BMJ Open is an open access journal and the full, final ... · EJT; Visualisation AH, EJT, AK;...

BMJ Open is committed to open peer review. As part of this commitment we make the peer review history of every article we publish publicly available. When an article is published we post the peer reviewers’ comments and the authors’ responses online. We also post the versions of the paper that were used during peer review. These are the versions that the peer review comments apply to. The versions of the paper that follow are the versions that were submitted during the peer review process. They are not the versions of record or the final published versions. They should not be cited or distributed as the published version of this manuscript. BMJ Open is an open access journal and the full, final, typeset and author-corrected version of record of the manuscript is available on our site with no access controls, subscription charges or pay-per-view fees (http://bmjopen.bmj.com). If you have any questions on BMJ Open’s open peer review process please email

[email protected]

http://bmjopen.bmj.com/

[email protected]

For peer review onlySocioeconomic inequalities in obesity: modelling future

trends in Australia

Journal: BMJ Open

Manuscript ID bmjopen-2018-026525

Article Type: Research

Date Submitted by the Author: 06-Sep-2018

Complete List of Authors: Hayes, Alison; University of Sydney - Camperdown and Darlington Campus, Sydney School of Public HealthTan, Eng Joo; University of Sydney, Sydney School of Public HealthKilledar, Anagha; University of Sydney, Sydney School of Public HealthLung, Thomas; University of New South Wales, The George Institute for Global Health

Keywords: Obesity, microsimulation, socioeconomic inequalities, BMI trajectory, modelling

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open

For peer review only

1

Socioeconomic inequalities in obesity: modelling future trends in Australia

Alison Hayes, Eng Joo Tan, Anagha Killedar, Thomas Lung

Alison Hayes, Associate Professor, The University of Sydney, Faculty of Medicine and Health, School

of Public Health, NSW 2006, Australia

Eng Joo Tan, Health Economics Research Fellow, The University of Sydney, Faculty of Medicine and

Health, School of Public Health, NSW 2006, Australia

Anagha Killedar, PhD candidate, The University of Sydney, Faculty of Medicine and Health, School of

Public Health, NSW 2006, Australia

Thomas Lung, Health Economics Research Fellow, The George Institute for Global Health, University

of New South Wales, NSW 2042

Correspondence to: Eng Joo Tan

Rm314, Edward Ford Building

Sydney School of Public Health, University of Sydney

NSW 2006 Australia

[email protected]

Contributions: The author’s responsibilities were as follows: AH conceived the study. Model

conceptualization AH, TL; Software TL, AH; Analysed the data AH, TL, EJT; performed experiments TL,

EJT; Visualisation AH, EJT, AK; Writing First draft: AH, AK. All authors revised the manuscript for

important intellectual content. AH, EJT and TL had full access to the data and take responsibility for

the integrity of the data analysis. AH is the guarantor. All authors have given final approval of the

version to be published.

Page 1 of 41


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


2

Funding: Dr Tan receives funding support from the National Health and Medical Research Council

Centre of Research Excellence in Early Prevention of Obesity in Childhood (APP1101675). Ms Killedar

is supported by the Kassulke Scholarship for PhD study. Dr. Lung is supported by a National Health

and Medical Research Council Early Career Fellowship and a Heart Foundation Postdoctoral

Fellowship (APP 1141392).

Competing interests: None declared

Patient consent: Information in the Australian National Health Surveys have been collected under

the Census and Statistics Act 1905 (CSA) by the Australian Bureau of Statistics.

Data sharing statement: The model code is available on request. Data on which analyses are based

are available from the Australian Bureau of Statistics.

Word Count: 2975

Page 2 of 41


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


3

Abstract

Objectives: To develop a model to predict future socioeconomic inequalities in body-mass index

(BMI) and obesity.

Design: Microsimulation modelling using BMI data from adult participants of Australian Health

Surveys, and published data on the relative risk of mortality in relation to BMI and socioeconomic

position (SEP), based on education.

Setting: Australia.

Participants: 74,329 adults, aged 20 and over from Australian Health Surveys, 1995-2015.

Primary and secondary outcome measures: The primary outcomes were BMI trajectories and

obesity prevalence by SEP for 4 birth cohorts, born 10 years apart, centred on 1940, 1950, 1960 and

1970.

Results: Simulations predicted persistent or widening socioeconomic inequality in BMI and obesity

over the adult life course, for all birth cohorts. Recent birth cohorts will experience greater

socioeconomic inequality by middle age, compared with earlier cohorts. For example among men,

there was no inequality in obesity prevalence at age 60 for the 1940 birth cohort (low SEP 25% [95%

CI 17 to 34]; high SEP 26% [95% CI 19 to 34]), yet for the 1970 birth cohort, obesity prevalence was

predicted to be 51% [95% CI 43 to 58] and 41% [95% CI 36 to 46] for the low and high SEP groups,

respectively. Notably, for more recent birth cohorts, the model predicted greatest socioeconomic

inequality in severe obesity (BMI>35kg/m2) by age 60.

Conclusions: Lower SEP groups and more recent birth cohorts are at higher risk of obesity and

severe obesity, and its consequences in middle age. Prevention efforts should focus on these

vulnerable population groups in order to avoid future disparities in health outcomes. The model

provides a framework to investigate which interventions will be most effective in narrowing the gap

in socioeconomic disparities in obesity in adulthood. Further research will facilitate modelled health

economic evaluations among different socioeconomic groups.


