0.714 1.000 0.850 0.0095 NC 0.1145
rs17035358 1.72 × 10-9 A 0.405 0.066 0.244 G 0.595 0.934 0.756 1.00 × 10-5 0.6642 0.0039
rs17368935 1.72 × 10-9 G 0.405 0.066 0.244 A 0.595 0.934 0.756 1.00 × 10-5 0.6642 0.0039
rs6679280 1.72 × 10-9 T 0.405 0.066 0.244 C 0.595 0.934 0.756 1.00 × 10-5 0.6642 0.0039
rs3867246 1.88 × 10-9 T 0.440 0.132 0.294 G 0.560 0.868 0.706 3.40 × 10-7 0.6272 0.0015
rs689462 2.08 × 10-9 C 0.464 0.092 0.288 A 0.536 0.908 0.713 0.0017 0.1887 0.3788
rs9285128 2.15 × 10-9 A 0.571 0.250 0.419 C 0.429 0.750 0.581 1.20 × 10-6 0.7456 0.0056
rs822027 2.52 × 10-9 A 0.321 0.013 0.175 G 0.679 0.987 0.825 0.0021 0.9345 0.0578
rs11168709 2.52 × 10-9 T 0.321 0.013 0.175 C 0.679 0.987 0.825 0.0021 0.9345 0.0578
rs12055682 2.99 × 10-9 G 0.631 0.263 0.456 A 0.369 0.737 0.544 0.0137 0.0003 0.1315
rs17047694 3.66 × 10-9 T 0.571 0.303 0.444 A 0.429 0.697 0.556 1.20 × 10-6 0.2427 0.0313
rs10978470 4.32 × 10-9 G 0.512 0.158 0.344 A 0.488 0.842 0.656 1.50 × 10-5 0.9488 0.0273
Footnotes for Supplementary Table 7.1 are on page 214.
200 Appendices
Supplementary Table 7.1. Continued Calculation of Hardy-Weinberg equilibrium (HWE) p-value and allele frequencies for 442 SNPs reported by Schlauch and colleagues
(2016) as being associated with chronic fatigue syndrome. SNP ID’s highlighted in red do not meet the reported MAF (< 0.05) or HWE (p < 8 E-4) thresholds in cases,
controls, or the total cohort separately.
SNP Reported GWA p-value RA Freq RA
OA Freq OA HWE p-value
Case Control Total Case Control Total Case Control Total
rs890527 4.60 × 10-9 T 0.560 0.289 0.431 A 0.440 0.711 0.569 3.40 × 10-7 0.5199 0.0074
rs11062852 4.84 × 10-9 C 0.500 0.197 0.356 A 0.500 0.803 0.644 6.00 × 10-5 0.0099 0.2939
rs16992281 5.22 × 10-9 A 0.357 0.000 0.188 C 0.643 1.000 0.813 0.2695 NC 0.0021
rs6675622 5.94 × 10-9 T 0.857 0.395 0.638 C 0.143 0.605 0.363 0.2801 0.9573 0.0916
rs6863118 6.22 × 10-9 G 0.464 0.105 0.294 A 0.536 0.895 0.706 0.0002 0.4683 0.0354
rs10121299 6.62 × 10-9 C 0.643 0.289 0.475 T 0.357 0.711 0.525 0.0003 0.5199 0.3581
rs12014391 6.66 × 10-9 A 0.714 0.211 0.475 G 0.286 0.789 0.525 0.2348 2.10 × 10-7 8.40 × 10-8
rs12391243 6.68 × 10-9 C 0.821 0.355 0.600 G 0.179 0.645 0.400 0.1589 0.0233 0.0039
rs1041296 6.89 × 10-9 G 0.560 0.211 0.394 A 0.440 0.789 0.606 0.0001 0.1991 0.2604
rs11027583 7.03 × 10-9 T 0.429 0.092 0.269 G 0.571 0.908 0.731 0.0003 0.1887 0.1140
rs12305678 7.87 × 10-9 G 0.560 0.224 0.400 A 0.440 0.776 0.600 7.60 × 10-6 0.9265 0.0253
rs9581771 7.96 × 10-9 T 0.440 0.092 0.275 C 0.560 0.908 0.725 0.0001 0.5317 0.0231
rs4022211 9.09 × 10-9 G 0.774 0.316 0.556 T 0.226 0.684 0.444 0.0120 0.1791 0.0175
rs16883408 1.06 × 10-8 C 0.631 0.237 0.444 G 0.369 0.763 0.556 0.0018 0.9060 0.4272
rs41456945 1.07 × 10-8 C 0.333 0.026 0.188 T 0.667 0.974 0.813 0.0012 0.8677 0.0390
rs361236 1.07 × 10-8 A 0.560 0.276 0.425 G 0.440 0.724 0.575 7.60 × 10-6 0.0887 0.1145
rs1007540 1.17 × 10-8 G 0.655 0.368 0.519 A 0.345 0.632 0.481 0.0006 0.0074 0.8326
rs7143222 1.54 × 10-8 T 0.262 0.000 0.138 C 0.738 1.000 0.863 0.0215 NC 0.1539
rs17092382 1.54 × 10-8 A 0.262 0.000 0.138 G 0.738 1.000 0.863 0.0215 NC 0.1539
rs7549528 1.54 × 10-8 C 0.262 0.000 0.138 T 0.738 1.000 0.863 0.0215 NC 0.1539
rs6854376 1.71 × 10-8 T 0.369 0.053 0.219 C 0.631 0.947 0.781 0.0002 0.7320 0.0123
rs16902672 1.77 × 10-8 C 0.583 0.211 0.406 G 0.417 0.789 0.594 0.0366 0.0012 0.1948
rs4473594 1.81 × 10-8 A 0.464 0.092 0.288 G 0.536 0.908 0.713 0.0119 0.5317 0.3788
rs10737169 2.51 × 10-8 A 0.607 0.224 0.425 G 0.393 0.776 0.575 0.0037 0.3047 0.8369
rs7883119 2.57 × 10-8 G 0.798 0.355 0.588 T 0.202 0.645 0.413 0.4912 1.10 × 10-5 1.50 × 10-5
rs4623336 2.68 × 10-8 T 0.476 0.158 0.325 C 0.524 0.842 0.675 3.30 × 10-6 0.9488 0.0055
rs2748997 2.76 × 10-8 C 0.440 0.079 0.269 T 0.560 0.921 0.731 0.0487 0.0887 0.8996
Footnotes for Supplementary Table 7.1 are on page 214.
Appendices 201
Supplementary Table 7.1. Continued Calculation of Hardy-Weinberg equilibrium (HWE) p-value and allele frequencies for 442 SNPs reported by Schlauch and colleagues
(2016) as being associated with chronic fatigue syndrome. SNP ID’s highlighted in red do not meet the reported MAF (< 0.05) or HWE (p < 8 E-4) thresholds in cases,
controls, or the total cohort separately.
SNP Reported GWA p-value RA Freq RA
OA Freq OA HWE p-value
Case Control Total Case Control Total Case Control Total
rs584569 2.84 × 10-8 A 0.381 0.066 0.231 G 0.619 0.934 0.769 6.70 × 10-5 0.6642 0.0071
rs13339179 2.89 × 10-8 T 0.345 0.039 0.200 C 0.655 0.961 0.800 0.0006 0.8000 0.0253
rs1222400 2.89 × 10-8 T 0.345 0.039 0.200 C 0.655 0.961 0.800 0.0006 0.8000 0.0253
rs2882361 3.02 × 10-8 G 0.917 0.513 0.725 C 0.083 0.487 0.275 0.5558 0.5134 0.2742
rs41464146 3.22 × 10-8 C 0.333 0.026 0.188 T 0.667 0.974 0.813 0.0109 0.8677 0.1835
rs9446695 3.46 × 10-8 T 0.321 0.026 0.181 A 0.679 0.974 0.819 0.0021 0.8677 0.0477
rs7290437 3.52 × 10-8 G 0.476 0.250 0.369 A 0.524 0.750 0.631 3.30 × 10-6 0.0231 0.0625
rs12607783 4.31 × 10-8 A 0.417 0.132 0.281 T 0.583 0.868 0.719 3.70 × 10-6 0.6272 0.0032
rs606324 4.39 × 10-8 A 0.357 0.079 0.225 G 0.643 0.921 0.775 0.0003 0.0887 0.0505
rs2869820 4.39 × 10-8 T 0.250 0.000 0.131 A 0.750 1.000 0.869 0.0308 NC 0.1766
rs6643261 4.55 × 10-8 A 0.452 0.026 0.250 G 0.548 0.974 0.750 0.0008 7.10 × 10-10 2.50 × 10-9
rs17133553 4.74 × 10-8 A 0.679 0.276 0.488 T 0.321 0.724 0.513 0.0182 0.3728 0.6586
rs2816936 4.91 × 10-8 A 0.940 0.513 0.738 G 0.060 0.487 0.263 0.6817 0.1955 0.0015
rs1915603 5.15 × 10-8 G 0.286 0.039 0.169 A 0.714 0.961 0.831 0.0095 5.70 × 10-5 0.3083
rs17052315 5.97 × 10-8 A 0.357 0.053 0.213 G 0.643 0.947 0.788 0.0003 0.7320 0.0158
rs6502875 5.97 × 10-8 G 0.643 0.316 0.488 A 0.357 0.684 0.513 0.0003 0.8744 0.1776
rs10047684 7.31 × 10-8 A 0.583 0.237 0.419 T 0.417 0.763 0.581 0.0008 0.0557 0.0209
rs11038285 7.69 × 10-8 G 0.286 0.013 0.156 T 0.714 0.987 0.844 0.0095 0.9345 0.0977
rs7701654 7.69 × 10-8 G 0.286 0.013 0.156 A 0.714 0.987 0.844 0.0095 0.9345 0.0977
rs6662412 7.69 × 10-8 G 0.286 0.013 0.156 A 0.714 0.987 0.844 0.0095 0.9345 0.0977
rs7301442 7.69 × 10-8 T 0.286 0.013 0.156 A 0.714 0.987 0.844 0.0095 0.9345 0.0977
rs2193766 7.80 × 10-8 G 1.000 0.763 0.888 A 0.000 0.237 0.113 NC 0.0557 0.2569
rs17060061 7.80 × 10-8 G 1.000 0.763 0.888 A 0.000 0.237 0.113 NC 0.0557 0.2569
rs9301483 8.08 × 10-8 A 0.643 0.250 0.456 G 0.357 0.750 0.544 0.0241 0.1597 0.5439
rs7537461 8.53 × 10-8 C 0.607 0.289 0.456 A 0.393 0.711 0.544 0.0004 0.1521 0.4563
rs1362859 8.99 × 10-8 G 0.583 0.329 0.463 C 0.417 0.671 0.538 6.50 × 10-5 0.0338 0.3421
rs9964872 9.92 × 10-8 A 0.321 0.026 0.181 G 0.679 0.974 0.819 0.0182 0.8677 0.2200
Footnotes for Supplementary Table 7.1 are on page 214.
202 Appendices
Supplementary Table 7.1. Continued Calculation of Hardy-Weinberg equilibrium (HWE) p-value and allele frequencies for 442 SNPs reported by Schlauch and colleagues
(2016) as being associated with chronic fatigue syndrome. SNP ID’s highlighted in red do not meet the reported MAF (< 0.05) or HWE (p < 8 E-4) thresholds in cases,
controls, or the total cohort separately.
SNP Reported GWA p-value RA Freq RA
OA Freq OA HWE p-value
Case Control Total Case Control Total Case Control Total
rs17643851 9.93 × 10-8 G 0.429 0.079 0.263 A 0.571 0.921 0.738 0.0872 0.0887 0.7783
rs10788258 1.04 × 10-7 T 0.429 0.092 0.269 C 0.571 0.908 0.731 0.0030 0.5317 0.1140
rs8057267 1.06 × 10-7 G 0.476 0.171 0.331 A 0.524 0.829 0.669 5.40 × 10-5 0.3096 0.0565
rs10074876 1.07 × 10-7 C 0.310 0.026 0.175 T 0.690 0.974 0.825 0.0037 0.8677 0.0578
rs1961484 1.21 × 10-7 A 0.429 0.066 0.256 G 0.571 0.934 0.744 0.2801 0.6642 0.6613
rs16973831 1.22 × 10-7 T 0.238 0.000 0.125 A 0.762 1.000 0.875 0.0428 NC 0.2013
rs4510466 1.22 × 10-7 C 0.238 0.000 0.125 A 0.762 1.000 0.875 0.0428 NC 0.2013
rs13393078 1.24 × 10-7 C 0.393 0.079 0.244 T 0.607 0.921 0.756 0.0004 0.5972 0.0228
rs2204978 1.26 × 10-7 A 0.488 0.211 0.356 G 0.512 0.789 0.644 7.60 × 10-7 0.7579 0.0027
rs41330648 1.28 × 10-7 G 0.524 0.118 0.331 A 0.476 0.882 0.669 0.7459 0.4076 0.2621
rs6479969 1.28 × 10-7 G 0.774 0.382 0.588 C 0.226 0.618 0.413 0.0582 0.7141 0.8581
rs16877795 1.49 × 10-7 G 0.500 0.171 0.344 A 0.500 0.829 0.656 0.0007 0.3096 0.2241
rs197770 1.50 × 10-7 G 0.631 0.382 0.513 A 0.369 0.618 0.488 0.0002 0.0899 0.3678
rs17426290 1.54 × 10-7 T 0.464 0.250 0.363 C 0.536 0.750 0.638 5.80 × 10-7 0.1597 0.0077
rs2017563 1.55 × 10-7 A 0.357 0.013 0.194 G 0.643 0.987 0.806 0.2695 0.9345 0.0042
rs690607 1.56 × 10-7 A 0.476 0.158 0.325 G 0.524 0.842 0.675 0.0006 0.1991 0.2118
rs1859790 1.56 × 10-7 T 0.405 0.092 0.256 C 0.595 0.908 0.744 0.0002 0.5317 0.0126
rs7119924 1.58 × 10-7 C 0.452 0.184 0.325 T 0.548 0.816 0.675 2.20 × 10-6 0.4431 0.0055
rs275154 1.72 × 10-7 G 0.417 0.145 0.288 A 0.583 0.855 0.713 6.50 × 10-5 0.1146 0.0486
rs6508891 1.75 × 10-7 T 0.429 0.118 0.281 A 0.571 0.882 0.719 2.30 × 10-5 0.4076 0.0032
rs11914436 1.93 × 10-7 A 0.607 0.368 0.494 G 0.393 0.632 0.506 0.0004 0.0074 0.8219
rs41385645 1.99 × 10-7 T 0.488 0.132 0.319 A 0.512 0.868 0.681 0.0134 0.6272 0.5613
rs7552454 2.11 × 10-7 A 0.488 0.197 0.350 G 0.512 0.803 0.650 7.60 × 10-7 0.6229 0.0008
rs11021876 2.21 × 10-7 T 0.274 0.013 0.150 C 0.726 0.987 0.850 0.0145 0.9345 0.1145
rs7739542 2.38 × 10-7 C 1.000 0.776 0.894 C 0.000 0.224 0.106 NC 0.0757 0.2876
rs16994314 2.42 × 10-7 T 0.464 0.118 0.300 A 0.536 0.882 0.700 0.0119 0.4076 0.2415
rs12086522 2.61 × 10-7 T 0.381 0.092 0.244 G 0.619 0.908 0.756 0.0009 0.1887 0.0950
Footnotes for Supplementary Table 7.1 are on page 214.
Appendices 203
Supplementary Table 7.1. Continued Calculation of Hardy-Weinberg equilibrium (HWE) p-value and allele frequencies for 442 SNPs reported by Schlauch and colleagues
(2016) as being associated with chronic fatigue syndrome. SNP ID’s highlighted in red do not meet the reported MAF (< 0.05) or HWE (p < 8 E-4) thresholds in cases,
controls, or the total cohort separately.
SNP Reported GWA p-value RA Freq RA
OA Freq OA HWE p-value
Case Control Total Case Control Total Case Control Total
rs12408925 2.61 × 10-7 G 0.667 0.276 0.481 A 0.333 0.724 0.519 0.6434 3.50 × 10-5 0.0008
rs2980018 2.65 × 10-7 T 0.738 0.250 0.506 C 0.262 0.750 0.494 0.3717 0.0231 0.0001
rs6504560 2.69 × 10-7 T 0.321 0.026 0.181 C 0.679 0.974 0.819 0.0979 0.8677 0.6360
rs5942699 2.70 × 10-7 G 0.464 0.053 0.269 A 0.536 0.947 0.731 0.0676 0.0036 4.00 × 10-5
rs12559754 2.86 × 10-7 G 0.857 0.447 0.663 A 0.143 0.553 0.338 0.2801 0.0039 0.0006
rs1610024 2.89 × 10-7 A 0.607 0.250 0.438 G 0.393 0.750 0.563 0.1083 0.0017 0.0939
rs213981 2.94 × 10-7 G 0.679 0.329 0.513 T 0.321 0.671 0.488 0.0182 0.0338 0.3738
rs17024760 3.22 × 10-7 T 0.655 0.211 0.444 C 0.345 0.789 0.556 0.1735 0.5043 0.0175
rs4242794 3.27 × 10-7 A 0.226 0.000 0.119 G 0.774 1.000 0.881 0.0582 NC 0.2281
rs4792493 3.27 × 10-7 G 0.226 0.000 0.119 A 0.774 1.000 0.881 0.0582 NC 0.2281
rs4982735 3.27 × 10-7 C 0.226 0.000 0.119 A 0.774 1.000 0.881 0.0582 NC 0.2281
rs6973776 3.27 × 10-7 T 0.226 0.000 0.119 C 0.774 1.000 0.881 0.0582 NC 0.2281
rs4422316 3.27 × 10-7 T 0.226 0.000 0.119 C 0.774 1.000 0.881 0.0582 NC 0.2281
rs233122 3.27 × 10-7 C 0.250 0.000 0.131 T 0.750 1.000 0.869 0.6070 NC 0.5420
rs2421122 3.27 × 10-7 C 0.238 0.000 0.125 T 0.762 1.000 0.875 0.2401 NC 0.7983
rs7847862 3.47 × 10-7 G 0.631 0.329 0.488 A 0.369 0.671 0.513 0.0002 0.4139 0.0249
rs283825 4.41 × 10-7 G 0.595 0.342 0.475 A 0.405 0.658 0.525 0.0002 0.0658 0.3581
rs822020 5.15 × 10-7 C 0.357 0.026 0.200 T 0.643 0.974 0.800 0.6657 0.8677 0.0504
rs12629385 5.27 × 10-7 T 0.381 0.066 0.231 A 0.619 0.934 0.769 0.1704 0.0274 0.6498
rs11056347 5.27 × 10-7 A 0.345 0.053 0.206 G 0.655 0.947 0.794 0.0062 0.7320 0.1008
rs9946817 5.35 × 10-7 C 0.357 0.158 0.263 T 0.643 0.842 0.738 0.0034 0.0002 0.7672
rs579751 5.59 × 10-7 C 0.810 0.382 0.606 A 0.190 0.618 0.394 0.6002 0.0021 0.0004
rs17041554 5.63 × 10-7 A 0.619 0.237 0.438 G 0.381 0.763 0.563 0.0428 0.9060 0.8871
rs7726463 5.64 × 10-7 G 0.679 0.355 0.525 A 0.321 0.645 0.475 0.0021 0.5729 0.1715
rs12965947 5.82 × 10-7 A 0.405 0.053 0.238 G 0.595 0.947 0.763 0.9392 0.7320 0.1246
rs609539 6.14 × 10-7 C 0.262 0.013 0.144 T 0.738 0.987 0.856 0.0215 0.9345 0.1332
rs6957524 6.14 × 10-7 G 0.262 0.013 0.144 A 0.738 0.987 0.856 0.0215 0.9345 0.1332
Footnotes for Supplementary Table 7.1 are on page 214.
204 Appendices
Supplementary Table 7.1. Continued Calculation of Hardy-Weinberg equilibrium (HWE) p-value and allele frequencies for 442 SNPs reported by Schlauch and colleagues
(2016) as being associated with chronic fatigue syndrome. SNP ID’s highlighted in red do not meet the reported MAF (< 0.05) or HWE (p < 8 E-4) thresholds in cases,
controls, or the total cohort separately.
SNP Reported GWA p-value RA Freq RA
OA Freq OA HWE p-value
Case Control Total Case Control Total Case Control Total
rs6093591 6.67 × 10-7 T 0.476 0.184 0.338 C 0.524 0.816 0.663 3.30 × 10-6 0.1639 0.0004
rs41441747 6.72 × 10-7 C 0.512 0.197 0.363 G 0.488 0.803 0.638 0.0002 0.1296 0.0077
rs16956158 6.93 × 10-7 G 0.762 0.447 0.613 A 0.238 0.553 0.388 0.0428 0.0039 0.0603
rs7613828 7.03 × 10-7 A 1.000 0.737 0.875 G 0.000 0.263 0.125 NC 0.2523 0.0049
rs7768988 7.15 × 10-7 T 0.655 0.237 0.456 C 0.345 0.763 0.544 0.4923 0.0936 0.0501
rs17130776 7.17 × 10-7 T 0.536 0.197 0.375 C 0.464 0.803 0.625 0.0017 0.6229 0.1211
rs2223341 7.27 × 10-7 G 0.464 0.053 0.269 A 0.536 0.947 0.731 0.0144 0.0036 2.90 × 10-6
rs17865437 7.39 × 10-7 A 0.869 0.671 0.775 C 0.131 0.329 0.225 0.0020 0.0025 0.5008
rs17861907 7.39 × 10-7 G 0.869 0.671 0.775 A 0.131 0.329 0.225 0.0020 0.0025 0.5008
rs7895391 7.78 × 10-7 T 0.476 0.145 0.319 C 0.524 0.855 0.681 0.0051 0.7893 0.2732
rs9283919 7.87 × 10-7 G 0.381 0.053 0.225 A 0.619 0.947 0.775 0.9503 0.0036 0.0586
rs9977796 8.56 × 10-7 G 0.310 0.000 0.163 A 0.690 1.000 0.838 0.0041 NC 1.30 × 10-6
rs7672066 8.56 × 10-7 G 0.214 0.000 0.113 A 0.786 1.000 0.888 0.0771 NC 0.2569
rs11873202 8.56 × 10-7 A 0.214 0.000 0.113 G 0.786 1.000 0.888 0.0771 NC 0.2569
rs866781 8.56 × 10-7 A 0.214 0.000 0.113 G 0.786 1.000 0.888 0.0771 NC 0.2569
rs3732196 8.56 × 10-7 T 0.214 0.000 0.113 C 0.786 1.000 0.888 0.0771 NC 0.2569
rs16987453 8.56 × 10-7 G 0.226 0.000 0.119 T 0.774 1.000 0.881 0.3112 NC 0.8911
rs12407818 8.75 × 10-7 C 0.536 0.263 0.406 T 0.464 0.737 0.594 1.20 × 10-5 0.7579 0.0159
rs4289946 8.81 × 10-7 C 0.571 0.237 0.413 G 0.429 0.763 0.588 4.40 × 10-6 0.3098 0.0001
rs5909213 8.81 × 10-7 C 0.571 0.237 0.413 T 0.429 0.763 0.588 4.40 × 10-6 0.3098 0.0001
rs5909214 8.81 × 10-7 G 0.571 0.237 0.413 A 0.429 0.763 0.588 4.40 × 10-6 0.3098 0.0001
rs5909220 8.81 × 10-7 T 0.571 0.237 0.413 C 0.429 0.763 0.588 4.40 × 10-6 0.3098 0.0001
rs9419277 8.93 × 10-7 G 0.286 0.026 0.163 A 0.714 0.974 0.838 0.0095 0.8677 0.0827
rs17085969 8.99 × 10-7 T 0.988 0.750 0.875 C 0.012 0.250 0.125 0.9378 0.0399 0.2013
rs16842140 9.07 × 10-7 G 0.512 0.263 0.394 A 0.488 0.737 0.606 0.0020 0.0048 0.8502
rs13398697 1.05 × 10-6 A 0.345 0.066 0.213 T 0.655 0.934 0.788 0.0006 0.6642 0.0158
rs10146102 1.09 × 10-6 T 0.381 0.079 0.238 C 0.619 0.921 0.763 0.0074 0.5972 0.1209
Footnotes for Supplementary Table 7.1 are on page 214.
Appendices 205
Supplementary Table 7.1. Continued Calculation of Hardy-Weinberg equilibrium (HWE) p-value and allele frequencies for 442 SNPs reported by Schlauch and colleagues
(2016) as being associated with chronic fatigue syndrome. SNP ID’s highlighted in red do not meet the reported MAF (< 0.05) or HWE (p < 8 E-4) thresholds in cases,
controls, or the total cohort separately.
SNP Reported GWA p-value RA Freq RA
OA Freq OA HWE p-value
Case Control Total Case Control Total Case Control Total
rs1873717 1.23 × 10-6 T 0.524 0.184 0.363 C 0.476 0.816 0.638 0.0293 0.1639 0.2242
rs41363145 1.32 × 10-6 C 0.369 0.079 0.231 G 0.631 0.921 0.769 0.0018 0.5972 0.0392
rs6871885 1.42 × 10-6 A 0.393 0.092 0.250 T 0.607 0.908 0.750 0.0037 0.5317 0.0736
rs7321094 1.45 × 10-6 T 0.571 0.237 0.413 C 0.429 0.763 0.588 0.0030 0.9060 0.2281
rs17781246 1.48 × 10-6 G 0.512 0.289 0.406 A 0.488 0.711 0.594 1.50 × 10-5 0.1521 0.0514
rs7529216 1.52 × 10-6 G 0.274 0.039 0.163 T 0.726 0.961 0.838 0.0954 5.70 × 10-5 0.9264
rs12331711 1.52 × 10-6 G 0.381 0.066 0.231 A 0.619 0.934 0.769 0.1704 0.6642 0.8611
rs17079111 1.53 × 10-6 G 0.333 0.053 0.200 A 0.667 0.947 0.800 0.0109 0.7320 0.1242
rs7610618 1.57 × 10-6 T 0.262 0.013 0.144 C 0.738 0.987 0.856 0.1333 0.9345 0.5530
rs16985794 1.58 × 10-6 C 0.298 0.066 0.188 G 0.702 0.934 0.813 0.0060 0.0274 0.1835
rs11010290 1.61 × 10-6 T 0.798 0.395 0.606 C 0.202 0.605 0.394 0.4912 0.0056 0.0020
rs9913705 1.65 × 10-6 G 0.250 0.013 0.138 A 0.750 0.987 0.863 0.0308 0.9345 0.1539
rs14541 1.70 × 10-6 G 0.548 0.224 0.394 A 0.452 0.776 0.606 0.0251 0.0499 0.7798
rs16861920 1.76 × 10-6 C 0.321 0.053 0.194 T 0.679 0.947 0.806 0.0021 0.7320 0.0316
rs17127809 1.76 × 10-6 T 0.321 0.053 0.194 C 0.679 0.947 0.806 0.0021 0.7320 0.0316
rs2207301 1.83 × 10-6 G 0.571 0.132 0.363 A 0.429 0.868 0.638 0.0384 0.0009 4.40 × 10-6
rs4843884 1.83 × 10-6 G 0.369 0.066 0.225 A 0.631 0.934 0.775 0.0714 0.6642 0.5008
rs12417706 1.90 × 10-6 T 0.452 0.105 0.288 C 0.548 0.895 0.713 0.3204 0.3187 0.4489
rs3095168 1.91 × 10-6 A 0.500 0.237 0.375 G 0.500 0.763 0.625 3.70 × 10-6 0.9060 0.0029
rs17780243 1.95 × 10-6 T 0.595 0.368 0.488 A 0.405 0.632 0.513 1.00 × 10-5 0.9123 0.0071
rs6927507 1.99 × 10-6 G 0.429 0.184 0.313 A 0.571 0.816 0.688 1.20 × 10-6 0.7547 0.0004
rs1350060 2.01 × 10-6 A 0.452 0.211 0.338 G 0.548 0.789 0.663 2.20 × 10-6 0.7579 0.0022
rs6940702 2.18 × 10-6 A 0.202 0.000 0.106 G 0.798 1.000 0.894 0.1001 NC 0.2876
rs10114442 2.18 × 10-6 T 0.202 0.000 0.106 C 0.798 1.000 0.894 0.1001 NC 0.2876
rs9628158 2.18 × 10-6 T 0.202 0.000 0.106 C 0.798 1.000 0.894 0.1001 NC 0.2876
rs7154569 2.18 × 10-6 C 0.202 0.000 0.106 T 0.798 1.000 0.894 0.1001 NC 0.2876
rs2803453 2.18 × 10-6 A 0.226 0.000 0.119 G 0.774 1.000 0.881 0.8956 NC 0.3516
Footnotes for Supplementary Table 7.1 are on page 214.
206 Appendices
Supplementary Table 7.1. Continued Calculation of Hardy-Weinberg equilibrium (HWE) p-value and allele frequencies for 442 SNPs reported by Schlauch and colleagues
(2016) as being associated with chronic fatigue syndrome. SNP ID’s highlighted in red do not meet the reported MAF (< 0.05) or HWE (p < 8 E-4) thresholds in cases,
controls, or the total cohort separately.
SNP Reported GWA p-value RA Freq RA
OA Freq OA HWE p-value
Case Control Total Case Control Total Case Control Total
rs2516025 2.19 × 10-6 C 0.357 0.013 0.194 T 0.643 0.987 0.806 0.0018 0.9345 5.50 × 10-7
rs4812100 2.21 × 10-6 G 0.631 0.289 0.469 A 0.369 0.711 0.531 0.0714 0.0264 0.2769
rs7347140 2.23 × 10-6 T 0.286 0.026 0.163 C 0.714 0.974 0.838 0.0663 0.8677 0.3607
rs10928930 2.23 × 10-6 T 0.286 0.026 0.163 C 0.714 0.974 0.838 0.0663 0.8677 0.3607
rs9920285 2.36 × 10-6 A 0.298 0.039 0.175 G 0.702 0.961 0.825 0.0060 0.8000 0.0578
rs10133617 2.36 × 10-6 T 0.298 0.039 0.175 A 0.702 0.961 0.825 0.0060 0.8000 0.0578
rs16844808 2.42 × 10-6 T 0.298 0.079 0.194 C 0.702 0.921 0.806 0.0447 8.40 × 10-5 0.9982
rs17098846 2.42 × 10-6 A 0.274 0.026 0.156 G 0.726 0.974 0.844 0.0145 0.8677 0.0977
rs682564 2.42 × 10-6 A 0.274 0.026 0.156 T 0.726 0.974 0.844 0.0145 0.8677 0.0977
rs3017495 2.42 × 10-6 T 0.274 0.026 0.156 C 0.726 0.974 0.844 0.0145 0.8677 0.0977
rs5909082 2.49 × 10-6 T 0.548 0.224 0.394 C 0.452 0.776 0.606 4.00 × 10-6 0.3998 5.70 × 10-5
rs17475512 2.56 × 10-6 G 0.690 0.382 0.544 A 0.310 0.618 0.456 0.0290 0.0172 0.2902
rs271662 2.60 × 10-6 C 0.464 0.132 0.306 T 0.536 0.868 0.694 0.2027 0.0567 0.4309
rs12300888 2.65 × 10-6 C 0.369 0.079 0.231 A 0.631 0.921 0.769 0.0714 0.0887 0.8611
rs11679695 2.80 × 10-6 G 0.440 0.224 0.338 A 0.560 0.776 0.663 7.60 × 10-6 0.3047 0.0106
rs243391 2.96 × 10-6 G 0.560 0.276 0.425 A 0.440 0.724 0.575 0.0001 0.9362 0.0418
rs889083 3.14 × 10-6 G 0.452 0.224 0.344 A 0.548 0.776 0.656 0.0042 0.0038 0.8223
rs41368852 3.28 × 10-6 G 0.345 0.039 0.200 A 0.655 0.961 0.800 0.9968 0.8000 0.2085
rs12551218 3.31 × 10-6 T 0.310 0.079 0.200 A 0.690 0.921 0.800 0.0037 0.0887 0.1242
rs16890805 3.34 × 10-6 T 0.345 0.105 0.231 C 0.655 0.895 0.769 0.0006 0.3187 0.0392
rs6757543 3.68 × 10-6 G 0.512 0.237 0.381 T 0.488 0.763 0.619 0.0020 0.0936 0.4403
rs6735919 3.82 × 10-6 T 0.476 0.250 0.369 C 0.524 0.750 0.631 3.30 × 10-6 0.5887 0.0048
rs4714199 3.92 × 10-6 C 0.512 0.158 0.344 A 0.488 0.842 0.656 0.9971 0.2478 0.4433
rs589402 3.92 × 10-6 T 0.952 0.697 0.831 A 0.048 0.303 0.169 0.7459 0.0075 0.0694
rs9668748 4.02 × 10-6 C 0.429 0.079 0.263 T 0.571 0.921 0.738 0.4179 0.5972 0.0440
rs7121660 4.10 × 10-6 A 0.786 0.447 0.625 G 0.214 0.553 0.375 0.0771 0.7956 0.7205
rs16987589 4.27 × 10-6 A 0.238 0.013 0.131 G 0.762 0.987 0.869 0.0428 0.9345 0.1766
Footnotes for Supplementary Table 7.1 are on page 214.
Appendices 207
Supplementary Table 7.1. Continued Calculation of Hardy-Weinberg equilibrium (HWE) p-value and allele frequencies for 442 SNPs reported by Schlauch and colleagues
(2016) as being associated with chronic fatigue syndrome. SNP ID’s highlighted in red do not meet the reported MAF (< 0.05) or HWE (p < 8 E-4) thresholds in cases,
controls, or the total cohort separately.
SNP Reported GWA p-value RA Freq RA
OA Freq OA HWE p-value
Case Control Total Case Control Total Case Control Total
rs6656441 4.27 × 10-6 C 0.238 0.013 0.131 T 0.762 0.987 0.869 0.0428 0.9345 0.1766
rs9698174 4.45 × 10-6 A 0.452 0.237 0.350 G 0.548 0.763 0.650 4.70 × 10-9 0.9060 5.40 × 10-7
rs7830366 4.48 × 10-6 T 0.619 0.316 0.475 A 0.381 0.684 0.525 0.0009 0.1791 0.0236
rs10988052 4.71 × 10-6 G 0.345 0.079 0.219 A 0.655 0.921 0.781 0.0006 0.5972 0.0123
rs7272593 4.89 × 10-6 G 0.310 0.053 0.188 A 0.690 0.947 0.813 0.0037 0.7320 0.0390
rs7742257 4.89 × 10-6 T 0.310 0.053 0.188 C 0.690 0.947 0.813 0.0037 0.7320 0.0390
rs4144897 5.12 × 10-6 T 0.405 0.118 0.269 C 0.595 0.882 0.731 0.5726 0.0001 0.0668
rs6450296 5.16 × 10-6 A 0.381 0.118 0.256 G 0.619 0.882 0.744 0.0009 0.4679 0.0563
rs12001751 5.44 × 10-6 T 0.190 0.000 0.100 C 0.810 1.000 0.900 0.1273 NC 0.3203
rs6497951 5.44 × 10-6 T 0.190 0.000 0.100 C 0.810 1.000 0.900 0.1273 NC 0.3203
rs11163916 5.44 × 10-6 G 0.190 0.000 0.100 A 0.810 1.000 0.900 0.1273 NC 0.3203
rs7404102 5.44 × 10-6 A 0.190 0.000 0.100 G 0.810 1.000 0.900 0.1273 NC 0.3203
rs1486178 5.44 × 10-6 T 0.190 0.000 0.100 C 0.810 1.000 0.900 0.1273 NC 0.3203
rs340170 5.44 × 10-6 G 0.190 0.000 0.100 A 0.810 1.000 0.900 0.1273 NC 0.3203
rs7960674 5.44 × 10-6 C 0.190 0.000 0.100 A 0.810 1.000 0.900 0.1273 NC 0.3203
rs496731 5.52 × 10-6 T 0.762 0.474 0.625 G 0.238 0.526 0.375 0.0428 0.0238 0.1896
rs2981884 5.62 × 10-6 T 1.000 0.816 0.913 T 0.000 0.184 0.088 NC 0.1639 0.3911
rs9311374 5.62 × 10-6 C 1.000 0.803 0.906 T 0.000 0.197 0.094 NC 0.6229 0.6960
rs7019328 5.66 × 10-6 T 0.964 0.645 0.813 C 0.036 0.355 0.188 0.8103 0.1185 0.0021
rs41423649 5.73 × 10-6 C 0.964 0.711 0.844 T 0.036 0.289 0.156 0.8103 0.0849 0.4189
rs16970196 5.76 × 10-6 A 0.952 0.750 0.856 G 0.048 0.250 0.144 0.0021 0.0399 0.5530
rs10009657 5.76 × 10-6 G 0.952 0.750 0.856 A 0.048 0.250 0.144 0.0021 0.0399 0.5530
rs1157185 5.84 × 10-6 T 0.571 0.237 0.413 C 0.429 0.763 0.588 0.0872 0.0936 0.5221
rs1367696 5.84 × 10-6 T 0.571 0.237 0.413 C 0.429 0.763 0.588 0.0872 0.0936 0.5221
rs7011650 5.89 × 10-6 T 0.524 0.263 0.400 C 0.476 0.737 0.600 5.40 × 10-5 0.7579 0.0253
rs16980810 6.05 × 10-6 A 0.464 0.158 0.319 G 0.536 0.842 0.681 0.0119 0.2478 0.1073
rs6008155 6.32 × 10-6 G 0.286 0.039 0.169 C 0.714 0.961 0.831 0.0095 0.8000 0.0694
Footnotes for Supplementary Table 7.1 are on page 214.