Page 3 of 41


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


4

Article Summary

Strengths and limitations of this study

• This is an innovative study and the first to use micro-simulation to increase our

understanding of trends in socioeconomic disparities in obesity among adults in Australia.

• The model combines best evidence pertaining to obesity progression and mortality in

different socio-economic groups and includes Australian data, and published meta-analyses

of mortality in relation to weight status and SEP

• The model has been presented transparently and externally validated using the most

recently available national data on population level adult BMI from Australia.

• Translation of evidence— the model has the potential to inform policy on which

interventions are most effective in narrowing the gap in socioeconomic disparities in

overweight and obesity in adulthood.

• A limitation is the use of only one indicator of SEP based on educational attainment, which

had some missing data in the baseline population.

Page 4 of 41


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


5

INTRODUCTION

Obesity has been described as the public health challenge of our time.1 In the last 4 decades, high

income Western countries including Australia have seen unprecedented increases in age-

standardised adult BMI and the prevalence of obesity.2 More recently, severe obesity (BMI >35) has

emerged as a public health problem, and in Australia the prevalence has doubled in the last 20 years

(from 5 to 10% of the adult population).3 This has important implications, because the upper

extremes of the BMI spectrum confer acute health risks and because healthcare costs rise steeply

with BMI above 35.4

It is well established that in high income countries, obesity disproportionately affects the most

socioeconomically disadvantaged groups.5 Furthermore, there are major disparities in chronic

disease outcomes for which overweight and obesity is a risk factor.6 However, there is limited

evidence of how these inequalities are changing over time and epidemiologic and health economic

models of obesity rarely take account of socioeconomic position (SEP), thus overlooking a key policy-

relevant determinant of obesity.

People generally accrue BMI during their life course7 and, as a result, there has been great

interest in identifying BMI trajectories through longitudinal studies or modelling BMI growth

trajectories8

to understand the epidemiology of disease and to identify at–risk populations. Beyond

their value in epidemiological studies, predictive models are regarded as powerful tools for

informing policy decisions.9 They may facilitate health economic evaluations that take account of the

predicted long-term effects and resource considerations of prevention and treatment strategies.

This is particularly appropriate for the assessment of obesity policies where the resources are spent

in the short term but the weight loss or weight gain avoided, and therefore cost savings, will often

emerge in the long-term.

Page 5 of 41


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


6

A number of recent systematic reviews into the effectiveness of obesity prevention and treatment

strategies to reduce inequalities provide some evidence that targeted interventions are effective in

socially disadvantaged groups.10,11

However, each of these reviews highlights the lack of evidence

available to assess the impact of interventions on inequalities, particularly in the long-term.

Simulation modelling is a powerful tool for predicting the epidemiology of obesity progression12

but

currently there are few analytical tools to evaluate which interventions are most effective in

reducing inequalities.13

We have developed a micro-simulation model for obesity progression among

Australian adults14

that accurately simulates population level changes in BMI and obesity in Australia

over the last 25 years but does not currently incorporate any measure of socioeconomic position. In

this study we present a new version of the model in which all equations and parameters have been

updated to include a measure of socioeconomic position, based on educational attainment. We

validate model predictions against observed BMI trajectories and obesity progression, stratified by

SEP, and use the model to determine future trends in socioeconomic inequalities in obesity – both

within and between generations.

METHODS

Study populations

Study populations included survey respondents of four Australian National Health Surveys between

1995 and 2015, including the 1995 National Nutrition Survey (NNS), and National Health Surveys

(NHS) in 2007/8, 2011/12 and 2014/15.3,15-17

Height and weight were objectively measured in all

surveys. Data from the 1995 and 2011/12 surveys were used to derive equations for age-related

weight gain, and all NHS beyond 1995 were used in validation of model predictions until 2014/15.

Overview of the simulation model

Page 6 of 41


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


7

Our approach uses individual-level (micro-simulation) modelling and predicts BMI trajectories for

members of the Australian adult population. Micro-simulation accounts for heterogeneity within a

population and thus can model obesity progression based on individual characteristics such as age,

sex and socioeconomic position (SEP). The model is initialized with nationally representative

individual-level data from participants of national health surveys, that have characteristics of age,

sex, SEP and measured BMI. The model runs on discrete time-steps, in which individuals grow older

by one year, they may gain or lose weight and/or they may die in any annual cycle. BMI is modelled

as a continuous variable, age is modelled in individual years and SEP is a binary variable defined by

completion of high school. The model equations predicting annual BMI change (Supplement

Methods 1.1) are based on Australian national data, using a synthetic cohort technique.18

The

modelling of age-, sex- and SEP- specific mortality (Supplement Methods 1.2) is based on the

2011/12 Australian life table,19

a published meta-analysis of the association of BMI and all-cause

mortality,20

and the published relative risk of mortality by SEP from a large Australian cohort study.21

This involves apportioning the conditional probability of death to those of different weight status

and SEP, using established methods.22

Validation and prediction of BMI and obesity trajectories by SEP

We carried out validation of the model predictions, for four birth cohorts 1966-75, 1956-65, 1946-55

and 1936-45 (centred around 1970, 1960, 1950 and 1940 and aged 20-29, 30-39, 40-49 and 50-59

years in 1995). Starting with a baseline population representing 4.5 million adult men and 4.9 million

adult women in 1995, we simulated BMI trajectories and determined the prevalence of healthy,

overweight, obesity, and severe obesity over three decades to 2025, stratified by SEP. Predicted

mean BMI and prevalence of weight status groups, using standard BMI cut points, were compared

with NHS data from the 2007/8, 2011/12 and 2014/15 surveys, matched to the same birth cohorts.