208 Appendices
Supplementary Table 7.1. Continued Calculation of Hardy-Weinberg equilibrium (HWE) p-value and allele frequencies for 442 SNPs reported by Schlauch and colleagues
(2016) as being associated with chronic fatigue syndrome. SNP ID’s highlighted in red do not meet the reported MAF (< 0.05) or HWE (p < 8 E-4) thresholds in cases,
controls, or the total cohort separately.
SNP Reported GWA p-value RA Freq RA
OA Freq OA HWE p-value
Case Control Total Case Control Total Case Control Total
rs479448 6.34 × 10-6 C 0.262 0.026 0.150 T 0.738 0.974 0.850 0.0215 0.8677 0.1145
rs11917596 6.34 × 10-6 C 0.262 0.026 0.150 T 0.738 0.974 0.850 0.0215 0.8677 0.1145
rs17085519 6.34 × 10-6 G 0.262 0.026 0.150 A 0.738 0.974 0.850 0.0215 0.8677 0.1145
rs2237406 6.34 × 10-6 T 0.262 0.026 0.150 C 0.738 0.974 0.850 0.0215 0.8677 0.1145
rs9362453 6.40 × 10-6 G 0.274 0.039 0.163 A 0.726 0.961 0.838 0.3727 5.70 × 10-5 0.4659
rs6654507 6.44 × 10-6 G 0.548 0.118 0.344 A 0.452 0.882 0.656 0.0061 0.0226 2.20 × 10-6
rs6950641 6.49 × 10-6 T 0.357 0.092 0.231 C 0.643 0.908 0.769 0.0003 0.5317 0.0071
rs6449669 6.54 × 10-6 T 0.595 0.289 0.450 A 0.405 0.711 0.550 0.0129 0.1521 0.9280
rs10137248 6.59 × 10-6 G 0.226 0.039 0.138 A 0.774 0.961 0.863 0.0582 5.70 × 10-5 0.6290
rs8130198 6.63 × 10-6 C 0.583 0.382 0.488 T 0.417 0.618 0.513 0.0064 0.0021 0.6586
rs9844641 6.65 × 10-6 A 0.988 0.750 0.875 G 0.012 0.250 0.125 0.9378 0.7456 0.4433
rs2079989 6.68 × 10-6 C 0.524 0.421 0.475 T 0.476 0.579 0.525 3.70 × 10-6 0.0686 0.0265
rs8050875 6.91 × 10-6 G 0.976 0.724 0.856 A 0.024 0.276 0.144 0.8744 0.4647 0.7527
rs17019070 6.92 × 10-6 C 0.452 0.184 0.325 T 0.548 0.816 0.675 0.0005 0.4431 0.0787
rs10129777 6.97 × 10-6 G 0.405 0.079 0.250 A 0.595 0.921 0.750 0.9392 0.5972 0.2330
rs5930683 7.06 × 10-6 A 0.226 0.224 0.225 G 0.774 0.776 0.775 1.50 × 10-9 0.3998 0.0001
rs11205084 7.25 × 10-6 G 0.548 0.303 0.431 A 0.452 0.697 0.569 4.00 × 10-5 0.6895 0.0262
rs4099911 7.55 × 10-6 A 0.500 0.408 0.456 C 0.500 0.592 0.544 0.0007 0.0017 0.7685
rs349391 7.67 × 10-6 C 0.429 0.303 0.369 T 0.571 0.697 0.631 0.0009 0.0075 0.3081
rs349390 7.67 × 10-6 C 0.429 0.303 0.369 T 0.571 0.697 0.631 0.0009 0.0075 0.3081
rs17722227 7.70 × 10-6 A 0.179 0.026 0.106 G 0.821 0.974 0.894 0.1589 7.10 × 10-10 0.9092
rs6470455 7.73 × 10-6 G 0.714 0.342 0.538 A 0.286 0.658 0.463 0.2801 0.2631 0.1941
rs7019283 8.05 × 10-6 C 0.905 0.632 0.775 T 0.095 0.368 0.225 0.2676 0.0277 0.5008
rs13285078 8.05 × 10-6 C 0.905 0.632 0.775 G 0.095 0.368 0.225 0.2676 0.0277 0.5008
rs4738955 8.10 × 10-6 A 0.512 0.382 0.450 G 0.488 0.618 0.550 7.60 × 10-7 0.3132 0.0051
rs9984519 8.22 × 10-6 T 0.476 0.316 0.400 C 0.524 0.684 0.600 0.0056 0.0044 0.5762
rs8046503 8.41 × 10-6 A 0.595 0.316 0.463 G 0.405 0.684 0.538 0.0002 0.1791 0.0060
Footnotes for Supplementary Table 7.1 are on page 214.
Appendices 209
Supplementary Table 7.1. Continued Calculation of Hardy-Weinberg equilibrium (HWE) p-value and allele frequencies for 442 SNPs reported by Schlauch and colleagues
(2016) as being associated with chronic fatigue syndrome. SNP ID’s highlighted in red do not meet the reported MAF (< 0.05) or HWE (p < 8 E-4) thresholds in cases,
controls, or the total cohort separately.
SNP Reported GWA p-value RA Freq RA
OA Freq OA HWE p-value
Case Control Total Case Control Total Case Control Total
rs1859512 8.54 × 10-6 T 0.321 0.066 0.200 C 0.679 0.934 0.800 0.0021 0.6642 0.0253
rs5942870 8.88 × 10-6 A 0.810 0.474 0.650 C 0.190 0.526 0.350 4.30 × 10-8 0.1075 5.40 × 10-7
rs2288374 9.09 × 10-6 T 0.381 0.132 0.263 C 0.619 0.868 0.738 0.0074 0.0567 0.3824
rs2891242 9.33 × 10-6 C 0.500 0.171 0.344 T 0.500 0.829 0.656 0.1228 0.3096 0.7864
rs17156195 9.34 × 10-6 T 0.417 0.118 0.275 A 0.583 0.882 0.725 0.0366 0.4076 0.2503
rs873276 9.37 × 10-6 G 0.476 0.237 0.363 T 0.524 0.763 0.638 0.0007 0.0557 0.0299
rs2602803 9.83 × 10-6 G 0.488 0.303 0.400 T 0.512 0.697 0.600 7.60 × 10-7 0.6895 0.0015
rs12572431 9.90 × 10-6 G 0.536 0.250 0.400 A 0.464 0.750 0.600 0.0002 0.2342 0.0069
rs7306948 1.01 × 10-5 G 0.429 0.145 0.294 A 0.571 0.855 0.706 0.0030 0.7893 0.1177
rs17019561 1.02 × 10-5 C 0.238 0.013 0.131 T 0.762 0.987 0.869 0.2401 0.9345 0.7108
rs2033069 1.03 × 10-5 A 0.869 0.539 0.713 G 0.131 0.461 0.288 0.3288 0.9692 0.4489
rs5930684 1.06 × 10-5 G 0.798 0.776 0.788 A 0.202 0.224 0.213 1.90 × 10-9 0.3998 0.0003
rs1366834 1.07 × 10-5 G 0.226 0.013 0.125 A 0.774 0.987 0.875 0.0582 0.9345 0.2013
rs12170932 1.07 × 10-5 T 0.226 0.013 0.125 C 0.774 0.987 0.875 0.0582 0.9345 0.2013
rs12443497 1.07 × 10-5 T 0.226 0.013 0.125 A 0.774 0.987 0.875 0.0582 0.9345 0.2013
rs9744291 1.08 × 10-5 G 0.690 0.276 0.494 A 0.310 0.724 0.506 0.4809 0.0887 0.0037
rs5770525 1.11 × 10-5 G 0.810 0.500 0.663 A 0.190 0.500 0.338 0.1273 0.0231 0.0519
rs2196007 1.11 × 10-5 T 0.310 0.053 0.188 G 0.690 0.947 0.813 0.0290 0.7320 0.1835
rs10402951 1.13 × 10-5 C 0.607 0.342 0.481 T 0.393 0.658 0.519 0.0243 0.0104 0.5098
rs10517378 1.13 × 10-5 C 0.512 0.250 0.388 A 0.488 0.750 0.613 0.0134 0.0231 0.9953
rs4808297 1.16 × 10-5 T 0.381 0.132 0.263 C 0.619 0.868 0.738 0.9503 2.10 × 10-6 0.0096
rs12120556 1.16 × 10-5 A 0.226 0.132 0.181 G 0.774 0.868 0.819 0.0582 2.10 × 10-6 0.3013
rs4692612 1.17 × 10-5 T 0.369 0.092 0.238 G 0.631 0.908 0.763 0.0137 0.5317 0.1209
rs2062758 1.18 × 10-5 T 0.452 0.105 0.288 A 0.548 0.895 0.713 0.8010 0.3187 0.0645
rs7987491 1.19 × 10-5 G 0.440 0.158 0.306 T 0.560 0.842 0.694 0.0013 0.2478 0.0178
rs1597474 1.24 × 10-5 T 0.226 0.053 0.144 C 0.774 0.947 0.856 0.8956 7.10 × 10-10 0.0330
rs17341595 1.27 × 10-5 C 0.810 0.803 0.806 T 0.190 0.197 0.194 4.30 × 10-8 0.1296 0.0042
Footnotes for Supplementary Table 7.1 are on page 214.
210 Appendices
Supplementary Table 7.1. Continued Calculation of Hardy-Weinberg equilibrium (HWE) p-value and allele frequencies for 442 SNPs reported by Schlauch and colleagues
(2016) as being associated with chronic fatigue syndrome. SNP ID’s highlighted in red do not meet the reported MAF (< 0.05) or HWE (p < 8 E-4) thresholds in cases,
controls, or the total cohort separately.
SNP Reported GWA p-value RA Freq RA
OA Freq OA HWE p-value
Case Control Total Case Control Total Case Control Total
rs11954603 1.30 × 10-5 C 0.405 0.237 0.325 T 0.595 0.763 0.675 1.00 × 10-5 0.4357 0.0055
rs16926249 1.31 × 10-5 G 0.738 0.474 0.613 A 0.262 0.526 0.388 0.1333 0.0004 0.0188
rs4236924 1.31 × 10-5 C 0.333 0.079 0.213 T 0.667 0.921 0.788 0.0012 0.5972 0.0158
rs16888306 1.31 × 10-5 C 0.333 0.079 0.213 T 0.667 0.921 0.788 0.0012 0.5972 0.0158
rs7095919 1.31 × 10-5 G 0.333 0.079 0.213 A 0.667 0.921 0.788 0.0012 0.5972 0.0158
rs10483750 1.31 × 10-5 T 0.333 0.079 0.213 C 0.667 0.921 0.788 0.0012 0.5972 0.0158
rs2801659 1.32 × 10-5 C 0.619 0.237 0.438 T 0.381 0.763 0.563 0.9503 0.0936 0.0332
rs16886994 1.32 × 10-5 G 0.274 0.026 0.156 C 0.726 0.974 0.844 0.3727 0.8677 0.9683
rs418216 1.32 × 10-5 T 0.214 0.000 0.113 C 0.786 1.000 0.888 0.3261 NC 0.0260
rs547977 1.32 × 10-5 T 0.179 0.000 0.094 G 0.821 1.000 0.906 0.1589 NC 0.3548
rs17647077 1.32 × 10-5 G 0.179 0.000 0.094 C 0.821 1.000 0.906 0.1589 NC 0.3548
rs2294584 1.32 × 10-5 A 0.179 0.000 0.094 G 0.821 1.000 0.906 0.1589 NC 0.3548
rs7789233 1.32 × 10-5 G 0.821 1.000 0.906 G 0.179 0.000 0.094 0.1589 NC 0.3548
rs7153874 1.32 × 10-5 G 0.190 0.000 0.100 A 0.810 1.000 0.900 0.6002 NC 0.8038
rs1195242 1.32 × 10-5 C 0.190 0.000 0.100 A 0.810 1.000 0.900 0.6002 NC 0.8038
rs11934366 1.32 × 10-5 A 0.190 0.000 0.100 G 0.810 1.000 0.900 0.6002 NC 0.8038
rs10980229 1.41 × 10-5 G 0.500 0.289 0.400 A 0.500 0.711 0.600 6.00 × 10-5 0.0849 0.0154
rs2644567 1.41 × 10-5 G 0.964 0.724 0.850 A 0.036 0.276 0.150 2.20 × 10-5 0.9362 0.0537
rs11009106 1.42 × 10-5 C 0.524 0.329 0.431 G 0.476 0.671 0.569 0.0007 0.0222 0.1547
rs2916699 1.46 × 10-5 A 0.262 0.026 0.150 T 0.738 0.974 0.850 0.1333 0.8677 0.4830
rs10056584 1.46 × 10-5 A 0.262 0.026 0.150 G 0.738 0.974 0.850 0.1333 0.8677 0.4830
rs1696407 1.46 × 10-5 G 0.476 0.197 0.344 A 0.524 0.803 0.656 0.0006 0.1296 0.0069
rs17043470 1.52 × 10-5 A 1.000 0.829 0.919 G 0.000 0.171 0.081 NC 0.2034 0.4289
rs7253295 1.52 × 10-5 A 1.000 0.816 0.913 G 0.000 0.184 0.088 NC 0.7547 0.5874
rs13421497 1.57 × 10-5 G 0.345 0.053 0.206 C 0.655 0.947 0.794 0.4974 0.0036 0.0140
rs4892034 1.58 × 10-5 A 0.321 0.118 0.225 T 0.679 0.882 0.775 0.0979 0.0001 0.5425
rs7159091 1.58 × 10-5 C 0.214 0.039 0.131 T 0.786 0.961 0.869 0.0771 5.70 × 10-5 0.7108
Footnotes for Supplementary Table 7.1 are on page 214.
Appendices 211
Supplementary Table 7.1. Continued Calculation of Hardy-Weinberg equilibrium (HWE) p-value and allele frequencies for 442 SNPs reported by Schlauch and colleagues
(2016) as being associated with chronic fatigue syndrome. SNP ID’s highlighted in red do not meet the reported MAF (< 0.05) or HWE (p < 8 E-4) thresholds in cases,
controls, or the total cohort separately.
SNP Reported GWA p-value RA Freq RA
OA Freq OA HWE p-value
Case Control Total Case Control Total Case Control Total
rs2192217 1.60 × 10-5 A 0.250 0.026 0.144 G 0.750 0.974 0.856 0.0308 0.8677 0.1332
rs17002379 1.60 × 10-5 T 0.512 0.145 0.338 C 0.488 0.855 0.663 0.5393 0.7893 0.0519
rs7107438 1.60 × 10-5 C 0.250 0.026 0.144 T 0.750 0.974 0.856 0.2578 7.10 × 10-10 0.0024
rs16937494 1.62 × 10-5 G 0.274 0.039 0.163 A 0.726 0.961 0.838 0.0145 0.8000 0.0827
rs7783582 1.62 × 10-5 T 0.274 0.039 0.163 C 0.726 0.961 0.838 0.0145 0.8000 0.0827
rs3778315 1.65 × 10-5 G 0.333 0.039 0.194 A 0.667 0.961 0.806 0.3545 0.8000 0.0320
rs3792615 1.66 × 10-5 T 0.976 0.724 0.856 C 0.024 0.276 0.144 0.8744 0.9362 0.2212
rs16927111 1.69 × 10-5 G 0.464 0.211 0.344 C 0.536 0.789 0.656 1.20 × 10-5 0.5043 0.0014
rs12629627 1.76 × 10-5 G 0.631 0.355 0.500 A 0.369 0.645 0.500 0.0018 0.8852 0.1797
rs798368 1.78 × 10-5 T 0.738 0.605 0.675 C 0.262 0.395 0.325 0.0215 0.0006 0.1937
rs17256392 1.81 × 10-5 A 0.167 0.026 0.100 G 0.833 0.974 0.900 0.1949 7.10 × 10-10 0.8038
rs6892871 1.81 × 10-5 G 0.167 0.026 0.100 A 0.833 0.974 0.900 0.1949 7.10 × 10-10 0.8038
rs6055456 1.83 × 10-5 C 0.369 0.105 0.244 T 0.631 0.895 0.756 0.0018 0.4683 0.0228
rs12165212 1.85 × 10-5 G 0.964 0.737 0.856 A 0.036 0.263 0.144 0.8103 0.0277 0.1332
rs41469844 1.86 × 10-5 C 0.726 0.342 0.544 A 0.274 0.658 0.456 0.9081 0.0658 0.0160
rs7020077 1.87 × 10-5 A 0.488 0.184 0.344 G 0.512 0.816 0.656 0.0134 0.7547 0.2241
rs2127978 1.97 × 10-5 T 0.333 0.092 0.219 C 0.667 0.908 0.781 0.0109 0.1887 0.2317
rs1367276 1.97 × 10-5 A 0.762 0.618 0.694 G 0.238 0.382 0.306 0.0428 0.0002 0.0657
rs12034948 2.00 × 10-5 G 0.321 0.053 0.194 A 0.679 0.947 0.806 0.3434 0.7320 0.9982
rs11984468 2.04 × 10-5 C 0.524 0.184 0.363 G 0.476 0.816 0.638 0.3459 0.4431 0.4717
rs16975878 2.06 × 10-5 G 0.988 0.789 0.894 A 0.012 0.211 0.106 0.9378 0.1002 0.2876
rs7747443 2.13 × 10-5 C 0.357 0.053 0.213 G 0.643 0.947 0.788 0.2695 0.7320 0.0236
rs6923953 2.13 × 10-5 C 0.357 0.053 0.213 T 0.643 0.947 0.788 0.2695 0.7320 0.0236
rs6926583 2.13 × 10-5 C 0.357 0.053 0.213 T 0.643 0.947 0.788 0.2695 0.7320 0.0236
rs997139 2.13 × 10-5 G 0.357 0.053 0.213 A 0.643 0.947 0.788 0.2695 0.7320 0.0236
rs5956823 2.13 × 10-5 A 0.917 0.658 0.794 C 0.083 0.342 0.206 0.5558 0.0130 0.1008
rs985257 2.15 × 10-5 A 0.571 0.237 0.413 T 0.429 0.763 0.588 0.2801 0.0936 0.2707
Footnotes for Supplementary Table 7.1 are on page 214.
212 Appendices
Supplementary Table 7.1. Continued Calculation of Hardy-Weinberg equilibrium (HWE) p-value and allele frequencies for 442 SNPs reported by Schlauch and colleagues
(2016) as being associated with chronic fatigue syndrome. SNP ID’s highlighted in red do not meet the reported MAF (< 0.05) or HWE (p < 8 E-4) thresholds in cases,
controls, or the total cohort separately.
SNP Reported GWA p-value RA Freq RA
OA Freq OA HWE p-value
Case Control Total Case Control Total Case Control Total
rs7853174 2.17 × 10-5 G 0.571 0.513 0.544 A 0.429 0.487 0.456 0.0193 0.0001 0.2902
rs2274515 2.19 × 10-5 T 0.619 0.316 0.475 C 0.381 0.684 0.525 0.0074 0.1791 0.0694
rs7279994 2.22 × 10-5 C 0.500 0.211 0.363 A 0.500 0.789 0.638 0.0055 0.7579 0.2242
rs16970887 2.25 × 10-5 C 0.310 0.066 0.194 G 0.690 0.934 0.806 0.0037 0.6642 0.0316
rs17123453 2.25 × 10-5 C 0.310 0.066 0.194 G 0.690 0.934 0.806 0.0037 0.6642 0.0316
rs2648883 2.27 × 10-5 G 0.583 0.250 0.425 A 0.417 0.750 0.575 0.4122 0.0231 0.1043
rs11711551 2.31 × 10-5 G 0.440 0.132 0.294 A 0.560 0.868 0.706 0.1785 0.6272 0.9584
rs3861911 2.31 × 10-5 C 0.393 0.132 0.269 T 0.607 0.868 0.731 0.0004 0.3503 0.0066
rs7426702 2.31 × 10-5 T 0.393 0.132 0.269 C 0.607 0.868 0.731 0.0004 0.3503 0.0066
rs7540424 2.37 × 10-5 A 0.750 0.553 0.656 G 0.250 0.447 0.344 9.70 × 10-6 0.2922 0.0242
rs2988013 2.38 × 10-5 C 0.310 0.053 0.188 T 0.690 0.947 0.813 0.1440 0.7320 0.5510
rs4393091 2.42 × 10-5 G 0.500 0.171 0.344 C 0.500 0.829 0.656 3.70 × 10-6 0.8982 2.20 × 10-6
rs7863401 2.44 × 10-5 A 0.905 0.645 0.781 G 0.095 0.355 0.219 0.2676 0.0477 0.5880
rs12684292 2.44 × 10-5 T 0.905 0.645 0.781 C 0.095 0.355 0.219 0.2676 0.0477 0.5880
rs2421987 2.47 × 10-5 A 0.214 0.132 0.175 G 0.786 0.868 0.825 0.0771 2.10 × 10-6 0.2300
rs7307225 2.53 × 10-5 G 0.595 0.316 0.463 A 0.405 0.684 0.538 0.0650 0.0159 0.3959
rs10817082 2.53 × 10-5 C 0.917 0.632 0.781 G 0.083 0.368 0.219 0.5558 0.4195 0.9105
rs9485028 2.53 × 10-5 A 0.262 0.066 0.169 G 0.738 0.934 0.831 0.0215 0.0274 0.3083
rs654807 2.57 × 10-5 C 0.500 0.421 0.463 T 0.500 0.579 0.538 0.0007 0.0046 0.6168
rs5948927 2.60 × 10-5 T 0.524 0.197 0.369 C 0.476 0.803 0.631 6.20 × 10-5 0.6229 9.60 × 10-5
rs17733133 2.60 × 10-5 G 0.500 0.171 0.344 A 0.500 0.829 0.656 0.3545 0.3096 0.4433
rs10207238 2.62 × 10-5 C 0.214 0.013 0.119 T 0.786 0.987 0.881 0.0771 0.9345 0.2281
rs29110 2.62 × 10-5 C 0.214 0.013 0.119 A 0.786 0.987 0.881 0.0771 0.9345 0.2281
rs11506050 2.62 × 10-5 G 0.214 0.013 0.119 C 0.786 0.987 0.881 0.0771 0.9345 0.2281
rs6098723 2.62 × 10-5 G 0.214 0.013 0.119 A 0.786 0.987 0.881 0.0771 0.9345 0.2281
rs8016502 2.62 × 10-5 G 0.214 0.013 0.119 A 0.786 0.987 0.881 0.0771 0.9345 0.2281
rs8073194 2.62 × 10-5 T 0.214 0.013 0.119 C 0.786 0.987 0.881 0.0771 0.9345 0.2281
Footnotes for Supplementary Table 7.1 are on page 214.
Appendices 213
Supplementary Table 7.1. Continued Calculation of Hardy-Weinberg equilibrium (HWE) p-value and allele frequencies for 442 SNPs reported by Schlauch and colleagues
(2016) as being associated with chronic fatigue syndrome. SNP ID’s highlighted in red do not meet the reported MAF (< 0.05) or HWE (p < 8 E-4) thresholds in cases,
controls, or the total cohort separately.
SNP Reported GWA p-value RA Freq RA
OA Freq OA HWE p-value
Case Control Total Case Control Total Case Control Total
rs11895045 2.62 × 10-5 T 0.214 0.013 0.119 C 0.786 0.987 0.881 0.0771 0.9345 0.2281
rs11154872 2.64 × 10-5 C 0.345 0.053 0.206 T 0.655 0.947 0.794 0.4974 0.7320 0.0761
rs13052044 2.66 × 10-5 T 0.500 0.158 0.338 C 0.500 0.842 0.663 0.7576 0.9488 0.3453
rs3920498 2.68 × 10-5 C 0.952 0.711 0.838 G 0.048 0.289 0.163 0.7459 0.0849 0.3607
rs2685850 2.70 × 10-5 A 0.726 0.316 0.531 G 0.274 0.684 0.469 0.1508 0.0970 0.0009
rs1433429 2.71 × 10-5 C 0.488 0.197 0.350 T 0.512 0.803 0.650 0.0134 0.5946 0.3764
rs13148734 2.73 × 10-5 A 0.786 0.500 0.650 G 0.214 0.500 0.350 0.0771 0.1048 0.2796
rs5912816 2.74 × 10-5 C 0.821 0.487 0.663 T 0.179 0.513 0.338 0.0051 0.9966 0.0145
rs10144861 2.74 × 10-5 G 0.440 0.145 0.300 A 0.560 0.855 0.700 0.0487 0.7893 0.5229
rs622060 2.76 × 10-5 C 0.690 0.513 0.606 T 0.310 0.487 0.394 1.60 × 10-5 0.1926 0.0313
rs6877860 2.78 × 10-5 T 0.357 0.145 0.256 C 0.643 0.855 0.744 0.3617 2.70 × 10-5 0.1071
rs16944757 2.80 × 10-5 A 0.298 0.053 0.181 G 0.702 0.947 0.819 0.0447 0.7320 0.2200
rs2025499 2.82 × 10-5 G 0.333 0.079 0.213 A 0.667 0.921 0.788 0.0109 0.5972 0.0809
rs7834482 2.83 × 10-5 G 0.714 0.461 0.594 A 0.286 0.539 0.406 0.2801 0.0001 0.0072
rs11090847 2.92 × 10-5 T 0.429 0.171 0.306 C 0.571 0.829 0.694 0.0003 0.8982 0.0178
rs1977985 2.96 × 10-5 G 0.964 0.750 0.863 A 0.036 0.250 0.138 2.20 × 10-5 0.7456 0.1608
rs12761944 2.97 × 10-5 A 0.512 0.342 0.431 C 0.488 0.658 0.569 0.0020 0.0130 0.3334
rs1763788 2.97 × 10-5 A 0.512 0.342 0.431 G 0.488 0.658 0.569 0.0020 0.0130 0.3334
rs1577372 2.97 × 10-5 A 0.512 0.342 0.431 G 0.488 0.658 0.569 0.0020 0.0130 0.3334
rs1762529 2.97 × 10-5 A 0.512 0.342 0.431 G 0.488 0.658 0.569 0.0020 0.0130 0.3334
rs2490495 2.97 × 10-5 G 0.512 0.342 0.431 C 0.488 0.658 0.569 0.0020 0.0130 0.3334
rs2784574 2.97 × 10-5 G 0.512 0.342 0.431 A 0.488 0.658 0.569 0.0020 0.0130 0.3334
rs2995467 2.97 × 10-5 G 0.512 0.342 0.431 A 0.488 0.658 0.569 0.0020 0.0130 0.3334
rs2960770 3.01 × 10-5 C 0.869 0.539 0.713 T 0.131 0.461 0.288 0.0020 0.5390 0.0033
rs17675581 3.05 × 10-5 G 0.512 0.276 0.400 A 0.488 0.724 0.600 0.0002 0.1230 0.0154
rs1526415 3.05 × 10-5 T 0.381 0.132 0.263 A 0.619 0.868 0.738 6.70 × 10-5 0.3503 0.0015
rs10789931 3.08 × 10-5 T 0.619 0.303 0.469 C 0.381 0.697 0.531 0.0428 0.6895 0.7952
Footnotes for Supplementary Table 7.1 are on page 214.
214 Appendices
Supplementary Table 7.1. Continued Calculation of Hardy-Weinberg equilibrium (HWE) p-value and allele frequencies for 442 SNPs reported by Schlauch and colleagues
(2016) as being associated with chronic fatigue syndrome. SNP ID’s highlighted in red do not meet the reported MAF (< 0.05) or HWE (p < 8 E-4) thresholds in cases,
controls, or the total cohort separately.
SNP Reported GWA p-value RA Freq RA
OA Freq OA HWE p-value
Case Control Total Case Control Total Case Control Total
rs2015035 3.10 × 10-5 T 0.333 0.039 0.194 G 0.667 0.961 0.806 0.1052 0.8000 0.0042
rs2436739 3.15 × 10-5 G 0.190 0.000 0.100 A 0.810 1.000 0.900 0.6337 NC 0.1360
rs3798405 3.15 × 10-5 G 0.167 0.000 0.088 A 0.833 1.000 0.913 0.1949 NC 0.3911
rs7859623 3.15 × 10-5 C 0.167 0.000 0.088 T 0.833 1.000 0.913 0.1949 NC 0.3911
rs6744124 3.15 × 10-5 C 0.167 0.000 0.088 T 0.833 1.000 0.913 0.1949 NC 0.3911
rs4505649 3.15 × 10-5 G 0.167 0.000 0.088 A 0.833 1.000 0.913 0.1949 NC 0.3911
rs7517843 3.15 × 10-5 C 0.167 0.000 0.088 A 0.833 1.000 0.913 0.1949 NC 0.3911
rs1926721 3.15 × 10-5 G 0.167 0.000 0.088 A 0.833 1.000 0.913 0.1949 NC 0.3911
rs1458597 3.15 × 10-5 A 0.167 0.000 0.088 G 0.833 1.000 0.913 0.1949 NC 0.3911
rs2303409 3.19 × 10-5 C 0.417 0.132 0.281 A 0.583 0.868 0.719 0.8531 0.0009 0.0423
rs4130583 3.20 × 10-5 G 0.476 0.211 0.350 C 0.524 0.789 0.650 3.70 × 10-6 0.5043 5.60 × 10-5
rs6074914 3.20 × 10-5 A 0.940 0.658 0.806 G 0.060 0.342 0.194 0.0190 0.6904 0.0320
rs927651 3.21 × 10-5 G 0.714 0.461 0.594 A 0.286 0.539 0.406 0.0095 0.5390 0.5771
rs11972875 3.27 × 10-5 G 0.286 0.053 0.175 A 0.714 0.947 0.825 0.0095 0.7320 0.0578
rs1428323 3.28 × 10-5 A 0.464 0.237 0.356 G 0.536 0.763 0.644 1.20 × 10-5 0.9060 0.0027
rs9683305 3.28 × 10-5 C 0.798 0.447 0.631 T 0.202 0.553 0.369 0.4912 0.3601 0.1338
RA: risk allele; OA: other allele, Freq: frequency; HWE: Hardy-Weinberg equilibrium.
Appendices 215
Supplementary Table 7.2. Summary of association results for SNPs from previous chronic fatigue
syndrome (CFS) candidate gene association analyses within our CFS and fatigue cohorts.
SNP Chr SNP position RA OA Freq OR (95% CI) p-value Effect
directiona
CFS cohort (Bonferroni adjusted p-value = 0.0011)
rs655207 13 38368012 G T 0.346 2.67 (1.51-4.72) 0.0006 +
rs4738202 8 72940861 A G 0.218 2.77 (1.51-5.10) 0.0009 +
rs6650469 13 38367949 T C 0.364 2.47 (1.40-4.34) 0.0016 + rs6429157 1 239981643 G A 0.346 2.06 (1.17-3.63) 0.0114 +
rs12914385 15 78898723 T C 0.264 1.98 (1.10-3.57) 0.0224 +
rs951266 15 78878541 T C 0.218 2.03 (1.10-3.77) 0.0235 + rs2383844 8 72961252 G A 0.409 1.71 (0.98-2.99) 0.0570 +
rs12602006 17 16337288 A G 0.582 1.69 (0.95-3.03) 0.0748 +
rs2918419 5 142722353 G A 0.155 1.77 (0.88-3.57) 0.1060 - rs1317103 9 73195703 T C 0.727 1.70 (0.87-3.33) 0.1203 -
rs7520974 1 240067260 A G 0.500 1.54 (0.88-2.69) 0.1280 +
rs6578398 11 3638061 G A 0.700 1.49 (0.79-2.81) 0.2164 -
rs3829603 17 7347042 A C 0.273 1.44 (0.79-2.62) 0.2274 -
rs4620530 1 240063821 G T 0.473 1.38 (0.79-2.40) 0.2518 -
rs3743074 15 78909480 T C 0.600 1.35 (0.76-2.41) 0.2999 + rs589962 1 239989964 T C 0.664 1.33 (0.73-2.42) 0.3570 +
rs2741343 8 27326127 C T 0.436 1.29 (0.74-2.25) 0.3637 +
rs891398 8 27324822 C T 0.436 1.29 (0.74-2.25) 0.3637 + rs3763619 9 73225802 C A 0.600 1.29 (0.73-2.29) 0.3804 -
rs1424569 7 136569416 A G 0.418 1.28 (0.73-2.22) 0.3859 + rs7865858 9 73204431 G A 0.591 1.28 (0.72-2.26) 0.3953 -
rs2075748 11 62688269 A G 0.218 1.30 (0.68-2.47) 0.4259 +
rs1867263 1 239807920 A G 0.282 1.25 (0.69-2.28) 0.4578 - rs11563204 2 234917377 A G 0.182 1.30 (0.65-2.57) 0.4600 +
rs10780950 9 73193428 C T 0.791 1.29 (0.64-2.62) 0.4815 -
rs6188 5 142680344 T G 0.315 1.23 (0.69-2.21) 0.4818 - rs7543259 1 239979186 A G 0.227 1.23 (0.65-2.33) 0.5220 +
rs11142508 9 73231662 T C 0.627 1.21 (0.68-2.15) 0.5224 -
rs11823728 11 62676802 C T 0.964 1.74 (0.31-9.70) 0.5250 + rs11224816 11 101396286 T C 0.491 1.18 (0.68-2.05) 0.5592 +
rs763780 6 52101739 C T 0.045 1.43 (0.42-4.85) 0.5624 -
rs2767 2 233400074 T C 0.600 1.18 (0.67-2.08) 0.5747 + rs852977 5 142687494 G A 0.327 1.17 (0.65-2.08) 0.6056 -
rs10925941 1 239812538 A G 0.318 1.16 (0.65-2.08) 0.6195 -
rs10754677 1 239833100 A G 0.564 1.14 (0.65-1.99) 0.6434 + rs2071167 17 42287519 G A 0.727 1.16 (0.62-2.17) 0.6489 -
rs17865678 2 234919314 A G 0.264 1.13 (0.61-2.08) 0.7066 +
rs1328153 9 73416062 T C 0.800 1.13 (0.56-2.29) 0.7288 - rs7669882 4 40350651 G A 0.682 1.10 (0.61-2.00) 0.7543 -
rs4861065 4 40344395 T C 0.673 1.09 (0.60-1.97) 0.7744 -
rs6313 13 47469940 C T 0.591 1.07 (0.61-1.87) 0.8222 - rs1891301 9 74018496 C T 0.446 1.05 (0.60-1.83) 0.8638 -
rs685550 1 239924408 C T 0.236 1.05 (0.55-1.99) 0.8898 +
rs1860661 19 1650134 G A 0.355 1.03 (0.58-1.83) 0.9154 + rs603152 15 34294637 A C 0.355 1.03 (0.58-1.83) 0.9154 +
rs6311 13 47471478 C T 0.591 1.02 (0.58-1.79) 0.9441 -
rs10009228 4 40356422 A G 0.136 1.02 (0.46-2.26) 0.9681 - Fatigue cohort (Bonferroni adjusted p-value = 0.0007)
rs10115622 9 73306551 C A 0.684 1.04 (1.00-1.09) 0.0774 +
rs726169 1 239794277 T C 0.671 1.03 (0.99-1.08) 0.1396 + rs10009228 4 40356422 A G 0.184 1.04 (0.99-1.09) 0.1558 -
rs1867264 1 239845277 T A 0.358 1.03 (0.99-1.08) 0.1584 +
rs4861323 4 40355815 G A 0.183 1.04 (0.98-1.09) 0.1708 - rs3762529 2 233392449 A G 0.615 1.03 (0.99-1.08) 0.1898 +
rs4973537 2 233391965 A G 0.614 1.03 (0.98-1.07) 0.2113 +
rs2083817 1 239833605 T A 0.375 1.03 (0.98-1.07) 0.2355 + rs1867265 1 239840107 T C 0.357 1.03 (0.98-1.07) 0.2452 -
rs7865858 9 73204431 A G 0.372 1.02 (0.98-1.07) 0.2784 +
rs12463989 2 233395029 T C 0.628 1.02 (0.98-1.07) 0.2853 + rs7520974 1 240067260 G A 0.456 1.02 (0.98-1.06) 0.2949 -
rs7551001 1 239844600 G A 0.385 1.02 (0.98-1.07) 0.2951 -
rs2165872 1 239826988 A G 0.375 1.02 (0.98-1.07) 0.2964 - rs6669810 1 240068629 G C 0.454 1.02 (0.98-1.06) 0.3034 -
rs7180002 15 78873993 A T 0.663 1.02 (0.98-1.07) 0.3137 -
rs951266 15 78878541 G A 0.663 1.02 (0.98-1.07) 0.3186 - rs2767 2 233400074 A G 0.628 1.02 (0.98-1.07) 0.3279 +
rs12093821 1 239824248 A G 0.401 1.02 (0.98-1.06) 0.3925 -
Footnotes for Supplementary Table 7.2 are on page 216.