As only the 1995 and 2011/12 data were used in the derivation of model equations, this represents

both internal and external validation.23

Page 7 of 41


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


8

SEP inequalities in BMI and obesity by birth cohort

We simulated BMI trajectories from 1995 through to 2035 to predict how the population BMI

distribution and obesity prevalence progresses over time among different birth cohorts and different

SEP groups. In order to compare outcomes of different birth cohorts at a common age, we ran

simulations over the adult life course. We chose 60 years as a suitable age to compare outcomes, as

this is the age at which obesity related chronic disease starts to become apparent.24

For the two

most recent birth cohorts, this required running simulations prospectively i.e. beyond 2015. We then

determined inequality, calculated as absolute differences between high and low SEP in mean BMI,

prevalence of obesity (BMI>30 kg/m2) and prevalence of severe obesity (BMI>35 kg/m

2) at age 60

years.

Sensitivity analysis

Sensitivity analysis seeks to identify sensitive model parameters, i.e. those which are most important

in driving model outputs.25

We changed major model parameters by their upper and lower 95%

confidence limits and observed the change in the projected prevalence of mean BMI, overall obesity

and severe obesity by age 60 years, when compared with the base model. These sensitivity analyses

were carried out for men and women of high and low SEP, for 4 different age and birth cohorts,

centred around: 1940, 1950, 1960 and 1970.

Parameters investigated in the one–way sensitivity analyses were:

a. changing constants in the weight gain equations for men and women, by upper and lower

95% confidence limits

b. changing the hazard ratio for mortality (1.39 (95% CI 1.08 to 1.79) of low compared to high

education groups by the upper and lower 95% confidence limits.

Further details are provided in (Supplement Methods 1.3).

Page 8 of 41


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


9

Patient and public involvement

Patients and public were not involved in the design of the research study. This study is a modelling

study that used non-identifiable participant data from National Health Surveys and collected under

the Census Act.

RESULTS

Validation and projection of BMI and obesity trajectories by SEP

Figure 1 shows simulated and observed BMI trajectories between 1995 until 2025 for four birth

cohorts of men and women and two SEP groups. Overall, simulated BMI trajectories predicted a

widening or persisting socioeconomic inequality in mean BMI over time. For all cohorts, the model

showed good internal and external validation as NHS data were within the simulated 95% interval.

Similarly, inequalities in the prevalence of obesity were also projected to widen over time, and this

widening appeared to be greatest for the 1950 and 1960 birth cohorts (at least until 2025) which

was corroborated by NHS data for each birth cohort (Figure 2). Validation graphs of more weight

status groups by SEP are shown in Supplement Figures 1 and 2.


Figure 3 shows an example of the simulated progression of BMI distribution over time, for high and

low SEP, starting with a base population of 20-29 year old men. The baseline BMI distribution of the

low SEP group was already flatter and more right skewed than the high SEP group in 1995.

Simulated data show that by 2015 the distributions have advanced and the right skew increased but

this is more pronounced for low SEP. By 2035 the right skew is projected to increase further,

resulting in greater proportion of the distribution above BMI>35kg/m2 for low compared to high SEP.

Page 9 of 41


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


10

The model predicted that recent birth cohorts will experience unprecedented levels of obesity and

severe obesity by the time they reach middle age (Figure 4) – and the lower SEP group will be worst

affected. Obesity at age 60 (represented by total bar height in Figure 4) is predicted to be higher for

each successive birth cohort. For the 1970 birth cohort, the model predicts that 50% of the low SEP

group and around 40% of the high SEP will have obesity at age 60 years, approximately double that

of the 1940 birth cohort of around 24%, irrespective of SEP. Substantial socioeconomic inequalities

in mean BMI and prevalence of obesity at age 60 years were predicted for the three most recent

cohorts studied (Table 1); a difference of 1- 2 units of BMI, and 10-15% obesity prevalence between

low and high SEP. In contrast, there was virtually no inequality in any of the outcomes at age 60 for

the 1940 birth cohort. Whilst socioeconomic inequalities in BMI and obesity (BMI>30) at age 60 were

predicted to widen for the two successive birth cohorts beyond 1940, there was some attenuation of

these inequalities for the 1970 cohort (Table 1).

Notably, for the two most recent birth cohorts investigated (1960 and 1970) socioeconomic

inequality at age 60 years was predicted to be greatest in severe obesity (BMI>35kg/m2), with a

difference of about 10 percentage points between high and low SEP groups (Figure 4). In contrast,

there was negligible inequality in obesity (30kg/m2>BMI<35kg/m

2) between high and low SEP groups

of the same birth cohorts. In other words, most of the predicted socioeconomic inequality in BMI

above 30kg/m2 is due to inequality at the extreme upper bound (BMI>35kg/m

2).


The results of the sensitivity analyses are shown in Supplement Methods 1.3. Changing annual

weight gain by upper and lower CIs had major impacts on predicted BMI, obesity and severe obesity

at age 60, but only minimal effects on inequalities. For example, the prevalence of obesity at age 60

for the 1970 cohort changed by approximately +25% or -19% under the alternate weight gain

scenarios, yet the inequalities increased only slightly, by 3-5%. Furthermore, changing the hazard of

Page 10 of 41


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


11

mortality by SEP to upper and lower 95% CI had little or no effect on projected mean BMI, obesity

and severe obesity at age 60 years, and no effect on absolute inequalities. None of the sensitivity

analyses investigated affected the predicted pattern of obesity being higher with successive

generations and the finding that the 3 most recent cohorts would have greater socioeconomic

disparities at age 60, when compared with the 1940 birth cohort.

Page 11 of 41


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


12

Table 1. Simulated outcomes at age 60 for different birth cohorts of men and women, and absolute difference (inequality) in

outcomes between lower and higher SEP groups (High minus Low)

Mean BMI at age 60

(kg/m2)

Obesity

(BMI >30 kg/m2)

prevalence

(%)

Severe obesity

(BMI>35 kg/m2) prevalence

(%)

Low SEP High SEP Abs Diff Low SEP High SEP Abs Diff Low SEP High SEP Abs Diff

MEN

Birth Cohort 1940 27.6 27.7 0.1 24.5 25.6 1.1 1.4 2.6 1.2

1950 28.9 27.6 -1.3 36.3 23.1 -13.2 8.3 7.3 -1.0

1960 30.7 28.5 -2.2 47.9 33.6 -14.3 20.3 9.4 -10.9

1970 31.4 29.7 -1.7 50.7 41.1 -9.6 24.6 13.9 -10.7

WOMEN

Birth Cohort 1940 27.5 27.5 0.0 26.5 24.0 -2.5 9.4 10.4 1.0

1950 28.7 27.0 -1.7 38.3 22.1 -16.2 13.9 9.2 -4.7

1960 30.4 28.0 -2.4 43.4 30.4 -13.0 22.7 11.0 -11.7

1970 31.7 29.7 -2.0 53.7 42.3 -11.4 25.8 18.1 -7.7

Abbreviation: Abs Diff, absolute difference

Page 12 of 41


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


13

DISCUSSION

Our study provides insight into the future inequalities in obesity and severe obesity in a high-income

country. Persistent or widening inequalities were predicted between 1995 and 2025 for all birth

cohorts studied. Moreover, the model predicted that recent birth cohorts will experience

unprecedented levels of obesity and severe obesity by the time they reach middle age, and greater

socioeconomic inequality, compared with earlier birth cohorts. Of great concern is the predicted

shift towards inequality in severe obesity, and thus, the associated unequal burden of obesity

related disease.

The major strength of the study is our novel modelling methods which account for age – related BMI

change across the life course and age-related mortality within SEP groups. BMI is modelled as a

continuous variable, thus allowing for the prediction of prevalence of a range of weight status

groups, including severe obesity, which has not previously been possible with existing models.26

Another strength is the validation of model predictions using the most recently available national

data on adult BMI from Australia. This provides confidence in the model’s predictions into the

future. We have adhered to good reporting practices for modelling23

and the modelling is

comprehensive and transparent. Finally, the model is informed by objectively measured height and

weight, based on nationally representative population data.

As with any modelling study, there are a number of assumptions. The first is that age- and SEP-

related annual weight gain derived from a contemporary time period, up to 2012, is assumed to hold

beyond 2012. This may be a reasonable assumption, as recent studies suggest age-related annual

weight gain has been stable, or even slowed.18,27

Nonetheless, in sensitivity analysis we have

investigated the scenarios of annual weight gain being higher or lower, and the major conclusions

pertaining to the projected widening inequality in obesity and severe obesity prevalence still hold.

Page 13 of 41


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


14

Another limitation of the study is the use of completion of high school education as the only

indicator of SEP. As high school education is generally completed by early adulthood, it is a suitable

indicator to use in an adult life-course framework.28

There is evidence of education being an

important predictor of weight gain29,30

and the use of an individual level characteristic, is also

consistent with micro-simulation. However, the relevance of education as a marker of SEP, may

differ between birth cohorts because of secular trends in education levels. Interestingly, a recent

meta-analysis31

using occupational status as a marker of SEP, reported a very similar hazard ratio, for

mortality of low compared to high SEP, to that used in our modelling, based on educational status.

Investigations of inequalities in obesity progression using other measures of socioeconomic position

will be an important avenue for future research. Finally, there were some missing data on education

status, in our baseline population in 1995, particularly for the oldest birth cohort, which could lead

to bias. However, our results show good internal and external validation, suggesting that any bias did

not have major impact on the overall results.

To our knowledge, this is the first study in which a micro-simulation model has been developed,

validated and used to increase our understanding of trends in socioeconomic disparities in obesity

among adults in Australia. This study adds to the debate of whether inequalities in obesity are

growing. Our finding of widening inequalities in obesity corroborates with existing studies in the

UK,32

Australia26

and Europe,33

whilst other developed (OECD) countries report stable inequalities,34

and a US study found that socioeconomic inequality in obesity had largely disappeared by 2012.35

The majority of these studies analysed data from cross-sectional surveys. In contrast, our dynamic

model which accounts for the association of weight gain and mortality with SEP, has allowed us to

model into the future and hence to compare, side by side, 4 different birth cohorts of different SEP,

born 10 years apart.

Page 14 of 41


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


15

Socioeconomic disparities in obesity prevalence predicted by our model arise directly from the

higher rate of weight gain among low compared with high SEP groups. Studies in other countries

have also found disparities in weight gain by educational or occupational class.36,37

The mediators of

inequalities in weight gain are not clear, but there is some evidence that diet quality is poorer for

low SEP groups in Australia,38

and poorer health behaviours (smoking and physical activity)39

may

also play a role. However, the presence of inequalities in obesity at the beginning of adulthood

(Figure 1) suggests inequalities in weight gain during childhood also contribute to inequalities in

adulthood. The prediction that recent generations will have unprecedented levels of obesity and

severe obesity by middle age is probably due to exposure to obesogenic environments, for a greater

proportion of their lifetime, including wider access to low-nutrient, high-fat food and lower levels of

physical activity.