216 Appendices
Supplementary Table 7.2. Continued Summary of association results for SNPs from previous
chronic fatigue syndrome (CFS) candidate gene association analyses within our CFS and fatigue
cohorts.
SNP Chr SNP position RA OA Freq OR (95% CI) p-value Effect
directiona
rs16838637 1 239828350 G A 0.402 1.02 (0.98-1.06) 0.4339 -
rs10754677 1 239833100 G A 0.444 1.02 (0.97-1.06) 0.4580 -
rs12914385 15 78898723 C T 0.611 1.02 (0.97-1.06) 0.4605 - rs6188 5 142680344 A C 0.322 1.01 (0.97-1.06) 0.5016 -
rs852977 5 142687494 G A 0.324 1.01 (0.97-1.06) 0.5052 -
rs1106948 9 74017174 T C 0.547 1.01 (0.97-1.06) 0.5289 + rs1899616 1 239818568 T C 0.389 1.01 (0.97-1.06) 0.5381 -
rs1134 1 239872172 T C 0.405 1.01 (0.97-1.06) 0.5389 -
rs3738436 1 239872493 T G 0.405 1.01 (0.97-1.06) 0.5389 - rs763780 6 52101739 T C 0.957 1.03 (0.93-1.14) 0.5420 +
rs6578398 11 3638061 A G 0.265 1.01 (0.97-1.06) 0.5480 +
rs1891301 9 74018496 T C 0.538 1.01 (0.97-1.06) 0.5485 + rs6429157 1 239981643 G A 0.426 1.01 (0.97-1.05) 0.5551 +
rs12682832 9 73220691 G A 0.626 1.01 (0.97-1.06) 0.5581 -
rs7511970 1 239883255 A G 0.407 1.01 (0.97-1.06) 0.5638 - rs7513746 1 239862411 G A 0.404 1.01 (0.97-1.05) 0.5942 -
rs619214 1 239958622 T G 0.527 1.01 (0.97-1.05) 0.6080 +
rs4243084 15 78911672 G C 0.655 1.01 (0.97-1.05) 0.6210 + rs1373998 5 55255565 G A 0.877 1.02 (0.95-1.08) 0.6227 -
rs6694220 1 239883616 G A 0.483 1.01 (0.97-1.05) 0.6461 -
rs6684622 1 239877537 C G 0.43 1.01 (0.97-1.05) 0.6791 - rs7108612 11 3650086 T G 0.103 1.01 (0.95-1.08) 0.6930 +
rs655207 13 38368012 G T 0.424 1.01 (0.97-1.05) 0.6959 +
rs10802795 1 239870775 C T 0.431 1.01 (0.97-1.05) 0.7193 - rs2985167 13 38230542 A G 0.627 1.01 (0.97-1.05) 0.7414 +
rs6429147 1 239794794 C G 0.374 1.01 (0.97-1.05) 0.7614 -
rs2869546 15 78907345 T C 0.626 1.01 (0.96-1.05) 0.7680 + rs6560200 9 73980222 C T 0.531 1.01 (0.97-1.05) 0.7816 +
rs1160742 9 73314011 G A 0.601 1.01 (0.96-1.05) 0.7972 -
rs6650469 13 38367949 T C 0.43 1.00 (0.97-1.05) 0.8285 + rs511422 15 34282982 A G 0.66 1.00 (0.95-1.04) 0.8330 -
rs4861065 4 40344395 T C 0.684 1.00 (0.96-1.05) 0.8350 - rs2741343 8 27326127 A G 0.505 1.00 (0.96-1.05) 0.8442 -
rs3743075 15 78909452 T C 0.367 1.00 (0.95-1.04) 0.8600 -
rs603152 15 34294637 G T 0.645 1.00 (0.95-1.04) 0.8675 - rs1867263 1 239807920 G A 0.637 1.00 (0.96-1.04) 0.8681 +
rs3743074 15 78909480 G A 0.368 1.00 (0.95-1.04) 0.8707 -
rs10925941 1 239812538 G A 0.612 1.00 (0.96-1.04) 0.8954 + rs2302767 17 7350544 A G 0.713 1.00 (0.96-1.05) 0.9029 +
rs860458 5 142696036 G A 0.832 1.00 (0.95-1.06) 0.9059 +
rs2918419 5 142722353 T C 0.832 1.00 (0.95-1.06) 0.9076 + rs17865678 2 234919314 G A 0.736 1.00 (0.96-1.05) 0.9247 -
rs6758653 2 234912799 G A 0.667 1.00 (0.96-1.04) 0.9468 +
rs891398 8 27324822 T C 0.508 1.00 (0.96-1.04) 0.9755 - rs1799724 6 31542482 C T 0.936 1.00 (0.92-1.09) 0.9775 -
rs3752411 14 21968876 G A 0.863 1.00 (0.94-1.06) 0.9829 -
rs7669882 4 40350651 G A 0.69 1.00 (0.96-1.05) 0.9840 -
rs6700643 1 239798921 T C 0.612 1.00 (0.96-1.04) 0.9851 +
rs3829603 17 7347042 C A 0.715 1.00 (0.96-1.05) 0.9856 +
rs11142508 9 73231662 T C 0.623 1.00 (0.96-1.04) 0.9878 - rs646950 15 34291660 G A 0.646 1.00 (0.96-1.04) 0.9879 -
rs11563204 2 234917377 G A 0.79 1.00 (0.95-1.05) 0.9915 -
rs10015231 4 40337566 C T 0.787 1.00 (0.95-1.05) 0.9952 +
Chr: chromosome; RA: risk allele; OA: other allele; OR: odds ratio; CI: confidence interval. aSNPs with effects in the same direction are indicated by + while SNPs with effects in opposite directions are
indicated by -.
Appendices 217
Supplementary Table 7.3. Summary of association results for genes from previous chronic fatigue
syndrome (CFS) candidate gene association analyses within our CFS and fatigue cohorts.
Gene Chromosome Start Stop Number of SNPs p-value
CFS cohort (Bonferroni adjusted p-value = 0.0013)
CHRNA5 15 78857862 78887611 7 0.0358 TRPC4 13 38210773 38443939 72 0.0360
TRPC6 11 101322295 101454687 33 0.0372
TRPA1 8 72933486 72987819 28 0.0466 TRPV3 17 3413796 3461289 17 0.0546
EIF3A 10 120794541 120840959 5 0.0773
TRPM4 19 49661016 49715098 9 0.0790 CHRNE 17 4801064 4806369 3 0.0878
TRPV2 17 16318856 16340317 7 0.1176
CHRNA3 15 78885394 78913637 13 0.1314 BMP2K 4 79697532 79837519 5 0.1640
CHRNB4 15 78916636 78933587 5 0.1852
UBTF 17 42282401 42298994 3 0.1894
CHRM5 15 34260446 34357295 18 0.2313
SORL1 11 121322912 121504471 53 0.2806
CHRM3 1 239549876 240078750 119 0.2989 NR3C1 5 142657496 143113322 139 0.3168
CHRNA4 20 61974662 62009487 6 0.3189
TRPM8 2 234826043 234928166 57 0.3408 CHRNB1 17 7348406 7360932 4 0.3567
CHRNA10 11 3686817 3692614 1 0.4366 CHRM1 11 62676151 62689012 4 0.4504
IFNG 12 68548550 68553521 1 0.4723
CHRNA2 8 27317278 27336813 17 0.4845 TCF3 19 1609289 1652328 4 0.5453
CHRNG 2 233404424 233411038 1 0.5908
METTL3 14 21966282 21979517 7 0.6675 DISC1 1 231762561 232177018 136 0.6966
PEX16 11 45931220 45939674 2 0.7277
CHRND 2 233390870 233401375 2 0.7531 IL6ST 5 55230923 55290821 6 0.8501
CHRM2 7 136553399 136703720 41 0.8671
TRPM3 9 73143979 74061782 258 0.8729 HTR2A 13 47405677 47471211 41 0.8955
FAM126B 2 201838441 201936392 2 0.9344
TNF 6 31543344 31546113 3 0.9663 CHRNA9 4 40337346 40357234 10 0.9795
Fatigue cohort (Bonferroni adjusted p-value = 0.0014)
TRPA1 8 72933486 72987819 126 0.0376 TRPV2 17 16318856 16340317 9 0.0632
TRPC4 13 38210773 38443939 296 0.1494
HTR2A 13 47405677 47471211 76 0.1561 BMP2K 4 79697532 79837519 1 0.1734
TRPM3 9 73143979 74061782 1122 0.2399
CHRNA9 4 40337346 40357234 72 0.2451 CHRND 2 233390870 233401375 9 0.3086
CHRM3 1 239549876 240078750 384 0.4147
SORL1 11 121322912 121504471 122 0.4236 IFNG 12 68548550 68553521 3 0.5242
CHRNG 2 233404424 233411038 6 0.5350
PEX16 11 45931220 45939674 1 0.5439 TRPM4 19 49661016 49715098 4 0.5469
TRPV3 17 3413796 3461289 15 0.5518
IL6ST 5 55230923 55290821 28 0.5932 CHRM2 7 136553399 136703720 188 0.6297
UBTF 17 42282401 42298994 8 0.6546
CHRNA5 15 78857862 78887611 32 0.7246 CHRM5 15 34260446 34357295 95 0.7309
CHRNA2 8 27317278 27336813 31 0.7533
CHRM1 11 62676151 62689012 2 0.7681 CHRNA3 15 78885394 78913637 35 0.8059
CHRNB4 15 78916636 78933587 2 0.8265
METTL3 14 21966282 21979517 16 0.8288 NR3C1 5 142657496 143113322 371 0.8289
DISC1 1 231762561 232177018 235 0.8552
CHRNA4 20 61974662 62009487 3 0.8631 CHRNB1 17 7348406 7360932 4 0.8869
TRPM8 2 234826043 234928166 140 0.8915
CHRNE 17 4801064 4806369 3 0.8991
218 Appendices
Supplementary Table 7.3. Continued Summary of association results for genes from previous
chronic fatigue syndrome (CFS) candidate gene association analyses within our CFS and fatigue
cohorts.
Gene Chromosome Start Stop Number of SNPs p-value
FAM126B 2 201838441 201936392 39 0.9141 TCF3 19 1609289 1652328 1 0.9720
TRPC6 11 101322295 101454687 399 0.9811
TNF 6 31543344 31546113 2 0.9898
Appendices 219
Supplementary Table 7.4. Summary of association results for SNPs from previous chronic fatigue
syndrome (CFS) genome-wide association analyses within our CFS and fatigue cohorts.
SNP Chr SNP position RA OA Freq OR (95% CI) p-value Effect
directiona
CFS cohort (Bonferroni adjusted p-value = 0.0006)
rs400322 19 55172578 G A 0.627 1.94 (1.05-3.59) 0.0326 +
rs197770 3 37515827 T C 0.827 2.59 (1.04-6.48) 0.0359 -
rs9200 5 41142606 G A 0.482 1.81 (1.03-3.18) 0.0369 + rs10500964 11 23596570 T C 0.082 2.13 (0.89-5.12) 0.0858 +
rs10509412 10 89599354 C A 0.300 1.66 (0.93-2.95) 0.0870 +
rs3751488 14 23304094 G A 0.782 1.91 (0.90-4.06) 0.0909 + rs10506025 12 27726370 A G 0.327 1.59 (0.90-2.81) 0.1097 -
rs8177374 11 126162843 C T 0.855 2.12 (0.83-5.39) 0.1101 -
rs4692612 4 171537901 G T 0.955 4.43 (0.51-38.63) 0.1424 - rs2389957 4 120695322 G A 0.582 1.53 (0.86-2.73) 0.1448 +
rs1051007 17 4636813 C T 0.127 1.74 (0.82-3.69) 0.1478 +
rs11658971 17 4637698 A G 0.127 1.74 (0.82-3.69) 0.1478 +
rs1931035 13 79274470 G A 0.809 1.78 (0.81-3.92) 0.1480 +
rs1061147 1 196654324 C A 0.618 1.53 (0.85-2.76) 0.1546 +
rs1325904 10 90280938 C T 0.736 1.62 (0.83-3.18) 0.1584 - rs16956158 17 6594844 A G 0.691 1.56 (0.83-2.93) 0.1692 -
rs11575584 9 34661994 A G 0.055 2.06 (0.72-5.91) 0.1698 +
rs9946817 18 70367007 G A 0.173 1.55 (0.78-3.07) 0.2052 + rs4894505 3 175920884 T G 0.618 1.46 (0.81-2.62) 0.2081 -
rs7307225 12 71898358 T C 0.900 1.98 (0.66-5.91) 0.2151 - rs1359536 13 79275793 T C 0.873 1.81 (0.70-4.70) 0.2160 +
rs3802814 11 126162607 G A 0.873 1.81 (0.70-4.70) 0.2160 -
rs4242391 8 23000183 T C 0.373 1.42 (0.81-2.49) 0.2204 - rs6710681 2 64413221 T C 0.380 1.38 (0.79-2.42) 0.2630 -
rs3020729 2 87012293 T C 0.791 1.51 (0.73-3.14) 0.2664 +
rs7121660 11 45358483 G A 0.709 1.43 (0.75-2.69) 0.2731 - rs10737169 1 154653704 T C 0.082 1.64 (0.66-4.09) 0.2828 +
rs4251545 12 44180295 A G 0.082 1.64 (0.66-4.09) 0.2828 -
rs2648883 8 129076594 G A 0.146 1.49 (0.72-3.09) 0.2845 + rs12317807 12 130395910 T C 0.018 2.40 (0.43-13.41) 0.3044 +
rs10501068 11 26769636 G T 0.409 1.33 (0.76-2.31) 0.3181 +
rs228941 22 37523721 G C 0.682 1.36 (0.74-2.51) 0.3236 - rs33013 5 80060016 G A 0.682 1.36 (0.74-2.51) 0.3236 +
rs4151667 6 31914024 A T 0.027 2.00 (0.47-8.62) 0.3418 +
rs2274515 6 42933526 G A 0.955 2.19 (0.41-11.56) 0.3443 - rs7849492 9 122619031 T C 0.955 2.19 (0.41-11.56) 0.3443 -
rs7994531 13 42977439 T C 0.182 1.38 (0.70-2.72) 0.3578 -
rs10402951 19 7555092 G A 0.282 1.32 (0.73-2.39) 0.3665 + rs4714468 6 41452996 A G 0.155 1.39 (0.67-2.85) 0.3742 -
rs7616342 3 19433647 A G 0.536 1.27 (0.73-2.22) 0.3940 -
rs3774268 3 186954324 A G 0.100 1.44 (0.61-3.40) 0.3974 + rs4253760 22 46622384 T G 0.787 1.32 (0.65-2.68) 0.4426 -
rs547571 13 97231270 G A 0.873 1.38 (0.57-3.34) 0.4779 +
rs549908 11 112020916 G T 0.255 1.24 (0.67-2.30) 0.4894 + rs8057267 16 87541080 T C 0.936 1.53 (0.43-5.39) 0.5063 -
rs2200706 3 3673608 G A 0.864 1.33 (0.57-3.11) 0.5151 -
rs7853174 9 129419990 A G 0.455 1.20 (0.69-2.08) 0.5170 - rs3095168 17 16260198 G A 0.964 1.74 (0.31-9.70) 0.5250 -
rs17500510 6 32712818 A G 0.091 1.33 (0.54-3.27) 0.5407 +
rs2249954 14 92383999 C T 0.027 1.59 (0.35-7.27) 0.5501 + rs6757543 2 45977472 A C 0.891 1.32 (0.51-3.37) 0.5658 -
rs3803568 15 75108636 T C 0.046 1.41 (0.41-4.76) 0.5838 -
rs2421987 1 172100831 A G 0.164 1.21 (0.59-2.49) 0.6029 + rs14541 12 8800566 T C 0.870 1.25 (0.53-2.97) 0.6105 -
rs1144418 12 65293514 A C 0.246 1.18 (0.63-2.20) 0.6132 -
rs997139 6 136751118 T C 0.809 1.15 (0.56-2.36) 0.7022 - rs1859512 7 8535486 T C 0.091 1.19 (0.47-3.00) 0.7110 +
rs11257804 10 12496055 G A 0.700 1.12 (0.61-2.06) 0.7132 -
rs10505778 12 14125564 G A 0.327 1.11 (0.62-1.99) 0.7203 - rs10509958 10 114054601 G A 0.318 1.11 (0.62-1.99) 0.7360 -
rs6108 14 95058631 T A 0.573 1.10 (0.63-1.92) 0.7396 -
rs1050998 17 4638737 C T 0.382 1.10 (0.63-1.93) 0.7435 - rs10980229 9 112925977 C T 0.236 1.11 (0.58-2.10) 0.7537 +
rs2278831 19 52131119 G A 0.064 1.18 (0.40-3.51) 0.7603 +
rs1555589 13 100480664 A G 0.618 1.09 (0.62-1.93) 0.7671 + rs11711551 3 66697961 G A 0.055 1.18 (0.37-3.80) 0.7788 +
rs7540424 1 116721566 A G 0.973 1.29 (0.21-7.89) 0.7825 +
Footnotes for Supplementary Table 7.4 are on page 222.
220 Appendices
Supplementary Table 7.4. Continued Summary of association results for SNPs from previous
chronic fatigue syndrome (CFS) genome-wide association analyses within our CFS and fatigue
cohorts.
SNP Chr SNP position RA OA Freq OR (95% CI) p-value Effect
directiona
rs2395655 6 36645696 A G 0.646 1.06 (0.60-1.90) 0.8329 -
rs17643851 8 18437729 G A 0.027 1.18 (0.23-5.97) 0.8449 +
rs12120556 1 172115154 T C 0.182 1.07 (0.53-2.16) 0.8596 + rs829370 1 21933193 C T 0.100 1.07 (0.43-2.65) 0.8811 +
rs726817 10 95459817 T C 0.679 1.05 (0.57-1.91) 0.8850 +
rs11214105 11 112037653 A G 0.236 1.05 (0.55-1.99) 0.8898 + rs2071800 6 32714143 T C 0.091 1.06 (0.41-2.73) 0.9057 +
rs543736 8 104012949 A G 0.355 1.03 (0.58-1.83) 0.9154 +
rs10496982 2 146308114 A C 0.900 1.05 (0.42-2.65) 0.9189 + rs1048829 2 203430456 T G 0.473 1.02 (0.59-1.78) 0.9319 +
rs1433429 4 35899687 G A 0.082 1.04 (0.39-2.82) 0.9325 +
rs8336 4 95211610 A G 0.400 1.02 (0.58-1.78) 0.9507 + rs2247218 6 101966553 A G 0.700 1.01 (0.55-1.84) 0.9736 +
rs10510985 3 69663871 G A 0.564 1.00 (0.57-1.74) 0.9978 +
Fatigue cohort (Bonferroni adjusted p-value = 0.0005)
rs7306948 12 123345347 A G 0.957 1.17 (1.06-1.29) 0.0024 -
rs6721414 2 18494495 G C 0.589 1.06 (1.02-1.11) 0.0034 -
rs10121299 9 79397301 A G 0.935 1.12 (1.03-1.22) 0.0072 - rs1157185 2 38285735 T C 0.380 1.05 (1.00-1.09) 0.0405 +
rs1367696 2 38286914 A G 0.380 1.04 (1.00-1.09) 0.0598 +
rs4242391 8 23000183 T C 0.385 1.04 (1.00-1.08) 0.0659 - rs10789931 11 112842773 C T 0.882 1.06 (1.00-1.13) 0.0680 -
rs4245562 7 54403985 T C 0.699 1.04 (1.00-1.09) 0.0739 -
rs985257 2 38283228 A T 0.377 1.04 (0.99-1.08) 0.0862 + rs2016483 4 95229039 T A 0.565 1.04 (0.99-1.08) 0.0887 -
rs9320409 6 97530846 C T 0.463 1.03 (0.99-1.07) 0.1036 -
rs8336 4 95211610 C T 0.584 1.03 (0.99-1.08) 0.1163 - rs10501376 11 58971766 C G 0.900 1.05 (0.99-1.12) 0.1294 +
rs7537461 1 113383662 A C 0.886 1.05 (0.98-1.12) 0.1524 -
rs1325904 10 90280938 T C 0.270 1.03 (0.99-1.08) 0.1796 + rs4253760 22 46622384 G T 0.174 1.04 (0.98-1.10) 0.1877 +
rs1610024 9 111614766 A G 0.203 1.03 (0.98-1.09) 0.2025 + rs9200 5 41142606 C T 0.504 1.03 (0.98-1.07) 0.2231 +
rs400322 19 55172578 G A 0.716 1.03 (0.98-1.08) 0.2423 +
rs2602803 2 30818644 G T 0.100 1.04 (0.97-1.11) 0.2582 + rs10511961 9 71497485 G C 0.565 1.02 (0.98-1.07) 0.2617 -
rs17255510 14 22662856 C T 0.212 1.03 (0.98-1.08) 0.2632 +
rs4251545 12 44180295 A G 0.094 1.04 (0.97-1.11) 0.2778 - rs7834482 8 127785745 A G 0.787 1.03 (0.98-1.08) 0.2821 -
rs1881470 11 127333840 G C 0.740 1.02 (0.98-1.07) 0.3033 +
rs2247215 6 101966454 A G 0.701 1.02 (0.98-1.07) 0.3136 + rs10498445 14 52740441 G C 0.282 1.02 (0.98-1.07) 0.3185 -
rs11214105 11 112037653 A G 0.280 1.02 (0.98-1.07) 0.3270 +
rs10506025 12 27726370 G A 0.629 1.02 (0.98-1.06) 0.3341 + rs12055682 6 81189123 G A 0.138 1.03 (0.97-1.09) 0.3380 +
rs7529589 1 196658279 T C 0.398 1.02 (0.98-1.06) 0.3561 -
rs734640 11 17613348 A G 0.644 1.02 (0.98-1.06) 0.3580 -
rs2062758 12 39045452 T A 0.885 1.03 (0.97-1.10) 0.3759 +
rs7994531 13 42977439 T C 0.246 1.02 (0.97-1.07) 0.3762 -
rs1061147 1 196654324 A C 0.397 1.02 (0.98-1.06) 0.3897 - rs2228428 3 32995928 C T 0.701 1.02 (0.98-1.07) 0.4025 -
rs2247218 6 101966553 T C 0.702 1.02 (0.97-1.07) 0.4145 +
rs1926721 1 230864830 A G 0.968 1.04 (0.93-1.17) 0.4514 + rs3095168 17 16260198 C T 0.979 1.05 (0.92-1.21) 0.4518 -
rs4894505 3 175920884 T G 0.667 1.02 (0.97-1.06) 0.4552 -
rs2196007 3 144120914 C T 0.275 1.02 (0.97-1.07) 0.4563 - rs12408925 1 48473192 A G 0.757 1.02 (0.97-1.07) 0.4818 -
rs7307225 12 71898358 T C 0.888 1.02 (0.96-1.09) 0.4852 -
rs723886 7 68159592 G T 0.651 1.02 (0.97-1.06) 0.4880 + rs4151667 6 31914024 A T 0.039 1.04 (0.93-1.15) 0.4948 +
rs167337 12 52182052 G A 0.865 1.02 (0.96-1.08) 0.5065 -
rs11090847 22 48899419 C T 0.900 1.02 (0.96-1.10) 0.5080 - rs6926583 6 136752092 C G 0.841 1.02 (0.96-1.08) 0.5099 +
rs33013 5 80060016 A G 0.332 1.01 (0.97-1.06) 0.5233 -
rs997139 6 136751118 A G 0.841 1.02 (0.96-1.07) 0.5341 - rs10507556 13 47970075 G A 0.875 1.02 (0.96-1.08) 0.5392 -
Footnotes for Supplementary Table 7.4 are on page 222.
Appendices 221
Supplementary Table 7.4. Continued Summary of association results for SNPs from previous
chronic fatigue syndrome (CFS) genome-wide association analyses within our CFS and fatigue
cohorts.
SNP Chr SNP position RA OA Freq OR (95% CI) p-value Effect
directiona
rs7616342 3 19433647 A G 0.587 1.01 (0.97-1.06) 0.5411 -
rs12305678 12 2763539 A G 0.937 1.03 (0.94-1.11) 0.5504 -
rs7726463 5 37955073 T C 0.907 1.02 (0.95-1.09) 0.5508 - rs2490495 10 32726529 C G 0.487 1.01 (0.97-1.06) 0.5589 -
rs1763788 10 32917853 T C 0.488 1.01 (0.97-1.06) 0.5832 +
rs2014012 5 58612388 A T 0.655 1.01 (0.97-1.06) 0.5836 + rs10517378 4 36532369 C A 0.078 1.02 (0.95-1.10) 0.5878 +
rs382958 19 56439438 A G 0.545 1.01 (0.97-1.05) 0.5976 +
rs12417706 11 116059440 T C 0.050 1.03 (0.93-1.13) 0.6091 + rs11154872 6 136797757 A G 0.846 1.01 (0.96-1.07) 0.6114 -
rs11009106 10 33123413 C G 0.487 1.01 (0.97-1.05) 0.6209 +
rs2995467 10 33135952 G A 0.487 1.01 (0.97-1.05) 0.6209 + rs12407818 1 52904410 C T 0.056 1.02 (0.93-1.12) 0.6330 +
rs1762529 10 32968080 T C 0.488 1.01 (0.97-1.05) 0.6355 +
rs2784574 10 32976689 A G 0.488 1.01 (0.97-1.05) 0.6355 - rs1577372 10 32938382 C T 0.488 1.01 (0.97-1.05) 0.6362 -
rs11711551 3 66697961 A G 0.951 1.02 (0.93-1.13) 0.6368 -
rs10800118 1 165599774 G C 0.543 1.01 (0.97-1.05) 0.6444 - rs10501068 11 26769636 T G 0.529 1.01 (0.97-1.05) 0.6708 -
rs6923953 6 136726688 T C 0.844 1.01 (0.96-1.07) 0.6748 -
rs2389957 4 120695322 C T 0.634 1.01 (0.97-1.05) 0.6755 + rs1377828 3 176245042 T G 0.772 1.01 (0.96-1.06) 0.6759 -
rs3797302 5 145889123 G C 0.906 1.02 (0.95-1.09) 0.6774 +
rs2816936 1 199982900 G A 0.100 1.01 (0.95-1.09) 0.6902 - rs1157819 1 209604734 T A 0.355 1.01 (0.97-1.05) 0.6955 -
rs2882361 4 161379616 C G 0.353 1.01 (0.97-1.05) 0.7117 -
rs7747443 6 136713749 A G 0.844 1.01 (0.96-1.07) 0.7240 - rs13010656 2 203297068 G T 0.480 1.01 (0.97-1.05) 0.7241 -
rs372402 5 148752020 T C 0.511 1.01 (0.97-1.05) 0.7249 +
rs12761944 10 32803484 T G 0.487 1.01 (0.97-1.05) 0.7306 + rs353254 5 148748736 A G 0.512 1.01 (0.97-1.05) 0.7415 +
rs6449669 5 62929018 T A 0.157 1.01 (0.95-1.07) 0.7521 + rs2715898 2 201556388 C T 0.488 1.01 (0.97-1.05) 0.7645 -
rs9946817 18 70367007 G A 0.198 1.01 (0.95-1.06) 0.7786 +
rs197770 3 37515827 G A 0.142 1.01 (0.95-1.07) 0.7834 + rs726817 10 95459817 T C 0.728 1.01 (0.96-1.05) 0.7932 +
rs6074914 20 15519613 G A 0.035 1.01 (0.91-1.13) 0.8017 -
rs3778315 6 136850687 A G 0.847 1.01 (0.95-1.06) 0.8176 - rs1801058 4 3039150 T C 0.417 1.00 (0.96-1.05) 0.8196 -
rs10817082 9 113474166 C G 0.613 1.00 (0.96-1.05) 0.8207 +
rs4892034 18 70399988 A T 0.790 1.00 (0.95-1.06) 0.8647 + rs543736 8 104012949 G A 0.635 1.00 (0.95-1.04) 0.8878 -
rs10489599 1 16585818 A G 0.608 1.00 (0.96-1.05) 0.8973 -
rs1048829 2 203430456 G T 0.521 1.00 (0.96-1.04) 0.9080 - rs1050998 17 4638737 A G 0.585 1.00 (0.96-1.05) 0.9101 +
rs549908 11 112020916 T G 0.689 1.00 (0.95-1.04) 0.9141 -
rs496731 18 26491368 C A 0.749 1.00 (0.95-1.05) 0.9152 -
rs228945 22 37525880 T C 0.704 1.00 (0.96-1.05) 0.9156 +
rs11984468 8 16367555 G C 0.955 1.01 (0.91-1.11) 0.9170 -
rs228941 22 37523721 C G 0.711 1.00 (0.95-1.04) 0.9227 + rs2277680 17 4638563 G A 0.585 1.00 (0.96-1.05) 0.9282 -
rs9283919 6 54114066 A G 0.952 1.00 (0.91-1.10) 0.9301 -
rs654807 2 16457284 A G 0.555 1.00 (0.96-1.04) 0.9496 - rs4978076 9 26524684 T C 0.500 1.00 (0.96-1.04) 0.9559 -
rs2193766 3 8829321 C T 0.977 1.00 (0.88-1.15) 0.9596 +
rs10402951 19 7555092 G A 0.293 1.00 (0.96-1.05) 0.9845 + rs8050875 16 11223537 C T 0.676 1.00 (0.96-1.04) 0.9950 +
rs283825 2 79232491 A G 0.861 1.00 (0.94-1.06) 0.9964 -
rs2200706 3 3673608 C T 0.880 1.00 (0.94-1.07) 0.9983 -
Chr: chromosome; RA: risk allele; OA: other allele; OR: odds ratio; CI: confidence interval. aSNPs with
effects in the same direction are indicated by + while SNPs with effects in opposite directions are
indicated by -.
222 Appendices
Supplementary Table 7.5. Summary of association results for genes from previous chronic fatigue
syndrome (CFS) genome-wide association analyses within our CFS and fatigue cohorts.
Gene Chromosome Start Stop Number of SNPs p-value
CFS cohort (Bonferroni-adjusted p-value = 0.0002)
PLA2G4A 1 186798032 186958113 37 0.0001
MOB3B 9 27325207 27530223 78 0.0003
AFAP1 4 7760440 7941653 66 0.0075
RASEF 9 85594500 85678043 13 0.0108
CELF4 18 34823003 35146000 107 0.0112
ITGA9 3 37493813 37861281 98 0.0119
ARPC1B 7 98972298 99004226 5 0.0132
TXNRD1 12 104609537 104744085 23 0.0183
NR6A1 9 127279554 127533589 8 0.0228
PTGS2 1 186640944 186649559 3 0.0235
FSHR 2 49189296 49381666 68 0.0244
CDH18 5 19473140 20575972 144 0.0264
SLC13A5 17 6588038 6616886 18 0.0268
PDE4D 5 58264865 59783925 294 0.0285
DNMBP 10 101635334 101769676 21 0.0287
FBXL13 7 102453308 102715288 36 0.0383
SLC35F3 1 234040679 234460262 134 0.0396
FAM185A 7 102389399 102449672 3 0.0452
CLEC16A 16 11038345 11276046 63 0.0477
PARP11 12 3907410 3982608 16 0.0552
POLR3A 10 79734907 79789298 7 0.0556
KCNAB1 3 155838337 156256927 75 0.0657
DPF3 14 73085561 73360824 91 0.0809
LRFN2 6 40359373 40555126 64 0.0810
PRDM16 1 2985565 3355185 113 0.0879
CTSH 15 79214092 79237436 5 0.0926
NLRP13 19 56403058 56443702 21 0.0957
HLA-DOA 6 32971959 32977389 28 0.1005
MYO18B 22 26138117 26453345 138 0.1030
BPTF 17 65821644 65980494 15 0.1065
TRIP11 14 92434243 92506484 13 0.1140
ANO10 3 43407818 43663560 31 0.1231
ZCCHC11 1 52888947 53018764 12 0.1238
TBC1D19 4 26585546 26758232 5 0.1283
SH3TC2 5 148361713 148442737 20 0.1300
ZNF24 18 32912178 32924428 5 0.1347
NCAM1 11 112831969 113149158 91 0.1413
TRB 7 141998851 142510972 83 0.1424
GDA 9 74729511 74867140 35 0.1464
MED11 17 4634723 4636902 1 0.1478
C3orf52 3 111805175 111837073 7 0.1481
CDKN1A 6 36644237 36655116 3 0.1606
FBXO42 1 16573339 16678965 14 0.1607
TCERG1 5 145826873 145891071 8 0.1675
CCL27 9 34661893 34662689 1 0.1698
KLHL32 6 97372496 97588630 49 0.1711
IL20RB 3 136676707 136729927 5 0.1712
TNFRSF10D 8 22993101 23021543 10 0.1736
SHANK2 11 70313961 70935808 131 0.1825
FBLN5 14 92335755 92414046 35 0.1835
ZNF407 18 72265106 72777628 78 0.1929
CFB 6 31913721 31919861 20 0.2048
C22orf34 22 50013290 50051190 5 0.2111
RECK 9 36036910 36124452 17 0.2162
ABCG1 21 43619799 43724497 79 0.2205
CCR4 3 32993066 32996403 1 0.2412
Appendices 223
Supplementary Table 7.5. Continued Summary of association results for genes from previous
chronic fatigue syndrome (CFS) genome-wide association analyses within our CFS and fatigue
cohorts.
Gene Chromosome Start Stop Number of SNPs p-value
MASP1 3 186933873 187009810 37 0.2416
MSH3 5 79950467 80172634 43 0.2461
MRPL52 14 23299088 23304246 8 0.2562
VAPA 18 9913955 9960018 18 0.2587
SPSB4 3 140770743 140867453 17 0.2673
INTS2 17 59942728 60005377 3 0.2716
UTRN 6 144606434 145174170 106 0.2794
SOX5 12 23682438 24715383 353 0.2836
NDUFAF2 5 60240956 60448864 17 0.2845
APLP2 11 129939716 130014706 14 0.2875
CFH 1 196621008 196716634 37 0.2954
UVRAG 11 75526212 75855282 29 0.2982
CYP24A1 20 52769985 52790516 23 0.3045
PDZRN3 3 73431582 73674072 79 0.3059
RMDN2 2 38152462 38294285 41 0.3141
GALNT18 11 11292421 11643561 244 0.3208
NUP210L 1 153965166 154127592 14 0.3257
NXPH1 7 8473585 8792593 103 0.3259
POU6F2 7 39017609 39504390 102 0.3426
FADS6 17 72873451 72889737 4 0.3498
DNM3 1 171810618 172387606 112 0.3519
GRIK3 1 37261128 37499844 42 0.3543
BMPR2 2 203241033 203432474 17 0.3584
IRAK4 12 44152747 44183346 11 0.3641
C5orf66 5 134368970 134680370 88 0.3665
CACNA1C 12 2079952 2807115 202 0.3666
ANK2 4 113739239 114304896 157 0.3751
AGPAT3 21 45285116 45407475 26 0.3817
MARCH1 4 164445450 165305093 169 0.3867
TACC2 10 123748689 124014057 124 0.3921
TMX4 20 7961713 8000393 16 0.3955
RAP1GAP 1 21922708 21996010 23 0.4007
KCNIP4 4 20730234 21950424 303 0.4019
SIGLEC5 19 52114756 52133727 13 0.4023
PRKCH 14 61788515 62017698 93 0.4047
SAP130 2 128698791 128785667 11 0.4078
PPARA 22 46546458 46639653 48 0.4085
IL18 11 112013974 112034840 5 0.4099
PRR12 19 50094912 50129696 2 0.4239
OXTR 3 8792094 8811300 18 0.4248
PDE4DIP 1 144851424 145076186 2 0.4298
DSE 6 116601231 116762422 24 0.4320
HK1 10 71029740 71161638 52 0.4381
EPHA6 3 96533425 97467786 89 0.4399
PRKCE 2 45878454 46415129 237 0.4429
C6 5 41142248 41261588 28 0.4432
HLA-DQA2 6 32709156 32714664 14 0.4438
CLYBL 13 100258919 100549388 102 0.4484
CNTN5 11 98891706 100229616 486 0.4511
FAM19A5 22 48885288 49147744 127 0.4520
KCNJ6 21 38996778 39288741 101 0.4543
SERPINA5 14 95047706 95059457 19 0.4671
PRUNE2 9 79226292 79521136 125 0.4678
SHISA6 17 11144740 11467380 120 0.4742
MTAP 9 21802635 21941040 32 0.4765
224 Appendices
Supplementary Table 7.5. Continued Summary of association results for genes from previous
chronic fatigue syndrome (CFS) genome-wide association analyses within our CFS and fatigue
cohorts.