Notably, the model predicted that socioeconomic inequalities in obesity will be greater than in

previous generations, and that severe obesity, which has the greatest health implications and

medical expenditures,4 will disproportionately affect those in lower SEP groups. This study fills an

important gap in our understanding of how inequalities in obesity develop over time and has policy

implications for targeting of prevention efforts. Lower SEP groups and more recent birth cohorts are

at higher risk of obesity, severe obesity, and its consequences in middle age. Prevention efforts

should focus on these vulnerable population groups in order to avoid increasing disparities in the

long-term burden of obesity in the future.

Beyond its use in predicting epidemiology of obesity within different social strata, this model is part

of a wider research effort to develop a health economic model that has relevance for different SEP

groups. As the epidemiological predictions of the model are sound, we can have high confidence in

its health economic predictions. By modelling at the individual level, microsimulation will allow for

the investigation of intervention effects targeted at specific population groups (e.g. lower educated

Page 15 of 41


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


16

young men who are overweight). Simulation modelling has, to date, been underutilised in evaluation

of the impact of interventions on inequalities in health.13

We hope future research using this model

will assist policy makers in identifying not only which interventions are most effective and cost-

effective but will also determine which are most effective in narrowing the gap in socioeconomic

disparities in overweight and obesity in adulthood.

Acknowledgement: We thank the Australian Bureau of Statistics for provision of confidentialised

unit record data pertaining to National Health Surveys.

Page 16 of 41


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


17

References

1. Bassett MT, Perl S. Obesity: the public health challenge of our time. Am J Public Health

2004;94(9):1477.

2. NCD Risk Factor Collaboration. Trends in adult body-mass index in 200 countries from

1975 to 2014: a pooled analysis of 1698 population-based measurement studies with

19.2 million participants. Lancet 2016;387:1377–1396.

3. Australian Bureau of Statistics. National Health Survey: First Results, 2014-15. Canberra:

Australian Bureau Statistics; 2015.

4. Kent S, Fusco F, Gray A, et al. Body mass index and healthcare costs: a systematic

literature review of individual participant data studies. Obes Rev. 2017;18:869–879.

5. McLaren L. Socioeconomic status and obesity. Epidemiol Rev 2007;29:29–48.

6. Korda RJ, Soga K, Joshy G, et al. Socioeconomic variation in incidence of primary and

secondary major cardiovascular disease events: an Australian population-based

prospective cohort study. Int J Equity Health 2016;15:1–10.

7. Jacobsen BK, Njolstad I, Thune I, et al. Increase in weight in all birth cohorts in a general

population: the Tromso Study, 1974-1994. Arch Intern. Med 2001;161:466–472.

8. Ward ZJ, Long MW, Resch SC, et al. Simulation of growth trajectories of childhood obesity

into adulthood. N Engl J Med. 2017;377:2145–2153.

9. Richardson MB, Williams MS, Fontaine KR, et al. The development of scientific evidence

for health policies for obesity: Why and how? Int J Obes. 2017;41:840–848.

10. Cairns JM, Bambra C, Hillier-Brown FC, et al. Weighing up the evidence: a systematic

review of the effectiveness of workplace interventions to tackle socio-economic

inequalities in obesity. J Public Health 2015;37:659–670.

11. Hillier-Brown FC, Bambra CL, Cairns JM, et al. A systematic review of the effectiveness of

individual, community and societal-level interventions at reducing socio-economic

inequalities in obesity among adults. Int J Obes 2014;38:1483–1490.

Page 17 of 41


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


18

12. Levy DT, Mabry PL, Wang YC, et al. Simulation models of obesity: a review of the

literature and implications for research and policy. Obes Rev 2011;12(5):378–394.

13. Smith BT, Smith PM, Harper S, et al. Reducing social inequalities in health: the role of

simulation modelling in chronic disease epidemiology to evaluate the impact of

population health interventions. J Epidemiol Community Health 2014;68(4):384–389.

14. Hayes AJ, Lung TWC, Bauman A, et al. Modelling obesity trends in Australia: unravelling

the past and predicting the future. Int J Obes 2017;41:178-185.

15. Australian Bureau of Statistics. Information Paper: National Nutrition Survey 1995 (Cat.

No. 4805.0). Canberra: Australian Bureau of Statistics; 1995.

16. Australian Bureau of Statistics. National Health Survey: Users' Guide - Electronic

Publication, 2007-08 (Cat. No. 4363.0.55.001). Canberra: Australian Bureau of Statistics;

2008.

17. Australian Bureau of Statistics. Australian Health Survey: First Results, 2011–12.

Canberra: Australian Bureau of Statistics; 2012.

18. Hayes A, Gearon E, Backholer et al. Age-specific changes in BMI and BMI distribution

among Australian adults using cross-sectional surveys from 1980 to 2008. Int J Obes

2015;39:1209-1216.

19. Australian Government. Australian Life Tables 2010-12. Canberra: Commonwealth of

Australia; 2012. Available from: http://www.aga.gov.au/publications/life_table_2010-

12/.

20. Prospective Studies Collaboration. Body-mass index and cause-specific mortality in 900

000 adults: collaborative analyses of 57 prospective studies. Lancet 2009;373:1083–1096.