Gene Chromosome Start Stop Number of SNPs p-value
GRIA1 5 152870084 153193429 83 0.4795
DNAH3 16 20944476 21170762 46 0.4799
HDAC11 3 13521671 13547924 10 0.4812
STK10 5 171469074 171615346 47 0.4818
C3orf67 3 58720199 59035804 21 0.4819
PTPRD 9 8314246 10612723 1009 0.4833
IL17B 5 148753830 148760848 3 0.4864
ARL15 5 53180578 53606403 125 0.4892
CSMD3 8 113235157 114449242 135 0.4895
PIP5K1B 9 71320188 71624092 76 0.4898
TARP 7 38299243 38357285 18 0.4965
TRD 14 22891537 22935569 16 0.5023
DTX4 11 58939812 58976060 9 0.5027
PSD3 8 18384813 18871196 262 0.5043
CSRNP3 2 166326157 166545917 51 0.5046
PPFIBP1 12 27677045 27848497 63 0.5058
ARMC9 2 232063294 232238606 33 0.5064
PEX11G 19 7541756 7555884 5 0.5097
RNASEL 1 182542769 182558420 8 0.5111
CACNA2D1 7 81575760 82073031 131 0.5115
TEX12 11 112004926 112043279 12 0.5210
CCDC7 10 32735057 33171805 35 0.5267
ZBTB20 3 114033348 114866132 139 0.5327
PTPRU 1 29563028 29653325 16 0.5337
NLRP11 19 56296763 56348128 35 0.5374
HIP1R 12 123319045 123347508 5 0.5403
RBM19 12 114254543 114404176 54 0.5430
TRA 14 22090057 23021075 363 0.5465
INSR 19 7112266 7294011 60 0.5500
GRIN2B 12 13713684 14133022 157 0.5526
LRGUK 7 133812105 133948933 18 0.5634
STXBP5L 3 120627050 121143608 33 0.5637
LHX4 1 180199433 180244188 27 0.5661
CD247 1 167399877 167487847 36 0.5669
LRP1B 2 140988996 142889270 511 0.5678
GON4L 1 155719450 155829185 7 0.5743
SYNE3 14 95883831 95983000 74 0.5844
SHPRH 6 146205943 146285421 8 0.5854
DNAH8 6 38683117 38998574 102 0.5859
SLCO3A1 15 92396938 92715665 177 0.5862
SPTAN1 9 131314837 131395944 6 0.5892
MFAP5 12 8798540 8815433 6 0.5960
FRA10AC1 10 95427640 95462329 15 0.5990
KCNH8 3 19189946 19577135 56 0.6014
FSTL5 4 162305044 163085186 158 0.6085
ZC3H13 13 46528600 46626896 10 0.6119
GBE1 3 81538850 81810950 34 0.6150
KNTC1 12 123011796 123110947 15 0.6158
TIRAP 11 126152800 126164828 11 0.6212
TECTB 10 114043213 114064793 9 0.6303
HS6ST3 13 96743093 97491816 132 0.6338
COLEC11 2 3642422 3692234 17 0.6346
LMX1B 9 129376722 129463311 23 0.6357
ARHGAP20 11 110447759 110583912 20 0.6366
CAMK1D 10 12391583 12871735 218 0.6378
Appendices 225
Supplementary Table 7.5. Continued Summary of association results for genes from previous
chronic fatigue syndrome (CFS) genome-wide association analyses within our CFS and fatigue
cohorts.
Gene Chromosome Start Stop Number of SNPs p-value
LIMD1 3 45636323 45722755 19 0.6454
IL2RB 22 37521878 37552414 15 0.6516
IQUB 7 123091993 123174718 12 0.6527
UBE3C 7 156931655 157062066 9 0.6565
CADM2 3 85008133 86123579 123 0.6578
CLEC4M 19 7828035 7834491 5 0.6586
ANKFN1 17 54230836 54560007 67 0.6605
FBXL21 5 135266006 135277367 10 0.6610
NPAS2 2 101436613 101613289 75 0.6631
GRK4 4 2965232 3042474 14 0.6639
MAP7 6 136663419 136871957 16 0.6663
PCYOX1L 5 148737570 148749221 3 0.6704
TMPRSS15 21 19641433 19775970 70 0.6716
TPD52 8 80947103 81083836 28 0.6762
MAGI3 1 113933087 114228545 30 0.6796
BMP6 6 7727011 7881972 45 0.6810
ACOXL 2 111490150 111875799 107 0.6869
LGR5 12 71832931 71980090 35 0.6949
PTGDR 14 52734310 52743808 11 0.6972
WBSCR17 7 70597523 71178586 153 0.6975
SLC1A3 5 36606457 36688436 44 0.6980
THTPA 14 23980969 24028790 17 0.7003
STK32B 4 5053527 5502728 131 0.7015
MUSK 9 113430935 113566386 44 0.7035
ZNF100 19 21906417 21950430 4 0.7046
NKAIN2 6 124124991 125146786 279 0.7075
UBE2G1 17 4172512 4269969 11 0.7200
MGST3 1 165600110 165625373 31 0.7241
CD8A 2 87011728 87035519 6 0.7249
CCDC157 22 30752627 30772818 6 0.7271
CXCL16 17 4636828 4643223 7 0.7284
NKAIN3 8 63161501 63912211 152 0.7286
SHFM1 7 96318079 96339203 2 0.7335
PRKG1 10 52750911 54058110 391 0.7477
PEX6 6 42931611 42946981 5 0.7486
LMAN1L 15 75105194 75118099 6 0.7498
MLIP 6 53883714 54131078 67 0.7555
SIX6 14 60975938 60978525 3 0.7588
LCA5 6 80194708 80247147 11 0.7598
ARHGEF3 3 56761446 57113336 107 0.7612
SCN8A 12 51984050 52206648 35 0.7691
PALM2-AKAP2 9 112542577 112934792 155 0.7718
UBXN6 19 4445003 4457791 5 0.7777
TNFRSF1B 1 12227044 12269279 20 0.7875
EFCAB11 14 90261499 90421148 36 0.7884
PPP3CC 8 22298483 22398657 15 0.7887
ZNF469 16 88493879 88507165 8 0.7918
CFTR 7 117120017 117308719 27 0.7934
FAM19A1 3 68040734 68594772 148 0.7973
ZBED4 22 50247497 50283726 7 0.8006
MAG 19 35782989 35804710 7 0.8086
PGM2 4 37828282 37864559 16 0.8134
SLC2A9 4 9827848 10041872 75 0.8169
CELF2 10 10838851 11378674 216 0.8179
GPR161 1 168048780 168106905 12 0.8230
226 Appendices
Supplementary Table 7.5. Continued Summary of association results for genes from previous
chronic fatigue syndrome (CFS) genome-wide association analyses within our CFS and fatigue
cohorts.
Gene Chromosome Start Stop Number of SNPs p-value
EFNA5 5 106712590 107006596 99 0.8252
SRRM3 7 75831211 75916609 6 0.8454
GRIK2 6 101841584 102517958 151 0.8477
PRKACB 1 84543745 84704181 24 0.8485
SAMD4A 14 55033815 55260033 107 0.8490
IGL 22 22380474 23265085 91 0.8583
LCLAT1 2 30670102 30867091 36 0.8618
RAB3C 5 57877950 58155221 83 0.8631
DDAH1 1 85784168 86044046 78 0.8638
LILRB4 19 55174271 55181810 7 0.8690
ICOSLG 21 45642874 45660881 4 0.8692
ALG12 22 50296852 50312106 3 0.8712
DCC 18 49866542 51062273 265 0.8790
CDC20B 5 54408799 54469005 19 0.8974
TNFRSF19 13 24144509 24250244 33 0.8976
FAM49A 2 16730727 16847134 42 0.8978
TOX3 16 52471682 52581714 18 0.8991
SPOCK1 5 136310987 136835018 140 0.9030
MACROD2 20 13976146 16033842 666 0.9076
CRYL1 13 20977806 21100012 39 0.9139
OTOG 11 17568920 17667491 31 0.9183
STAB2 12 103981069 104160502 79 0.9185
PTPRT 20 40701392 41818557 379 0.9193
RBBP6 16 24550866 24584184 7 0.9447
RASGRF2 5 80256491 80525981 120 0.9456
GALNT7 4 174089904 174245118 18 0.9481
TYK2 19 10461204 10491248 12 0.9501
CEP128 14 80962821 81408105 73 0.9511
DNTTIP1 20 44420576 44440066 2 0.9589
SLC7A8 14 23594504 23652869 32 0.9682
SMARCAD1 4 95128759 95212443 15 0.9777
Fatigue cohort (Bonferroni-adjusted p-value = 0.0002)
PRUNE2 9 79226292 79521136 383 0.0003
RNASEL 1 182542769 182558420 6 0.0045
COLEC11 2 3642422 3692234 55 0.0080
TNFRSF10D 8 22993101 23021543 4 0.0275
HS6ST3 13 96743093 97491816 182 0.0281
ANK2 4 113739239 114304896 551 0.0458
ARL15 5 53180578 53606403 539 0.0476
MGST3 1 165600110 165625373 54 0.0503
LRGUK 7 133812105 133948933 239 0.0637
HIP1R 12 123319045 123347508 37 0.0711
ZNF407 18 72265106 72777628 198 0.0770
POU6F2 7 39017609 39504390 386 0.0799
CD247 1 167399877 167487847 47 0.0889
DTX4 11 58939812 58976060 23 0.0908
BMPR2 2 203241033 203432474 70 0.0995
PTPRU 1 29563028 29653325 61 0.1068
HLA-DOA 6 32971959 32977389 16 0.1118
PPARA 22 46546458 46639653 55 0.1227
LHX4 1 180199433 180244188 19 0.1503
ZC3H13 13 46528600 46626896 43 0.1540
CSRNP3 2 166326157 166545917 145 0.1640
SYNE3 14 95883831 95983000 78 0.1736
CFH 1 196621008 196716634 149 0.1807
Appendices 227
Supplementary Table 7.5. Continued Summary of association results for genes from previous
chronic fatigue syndrome (CFS) genome-wide association analyses within our CFS and fatigue
cohorts.
Gene Chromosome Start Stop Number of SNPs p-value
HDAC11 3 13521671 13547924 1 0.1815
FSHR 2 49189296 49381666 375 0.1829
NKAIN2 6 124124991 125146786 664 0.1858
KNTC1 12 123011796 123110947 45 0.1886
GPR161 1 168048780 168106905 65 0.2007
NDUFAF2 5 60240956 60448864 270 0.2024
MAGI3 1 113933087 114228545 140 0.2058
ICOSLG 21 45642874 45660881 11 0.2066
SH3TC2 5 148361713 148442737 124 0.2220
PRR12 19 50094912 50129696 22 0.2230
PRDM16 1 2985565 3355185 147 0.2232
ALG12 22 50296852 50312106 16 0.2324
SLC1A3 5 36606457 36688436 86 0.2386
SOX5 12 23682438 24715383 827 0.2395
MRPL52 14 23299088 23304246 5 0.2396
C3orf67 3 58720199 59035804 38 0.2401
STAB2 12 103981069 104160502 171 0.2465
RMDN2 2 38152462 38294285 235 0.2474
NR6A1 9 127279554 127533589 50 0.2690
SHANK2 11 70313961 70935808 366 0.2708
LCA5 6 80194708 80247147 36 0.2729
TARP 7 38299243 38357285 10 0.2742
WBSCR17 7 70597523 71178586 703 0.2794
MFAP5 12 8798540 8815433 1 0.2858
ACOXL 2 111490150 111875799 380 0.3006
CEP128 14 80962821 81408105 672 0.3039
LCLAT1 2 30670102 30867091 172 0.3072
LMAN1L 15 75105194 75118099 9 0.3190
TOX3 16 52471682 52581714 58 0.3194
RBM19 12 114254543 114404176 176 0.3202
CFTR 7 117120017 117308719 104 0.3369
PPFIBP1 12 27677045 27848497 251 0.3409
SMARCAD1 4 95128759 95212443 89 0.3425
PTGS2 1 186640944 186649559 2 0.3426
CELF2 10 10838851 11378674 532 0.3472
CACNA1C 12 2079952 2807115 375 0.3474
GRIK2 6 101841584 102517958 511 0.3505
SCN8A 12 51984050 52206648 44 0.3524
PTPRT 20 40701392 41818557 916 0.3536
ZNF24 18 32912178 32924428 9 0.3541
CNTN5 11 98891706 100229616 2003 0.3609
DNTTIP1 20 44420576 44440066 13 0.3613
GON4L 1 155719450 155829185 30 0.3674
PIP5K1B 9 71320188 71624092 161 0.3679
GRIA1 5 152870084 153193429 451 0.3709
CLYBL 13 100258919 100549388 247 0.3744
TMPRSS15 21 19641433 19775970 132 0.3768
IRAK4 12 44152747 44183346 38 0.3922
CCR4 3 32993066 32996403 1 0.4025
NLRP13 19 56403058 56443702 39 0.4095
FBLN5 14 92335755 92414046 69 0.4197
OXTR 3 8792094 8811300 7 0.4240
RECK 9 36036910 36124452 60 0.4309
TXNRD1 12 104609537 104744085 189 0.4312
MARCH1 4 164445450 165305093 688 0.4318
228 Appendices
Supplementary Table 7.5. Continued Summary of association results for genes from previous
chronic fatigue syndrome (CFS) genome-wide association analyses within our CFS and fatigue
cohorts.
Gene Chromosome Start Stop Number of SNPs p-value
GDA 9 74729511 74867140 125 0.4370
RAB3C 5 57877950 58155221 382 0.4396
NLRP11 19 56296763 56348128 10 0.4460
TYK2 19 10461204 10491248 9 0.4462
FAM185A 7 102389399 102449672 11 0.4474
AGPAT3 21 45285116 45407475 88 0.4505
NCAM1 11 112831969 113149158 519 0.4520
CDH18 5 19473140 20575972 781 0.4542
TACC2 10 123748689 124014057 269 0.4564
CFB 6 31913721 31919861 10 0.4592
EFNA5 5 106712590 107006596 289 0.4633
EFCAB11 14 90261499 90421148 189 0.4644
SIX6 14 60975938 60978525 2 0.4668
CYP24A1 20 52769985 52790516 12 0.4709
PSD3 8 18384813 18871196 845 0.4788
CSMD3 8 113235157 114449242 1093 0.4812
STK10 5 171469074 171615346 135 0.4838
HLA-DQA2 6 32709156 32714664 39 0.4874
CELF4 18 34823003 35146000 282 0.4880
CDC20B 5 54408799 54469005 3 0.4880
IGL 22 22380474 23265085 93 0.4884
VAPA 18 9913955 9960018 52 0.4906
ZBED4 22 50247497 50283726 94 0.4917
SRRM3 7 75831211 75916609 77 0.4988
DNMBP 10 101635334 101769676 228 0.5000
SHPRH 6 146205943 146285421 71 0.5072
UTRN 6 144606434 145174170 227 0.5098
RASGRF2 5 80256491 80525981 307 0.5148
FBXL21 5 135266006 135277367 8 0.5174
CCDC7 10 32735057 33171805 600 0.5177
C5orf66 5 134368970 134680370 163 0.5195
TMX4 20 7961713 8000393 9 0.5217
PALM2-AKAP2 9 112542577 112934792 338 0.5228
SERPINA5 14 95047706 95059457 20 0.5249
FAM19A5 22 48885288 49147744 367 0.5252
TRA 14 22090057 23021075 1109 0.5254
ARHGEF3 3 56761446 57113336 284 0.5388
CCDC157 22 30752627 30772818 11 0.5413
DNM3 1 171810618 172387606 544 0.5421
KLHL32 6 97372496 97588630 235 0.5487
PRKCE 2 45878454 46415129 433 0.5492
C6 5 41142248 41261588 123 0.5518
TEX12 11 112004926 112043279 27 0.5523
MAP7 6 136663419 136871957 97 0.5547
SLC35F3 1 234040679 234460262 249 0.5570
PDE4D 5 58264865 59783925 956 0.5702
KCNJ6 21 38996778 39288741 266 0.5721
FBXL13 7 102453308 102715288 101 0.5757
MOB3B 9 27325207 27530223 196 0.5852
CTSH 15 79214092 79237436 28 0.5854
DNAH8 6 38683117 38998574 312 0.5856
THTPA 14 23980969 24028790 1 0.5936
MTAP 9 21802635 21941040 105 0.5994
TNFRSF1B 1 12227044 12269279 32 0.6015
TECTB 10 114043213 114064793 1 0.6025
Appendices 229
Supplementary Table 7.5. Continued Summary of association results for genes from previous
chronic fatigue syndrome (CFS) genome-wide association analyses within our CFS and fatigue
cohorts.