21. Bihan H, Backholer K, Peeters A, et al. Socioeconomic position and premature mortality in

the AusDiab cohort of Australian adults. Am J Public Health 2016;106:470–477.

22. Olshansky SJ, Passaro DJ, Hershow RC, et al. A potential decline in life expectancy in the

United States in the 21st century. N Engl J Med 2005;352:1138–1145.

Page 18 of 41


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


19

23. Caro JJ, Briggs AH, Siebert U, et al. Modeling good research practices - overview: a report

of the ISPOR-SMDM modeling good research practices task force-1. Value Heal

2012;15:796–803.

24. Wannamethee SG, Shaper AG, Whincup PH, et al. Overweight and obesity and the

burden of disease and disability in elderly men. Int J Obes 2004;28:1374–1382.

25. Weinstein MC, O’Brien B, Hornberger J, et al. Principles of good practice for decision

analytic modeling in health-care evaluation: report of the ISPOR Task Force on Good

Research Practices-Modeling Studies. Value Heal. 2003;6:9–17.

26. Backholer K, Mannan HR, Magliano DJ, et al. Projected socioeconomic disparities in the

prevalence of obesity among Australian adults. Aust N Z J Public Health. 2012;36(6):557–

63.

27. Peeters A, Magliano DJ, Backholer K, et al. Changes in the rates of weight and waist

circumference gain in Australian adults over time: a longitudinal cohort study. BMJ Open

2014;4:e003667.

28. Galobardes B, Lynch J, Smith GD. Measuring socioeconomic position in health research.

Br Med Bull 2007;81–82(1):21–37.

29. Ball K, Crawford D, Ireland P, et al. Patterns and demographic predictors of 5-year weight

change in a multi-ethnic cohort of men and women in Australia. Public Health Nutr

2003;6(03):269–280.

30. Ball K, Crawford D. Socioeconomic status and weight change in adults: a review. Soc Sci

Med 2005;60(9):1987–2010.

31. Stringhini S, Carmeli C, Jokela M, et al. Socioeconomic status and the 25 × 25 risk factors

as determinants of premature mortality: a multicohort study and meta-analysis of 1·7

million men and women. Lancet. 2017;389(10075):1229–37.

Page 19 of 41


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


20

32. Zaninotto P, Head J, Stamatakis E, et al. Trends in obesity among adults in England from

1993 to 2004 by age and social class and projections of prevalence to 2012. J Epidemiol

Community Health 2009;63(2):140–146.

33. Hoffmann K, De Gelder R, Hu Y, et al. Trends in educational inequalities in obesity in 15

European countries between 1990 and 2010. Int J Behav Nutr Phys Act 2017;14:1–10.

34. Devaux M, Sassi F. Social inequalities in obesity and overweight in 11 OECD countries. Eur

J Public Health 2013;23:464–469.

35. Bilger M, Kruger EJ, Finkelstein EA. Measuring socioeconomic inequality in obesity:

looking beyond the obesity threshold. Health Econ 2017;26:1052–1066.

36. Shaw RJ, Green MJ, Popham F, et al. Differences in adiposity trajectories by birth cohort

and childhood social class: evidence from cohorts born in the 1930s, 1950s and 1970s in

the west of Scotland. J Epidemiol Community Health 2014; 68: 550–556.

37. Clarke P, O’Malley PM, Johnston LD, et al. Social disparities in BMI trajectories across

adulthood by gender, race/ ethnicity and lifetime socio-economic position: 1986-2004.

Int J Epidemiol 2009;38(2):499–509.

38. Grech A, Sui Z, Siu H, et al. Socio-demographic determinants of diet quality in Australian

adults using the validated healthy eating index for Australian adults (HEIFA-2013).

Healthcare 2017;5:1–12.

39. Williams ED, Tapp RJ, Magliano DJ, et al. Health behaviours, socioeconomic status and

diabetes incidence: the Australian Diabetes Obesity and Lifestyle Study (AusDiab).

Diabetologia 2010;53:2538–2545.

Page 20 of 41


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


21

Figure legends

Figure 1 Simulated compared with actual BMI trajectories for 4 birth cohorts stratified by SEP

(A) Birth cohort 1966-75 for men, (B) birth cohort 1966-75 for women, (C) birth cohort 1956-65 for

men, (D) birth cohort 1956-65 for women, (E) birth cohort 1946-55 for men, (F) birth cohort 1946-55

for women, (G) birth cohort 1936-45 for men, (H) birth cohort 1936-45 for women. Lines = simulated

BMI trajectory and 95% confidence interval; Circles = observed mean (95% CI) BMI from national

health surveys; turquoise = high SEP; brown = low SEP.

Figure 2. Simulated compared with actual obesity (BMI>30 kg/m2) prevalence for 4 birth cohorts

stratified by SEP

(A) Birth cohort 1966-75 for men, (B) birth cohort 1966-75 for women, (C) birth cohort 1956-65 for

men, (D) birth cohort 1956-65 for women, (E) birth cohort 1946-55 for men, (F) birth cohort 1946-55

for women, (G) birth cohort 1936-45 for men, (H) birth cohort 1936-45 for women. Lines = simulated

obesity prevalence and 95% confidence interval; Circles = observed obesity prevalence (95% CI) from

national health surveys; turquoise = high SEP; brown = low SEP.

Figure 3. Simulated BMI distributions in 1995, 2015 and 2035 for men, 1966-75 birth cohort

(A) High SEP (B) Low SEP. Light grey = 1995; dark grey= 2015; black= 2035. Dotted lines represent

obesity and severe obesity cut-points.

Figure 4. Simulated prevalence of obesity and severe obesity at age 60 for different birth cohorts,

men and women

Brown = obesity (30<BMI<35); red =severe obesity (BMI>35); solid bars= high SEP; hatched bars=

low SEP.

Page 21 of 41


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


Figure 1 Simulated compared with actual BMI trajectories for 4 birth cohorts stratified by SEP(A) Birth cohort 1966-75 for men, (B) birth cohort 1966-75 for women, (C) birth cohort 1956-65 for men,

(D) birth cohort 1956-65 for women, (E) birth cohort 1946-55 for men, (F) birth cohort 1946-55 for women, (G) birth cohort 1936-45 for men, (H) birth cohort 1936-45 for women. Lines = simulated BMI trajectory

and 95% confidence interval; Circles = observed mean (95% CI) BMI from national health surveys; turquoise = high SEP; brown = low SEP.

125x235mm (300 x 300 DPI)

Page 22 of 41


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


Figure 2. Simulated compared with actual obesity (BMI>30 kg/m2) prevalence for 4 birth cohorts stratified by SEP

(A) Birth cohort 1966-75 for men, (B) birth cohort 1966-75 for women, (C) birth cohort 1956-65 for men, (D) birth cohort 1956-65 for women, (E) birth cohort 1946-55 for men, (F) birth cohort 1946-55 for women,

(G) birth cohort 1936-45 for men, (H) birth cohort 1936-45 for women. Lines = simulated obesity prevalence and 95% confidence interval; Circles = observed obesity prevalence (95% CI) from national


140x242mm (300 x 300 DPI)

Page 23 of 41


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


Figure 3. Simulated BMI distributions in 1995, 2015 and 2035 for men, 1966-75 birth cohort(A) High SEP (B) Low SEP. Light grey = 1995; dark grey= 2015; black= 2035. Dotted lines represent


183x272mm (300 x 300 DPI)

Page 24 of 41


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


Figure 4. Simulated prevalence of obesity and severe obesity at age 60 for different birth cohorts, men and women

Brown = obesity (30<BMI<35); red =severe obesity (BMI>35); solid bars= high SEP; hatched bars= low SEP.

173x85mm (300 x 300 DPI)

Page 25 of 41


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


1

SUPPLEMENT MATERIAL



Alison Hayes, Associate Professor, The University of Sydney, Faculty of Medicine and Health, School of Public Health, NSW 2006, Australia

Eng Joo Tan, Health Economics Research Fellow, The University of Sydney, Faculty of Medicine and Health, School of Public Health, NSW 2006, Australia

Anagha Killedar, PhD candidate, The University of Sydney, Faculty of Medicine and Health, School of Public Health, NSW 2006, Australia

Thomas Lung, Health Economics Research Fellow, The George Institute for Global Health, University of New South Wales, NSW 2042

Table of Contents

Section 1: Supplement Methods ....................................................................................................... 2

1.1 Prediction equations for annual weight (BMI) gain, by age, sex and SEP ................................ 2

1.2 Modelling annual mortality ..................................................................................................... 7

1.3 Sensitivity analysis ................................................................................................................ 10

Section 2: Supplement Figures ........................................................................................................ 14

Supplement Figure 1. Simulated compared with actual weight status, for men by birth cohort

and SEP ....................................................................................................................................... 14

Supplement Figure 2. Simulated compared with actual weight status, for women, by birth

cohort and SEP ............................................................................................................................ 15

References ...................................................................................................................................... 16

Page 26 of 41


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


2

Section 1: Supplement Methods

1.1 Prediction equations for annual weight (BMI) gain, by age, sex and SEP

The study populations for deriving equations for annual weight (BMI) gain included data on 7508 persons aged between 20 and 59 from the 1995 National Nutrition Survey (NNS) and 9850 persons aged between 37 and 76 from the 2011/12 National Health survey who had full data on height, weight and education and were not pregnant. The 1995 NNS was the first nationally representative survey in Australia in which height and weight were objectively measured. The NHS administered by the Australian Bureau of statistics (ABS) use a stratified multistage area sampling design including private dwelling in all states and territories across Australia, and are designed to be population representative.

Further details are shown below.

Table A. Characteristics of birth cohorts used to derive weight (BMI) gain equations

Mean (95%CI) BMI

NNS 1995 NHS 2011/12 Birth

cohort Low SEP High SEP Missing Education Low SEP High SEP Missing

Education

1936-45 26.9 (26.2 – 27.6)

27.3 (26.8 – 27.8) 56.3% 29.1

(28.6 - 29.5) 27.5

(26.86 -28.17) 0%

1946-55 26.7 (26.3 – 27.2)

26.2 (25.8 – 26.6) 33.2% 29.6

(29.2 - 30.1) 27.7

(27.2 - 28.1) 0%

1956-65 26.5 (26.1 – 26.9)

25.4 (25.1 – 25.7) 22.0% 29.0

(28.5 - 29.4) 27.7

(27.3 - 28.1) 0%

1966-75 25.11 (24.6 – 25.6)

24.4 (24.0 – 24.7) 13.6% 28.7

(28.3 - 29.2) 27.2

(26.9 - 27.5) 0%

In discrete-time simulation with annual cycles, the BMI of person i at time t, is determined from their BMI at the end of the previous year plus BMI gained during the current year.

BMI it = BMI it-1 + ∆ BMI it

Annual BMI gain (∆ BMI it ) is a function of a number of covariates x1-x3 including age, BMI at the end of the previous year and socioeconomic position.