Gene Chromosome Start Stop Number of SNPs p-value
ANKFN1 17 54230836 54560007 146 0.6044
ANO10 3 43407818 43663560 12 0.6049
STK32B 4 5053527 5502728 246 0.6109
PCYOX1L 5 148737570 148749221 9 0.6142
NXPH1 7 8473585 8792593 324 0.6161
PRKCH 14 61788515 62017698 230 0.6207
CRYL1 13 20977806 21100012 164 0.6231
DPF3 14 73085561 73360824 276 0.6293
LGR5 12 71832931 71980090 238 0.6325
SPSB4 3 140770743 140867453 21 0.6364
IL18 11 112013974 112034840 18 0.6391
ZCCHC11 1 52888947 53018764 50 0.6402
ABCG1 21 43619799 43724497 65 0.6471
SPOCK1 5 136310987 136835018 603 0.6486
PTGDR 14 52734310 52743808 27 0.6501
PEX6 6 42931611 42946981 21 0.6546
GALNT7 4 174089904 174245118 32 0.6546
EPHA6 3 96533425 97467786 254 0.6547
FRA10AC1 10 95427640 95462329 56 0.6590
LRP1B 2 140988996 142889270 3044 0.6599
IQUB 7 123091993 123174718 67 0.6635
MYO18B 22 26138117 26453345 235 0.6704
FBXO42 1 16573339 16678965 128 0.6770
IL17B 5 148753830 148760848 9 0.6776
PRKG1 10 52750911 54058110 1607 0.6811
HK1 10 71029740 71161638 40 0.6892
PARP11 12 3907410 3982608 67 0.6915
SAMD4A 14 55033815 55260033 54 0.6926
BMP6 6 7727011 7881972 108 0.6970
RBBP6 16 24550866 24584184 40 0.6995
MAG 19 35782989 35804710 3 0.6995
UVRAG 11 75526212 75855282 71 0.7069
GBE1 3 81538850 81810950 153 0.7109
LIMD1 3 45636323 45722755 151 0.7258
KCNAB1 3 155838337 156256927 383 0.7336
KCNIP4 4 20730234 21950424 1557 0.7373
TRIP11 14 92434243 92506484 83 0.7443
CADM2 3 85008133 86123579 1366 0.7522
INTS2 17 59942728 60005377 2 0.7577
ARMC9 2 232063294 232238606 106 0.7592
CACNA2D1 7 81575760 82073031 530 0.7623
LMX1B 9 129376722 129463311 68 0.7623
AFAP1 4 7760440 7941653 352 0.7624
FAM49A 2 16730727 16847134 39 0.7721
PTPRD 9 8314246 10612723 3071 0.7734
UBE3C 7 156931655 157062066 169 0.7742
C22orf34 22 50013290 50051190 23 0.7761
KCNH8 3 19189946 19577135 121 0.7774
ITGA9 3 37493813 37861281 296 0.7823
TRB 7 141998851 142510972 255 0.7827
TCERG1 5 145826873 145891071 94 0.7885
SLCO3A1 15 92396938 92715665 259 0.7906
STXBP5L 3 120627050 121143608 73 0.7906
PLA2G4A 1 186798032 186958113 125 0.7972
MSH3 5 79950467 80172634 380 0.7979
230 Appendices
Supplementary Table 7.5. Continued Summary of association results for genes from previous
chronic fatigue syndrome (CFS) genome-wide association analyses within our CFS and fatigue
cohorts.
Gene Chromosome Start Stop Number of SNPs p-value
DCC 18 49866542 51062273 1918 0.8018
RASEF 9 85594500 85678043 55 0.8062
TRD 14 22891537 22935569 25 0.8071
GRK4 4 2965232 3042474 105 0.8166
GRIK3 1 37261128 37499844 29 0.8195
LRFN2 6 40359373 40555126 125 0.8290
CXCL16 17 4636828 4643223 12 0.8295
MUSK 9 113430935 113566386 133 0.8303
FSTL5 4 162305044 163085186 1194 0.8357
RAP1GAP 1 21922708 21996010 53 0.8407
CAMK1D 10 12391583 12871735 439 0.8452
POLR3A 10 79734907 79789298 34 0.8499
IL2RB 22 37521878 37552414 34 0.8516
INSR 19 7112266 7294011 153 0.8516
GALNT18 11 11292421 11643561 462 0.8585
CLEC4M 19 7828035 7834491 3 0.8595
TNFRSF19 13 24144509 24250244 140 0.8630
UBE2G1 17 4172512 4269969 38 0.8635
APLP2 11 129939716 130014706 40 0.8637
OTOG 11 17568920 17667491 27 0.8650
SIGLEC5 19 52114756 52133727 1 0.8675
SHISA6 17 11144740 11467380 268 0.8701
NPAS2 2 101436613 101613289 177 0.8707
CDKN1A 6 36644237 36655116 4 0.8758
PRKACB 1 84543745 84704181 85 0.8802
CD8A 2 87011728 87035519 1 0.8856
C3orf52 3 111805175 111837073 9 0.8866
CLEC16A 16 11038345 11276046 363 0.8878
TPD52 8 80947103 81083836 191 0.8895
PGM2 4 37828282 37864559 115 0.8948
SLC2A9 4 9827848 10041872 8 0.8950
MLIP 6 53883714 54131078 287 0.8968
TBC1D19 4 26585546 26758232 16 0.8973
NUP210L 1 153965166 154127592 7 0.9030
DDAH1 1 85784168 86044046 271 0.9140
PEX11G 19 7541756 7555884 13 0.9160
SLC7A8 14 23594504 23652869 37 0.9169
NKAIN3 8 63161501 63912211 817 0.9238
SAP130 2 128698791 128785667 22 0.9253
MASP1 3 186933873 187009810 79 0.9265
ZBTB20 3 114033348 114866132 384 0.9268
TIRAP 11 126152800 126164828 16 0.9279
PDZRN3 3 73431582 73674072 161 0.9305
SHFM1 7 96318079 96339203 14 0.9310
MACROD2 20 13976146 16033842 1999 0.9343
DNAH3 16 20944476 21170762 282 0.9522
FAM19A1 3 68040734 68594772 640 0.9560
DSE 6 116601231 116762422 114 0.9573
GRIN2B 12 13713684 14133022 589 0.9681
Appendices 231
Supplementary Table 7.6. Summary of association results for genes previously associated with self-
reported tiredness within our CFS and fatigue cohorts.
Gene Chromosome Start Stop Number of SNPs p-value
CFS cohort (Bonferroni-adjusted p-value = 0.0010)
DAG1 3 49506136 49573051 3 0.0269
ZBTB37 1 173837220 173856596 1 0.0412
NAPA 19 47990891 48018515 4 0.0539
ANO10 3 43407818 43663560 31 0.1231
RHCG 15 90014638 90039799 7 0.1243
THEM4 1 151843342 151882361 8 0.1276
NICN1 3 49459766 49466777 1 0.1425
RHOA 3 49396569 49449526 3 0.1504
PRRC2C 1 171454652 171562650 27 0.1906
METTL16 17 2319343 2415200 17 0.2055
BSN 3 49591922 49708982 11 0.2058
TCTA 3 49449639 49453909 1 0.2423
SMC1B 22 45739944 45809500 15 0.2451
SSBP4 19 18529679 18545372 2 0.2494
ISYNA1 19 18545198 18549111 1 0.2546
ZDHHC5 11 57435223 57468659 4 0.2745
ELL 19 18553473 18632937 7 0.2985
CTNND1 11 57520756 57586652 6 0.3080
FBXO21 12 117581278 117628305 7 0.3420
SNF8 17 47007458 47022484 3 0.3483
OPA1 3 193310933 193415600 15 0.3969
CATSPER2 15 43922772 43941039 3 0.4218
PRR12 19 50094912 50129696 2 0.4239
SLC44A5 1 75667816 76081698 67 0.4335
ATP11B 3 182511291 182639423 10 0.4371
NRXN1 2 50145643 51259674 340 0.4537
C3orf84 3 49215069 49229291 3 0.4574
TMX2 11 57479995 57508445 4 0.4730
RPE 2 210867289 210889784 3 0.4761
CSMD3 8 113235157 114449242 135 0.4895
KANSL1L 2 210885435 211036068 5 0.4995
PAFAH1B1 17 2496923 2588909 15 0.5192
CCDC36 3 49235861 49295636 9 0.5327
UBA7 3 49842638 49851391 2 0.5365
GIP 17 47035918 47045955 4 0.5693
RELT 11 73087405 73108519 2 0.5835
CCNT2 2 135676363 135716915 8 0.5871
ZNF780A 19 40575059 40596845 3 0.5888
SERPING1 11 57365027 57382326 6 0.5940
CAMK1D 10 12391583 12871735 218 0.6378
UBE2Z 17 46985731 47006422 8 0.6420
PLGRKT 9 5357966 5438381 24 0.6994
ADARB1 21 46494493 46646478 29 0.7425
ASXL3 18 31158541 31331159 43 0.8406
DRD2 11 113280317 113346413 18 0.8721
KLF7 2 207938861 208031970 32 0.8908
PLAC8 4 84011201 84035911 5 0.9139
FAM168A 11 73111837 73309234 13 0.9197
PSMC4 19 40476912 40487671 1 0.9199
SRRM4 12 119419300 119600856 62 0.9491
Fatigue cohort (Bonferroni-adjusted p-value = 0.0012)
PLGRKT 9 5357966 5438381 144 0.0027
KANSL1L 2 210885435 211036068 49 0.0033
RPE 2 210867289 210889784 14 0.0054
ZBTB37 1 173837220 173856596 25 0.0457
ASXL3 18 31158541 31331159 156 0.0959
232 Appendices
Supplementary Table 7.6. Continued Summary of association results for genes previously
associated with self-reported tiredness within our CFS and fatigue cohorts.
Gene Chromosome Start Stop Number of SNPs p-value
PRR12 19 50094912 50129696 22 0.2230
SNF8 17 47007458 47022484 28 0.2347
C3orf84 3 49215069 49229291 9 0.2385
KLF7 2 207938861 208031970 103 0.2677
UBE2Z 17 46985731 47006422 42 0.2966
SRRM4 12 119419300 119600856 198 0.3356
UBA7 3 49842638 49851391 7 0.3428
GIP 17 47035918 47045955 22 0.3995
THEM4 1 151843342 151882361 17 0.4092
CCDC36 3 49235861 49295636 29 0.4103
FBXO21 12 117581278 117628305 90 0.4361
FAM168A 11 73111837 73309234 84 0.4584
CSMD3 8 113235157 114449242 1093 0.4812
OPA1 3 193310933 193415600 169 0.5291
PRRC2C 1 171454652 171562650 31 0.5343
ELL 19 18553473 18632937 92 0.5729
ANO10 3 43407818 43663560 12 0.6049
SERPING1 11 57365027 57382326 30 0.6280
METTL16 17 2319343 2415200 51 0.6553
BSN 3 49591922 49708982 106 0.6689
TCTA 3 49449639 49453909 5 0.6888
RELT 11 73087405 73108519 6 0.7115
SLC44A5 1 75667816 76081698 459 0.7196
ZNF780A 19 40575059 40596845 7 0.7258
RHOA 3 49396569 49449526 65 0.7358
NICN1 3 49459766 49466777 5 0.7361
ZDHHC5 11 57435223 57468659 21 0.7462
SMC1B 22 45739944 45809500 86 0.7479
TMX2 11 57479995 57508445 17 0.7536
DAG1 3 49506136 49573051 69 0.7706
CTNND1 11 57520756 57586652 32 0.7715
NRXN1 2 50145643 51259674 1765 0.7827
PAFAH1B1 17 2496923 2588909 15 0.7985
CAMK1D 10 12391583 12871735 439 0.8452
ADARB1 21 46494493 46646478 128 0.8869
DRD2 11 113280317 113346413 86 0.9021
Appendices 233
Supplementary Table 7.7. Summary of association results for SNPs previously associated with
major depressive disorder within our CFS and fatigue cohorts.
SNP Chr SNP position RA OA Freq OR (95% CI) p-value Effect
directiona
CFS cohort (Bonferroni-adjusted p-value = 0.0045)
rs10514299 5 87663610 C T 0.709 2.16 (1.09-4.30) 0.0264 +
rs7044150 9 2982931 T C 0.346 1.53 (0.87-2.69) 0.1394 +
rs7973260 12 118375486 G A 0.827 1.58 (0.71-3.51) 0.2628 -
rs12065553 1 80793118 A G 0.700 1.40 (0.75-2.63) 0.2900 +
rs10786831 10 106614571 G A 0.546 1.23 (0.70-2.14) 0.4697 +
rs301806 1 8482078 T C 0.519 1.20 (0.69-2.09) 0.5193 +
rs2125716 12 84941429 C T 0.791 1.20 (0.60-2.41) 0.6126 +
rs9825823 3 61082153 T C 0.436 1.14 (0.65-1.98) 0.6499 +
rs12552 13 53625781 T C 0.473 1.07 (0.62-1.86) 0.8126 -
rs6476606 9 37005561 G A 0.627 1.05 (0.59-1.86) 0.8707 +
rs7647854 3 184876783 G A 0.173 1.06 (0.51-2.18) 0.8794 +
Fatigue cohort (Bonferroni-adjusted p-value = 0.0042)
rs8025231 15 37648402 A C 0.558 1.03 (0.99-1.07) 0.1403 +
rs9825823 3 61082153 C T 0.564 1.03 (0.98-1.07) 0.2388 -
rs7647854 3 184876783 A G 0.846 1.03 (0.97-1.09) 0.2848 -
rs1656369 3 158280085 T A 0.637 1.02 (0.98-1.07) 0.3868 -
rs1475120 6 105389953 G A 0.448 1.02 (0.97-1.06) 0.4738 +
rs4543289 5 164484948 G T 0.533 1.02 (0.97-1.06) 0.4831 -
rs10786831 10 106614571 G A 0.593 1.01 (0.97-1.06) 0.5416 +
rs2422321 1 73293393 A G 0.586 1.00 (0.95-1.04) 0.8411 +
rs1518395 2 58208074 G A 0.61 1.00 (0.95-1.04) 0.8506 -
rs12065553 1 80793118 A G 0.713 1.00 (0.96-1.05) 0.8948 +
rs12552 13 53625781 A G 0.433 1.00 (0.96-1.05) 0.9016 -
rs6476606 9 37005561 A G 0.378 1.00 (0.96-1.04) 0.9427 -
Chr: chromosome; RA: risk allele; OA: other allele; Freq: frequency; OR: odds ratio; CI: confidence interval. aSNPs with effects in the same direction are indicated by + while SNPs with effects in opposite
directions are indicated by -.
Supplementary Table 7.8. Summary of association results for genes previously associated with
major depressive disorder within our CFS and fatigue cohorts.
Gene Chromosome Start Stop Number of
SNPs p-value
CFS cohort (Bonferroni-adjusted p-value = 0.0063)
KSR2 12 117890817 118406399 230 0.1991
L3MBTL2 22 41601312 41627276 6 0.2429
RERE 1 8412464 8877699 47 0.3513
SORCS3 10 106400859 107024993 122 0.3691
DCC 18 49866542 51062273 265 0.8790
PAX5 9 36833272 37034476 82 0.9057
FHIT 3 59735036 61237133 650 0.9307
OLFM4 13 53602876 53626196 23 0.9458
Fatigue cohort (Bonferroni-adjusted p-value = 0.0063)
FHIT 3 59735036 61237133 1839 0.2139
OLFM4 13 53602876 53626196 16 0.3112
L3MBTL2 22 41601312 41627276 61 0.5274
RERE 1 8412464 8877699 89 0.5350
SORCS3 10 106400859 107024993 1280 0.5695
PAX5 9 36833272 37034476 247 0.5696
DCC 18 49866542 51062273 1918 0.8018
KSR2 12 117890817 118406399 463 0.8058
234 Appendices
Supplementary Figure 7.1. Regional association plots of the corresponding to the region 50kb upstream and downstream of the risk loci associated with self-reported
tiredness in the UK Biobank dataset (Deary et al., 2017). For each plot the −log10 P values (y-axis) of the SNPs are shown according to their chromosomal positions (x-axis).
The estimated recombination rates from the 1000 Genomes Project March 2012 release are shown as blue lines, and the genomic locations of genes within the regions of
interest in the NCBI Build 37 human assembly are shown as arrows. SNP colour represents LD with the most highly associated SNP at each locus. The figures were created
with Locus Zoom (Pruim et al., 2010). Figures A and B show the region surrounding 1:64178756_C_T in the CFS and fatigue cohorts, respectively. Figures C and D show
the region surrounding rs142592148, on chromosome 1, in the CFS and fatigue cohorts, respectively. Figures E and F show the region surrounding rs7219015, on
chromosome 17, in the CFS and fatigue cohorts, respectively.
A B C
D E F
Appendices 235
Supplementary Figure 7.2. Quantile-Quantile plot of observed vs. expected p-values for the genome-
wide association analysis of chronic fatigue syndrome (λ = 0.99).
236 Appendices
Supplementary Figure 7.3. Regional association plots corresponding to the 400kb region surrounding A. rs12473577, B. rs652252, and C. rs1888140 in the CFS cohort. For
each plot the −log10 P values (y-axis) of the SNPs are shown according to their chromosomal positions (x-axis). The estimated recombination rates from the 1000 Genomes
Project March 2012 release are shown as blue lines, and the genomic locations of genes within the regions of interest in the NCBI Build 37 human assembly are shown as
arrows. SNP colour represents LD with the most highly associated SNP at each locus. The figures were created with Locus Zoom (Pruim et al., 2010).
A B C
Appendices 237
Supplementary Figure 7.4. Quantile-Quantile plot of observed vs. expected p-values for the genome-
wide association analysis of fatigue (λ = 1.02).
238 Appendices
Supplementary Figure 7.5. Regional association plots corresponding to the 400kb region surrounding A. rs874681, B. rs16849948, C. rs1701470, D. rs352582, E. rs359477,
and F. rs4237354 in the fatigue cohort. For each plot the −log10 P values (y-axis) of the SNPs are shown according to their chromosomal positions (x-axis). The estimated
recombination rates from the 1000 Genomes Project March 2012 release are shown as blue lines, and the genomic locations of genes within the regions of interest in the
NCBI Build 37 human assembly are shown as arrows. SNP colour represents LD with the most highly associated SNP at each locus. The figures were created with Locus
Zoom (Pruim et al., 2010).
A B C
D E F