∆ BMI it = c + β1x1 + β2x2 + β3x3 +є

Estimates of annual BMI change for different sectors of the population were derived using a synthetic cohort technique (1) which matches members of national level cross-sectional health surveys by birth year to estimate change in BMI over longitudinal time for different age and sex cohorts, stratified by socio-economic position and quantiles of BMI. BMI in all surveys was based on objectively measured height and weight. Socio-economic position was defined by completion of senior school education. When analysing data on adults, this is a fixed, time invariant measure, and thus particularly suited to synthetic cohort methodology. The synthetic cohorts were constructed between 1995 and 2012 for men and women aged 20-29, 30-39, 40-49, and 50-59 years in 1995, representing 4 birth cohorts 1936-45, 1946-55, 1956-65 and 1966-75, centred around 1940, 1950, 1960 and 1970, respectively. To capture BMI growth at different positions across the BMI spectrum, we determined BMI change over time in each decile of BMI within synthetic cohorts.

Page 27 of 41


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


3

Annual weight gain based on age, sex, SEP and position on the BMI spectrum was then determined by assuming a fixed annual rate within the 17 years span. Using age and BMI from the mid–point of the matched surveys, this provided 40 estimates (10 from each of 4 synthetic cohort) of BMI change across matched deciles within each synthetic cohort and by SEP group. Finally we used multiple linear regression analysis and followed methods described in (2) to derive separately for men and women, prediction equations for annual change in BMI based on age, current BMI and SEP. For older adults (>75 years) we assumed a small annual weight loss, informed by observations of BMI change from a large Australian longitudinal study (3).

A summary of all the BMI gain equations for men and women are shown below:

Weight gain equations for men

Coefficients and 95% CI of the weight gain equations for men and women are shown in Tables A and B. Table B. Weight gain equations for men

Men under 50 Men over 50

Coefficient (95% CI) p-value Coefficient (95% CI) p-

value

Age -0.0065 (-0.0078, -0.0054) <0.001 - -

BMI 0.0118 (0.0104, 0.0132) <0.001 0.0151 (0.0013, 0.0176) <0.001

School completion -0.0129 (-0.025, -0.0008) 0.038 -0.0731 (-0.0935, -0.0527) <0.001

Constant 0.0909 (0.0317,0.150) 0.004 -0.2987 (-0.3701, -0.2273) <0.001

Adjusted R2 0.91 0.85

Some examples of annual weight gain by SEP Example 1: For a man aged 25, who completed high school and has BMI of 30; Annual weight gain = 0.0909 – 0.0065*25 + 0.0118*30 – 0.0129 = 0.2491 units BMI Example 2: For a man aged 25, who did not complete high school and has a BMI of 30; Annual weight gain = 0.0909 – 0.0065*25 + 0.0118*30 = 0.2620 units BMI Example 3: For a man aged 60, who completed high school and has a BMI of 35; Annual weight gain = -0.2987 + 0.0151*35 – 0.0731 = 0.1567 units BMI Example 4: For a man aged 60, who did not complete high school and has a BMI of 35; Annual weight gain = -0.2987 + 0.0151*35 = 0.2298 units BMI

Page 28 of 41


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


4

Figure A: Annual weight gain in 4 synthetic cohorts centred around age 34 (synthetic cohort 1), age

44 (synthetic cohort 2), age 54 (synthetic cohort 3), and age 64 (synthetic cohort 4). Brown circles = low SEP group; Turquoise circles = high SEP group; Each point represents annual BMI change in deciles of BMI. Brown lines = annual BMI change from regression equation(s) for low SEP; Turquoise lines = annual BMI change from regression equation(s) for low SEP

Page 29 of 41


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


5

Weight gain equations for women

As there was no significant difference in BMI gain between the high and low SEP groups for younger females (p=0.58), an equation already derived and not stratified by SEP (1) was used to predict annual change in BMI for young women. For older females, equations for high and low SEP groups were derived separately. Polynomial splines were used to account for the plateauing of BMI gain for people in higher BMI range the upper part of the BMI spectrum.

Table C. Weight gain equations for women

Women under 50

(High and low SEP)

Women over 50

(High SEP group)

Women over 50

(Low SEP group)

Coefficient

(95% CI) p-value

Coefficient

(95% CI) p-value

Coefficient

(95% CI) p-value

Age -0.0050 (-0.0055,-0.0046) <0.001 -0.0059

(-0.0076, -0.0042) <0.001 - -

BMI under 30

0.0185 (0.0170, 0.0200) <0.001 0.0080

(0.0055, 0.0105) <0.001 0.0187 (0.0158, 0.0216) <0.001

BMI over 30 - - - - 0.0066

(0.0032, 0.010) 0.001

Constant -0.0861 (-0.1268, -0.0454) <0.001 0.2091

(0.0932, 0.3251) 0.001 -0.3478 (-0.4231, -0.2725) <0.001

Adjusted R2 0.92 0.82 0.95

Some examples of annual weight gain by SEP Example 1: For a woman aged 25, and has BMI of 28; Annual weight gain = -0.0861 – 0.0050*25 + 0.0185*28 = 0.3069 units BMI (regardless of SEP status) Example 2: For a woman aged 25, who has a BMI of 35; Annual weight gain = -0.0861 – 0.0050*25 + 0.0185*30 = 0.3439 units BMI (regardless of SEP status) Example 3: For a woman aged 60, who completed high school and has a BMI of 28; Annual weight gain = 0.2091 – 0.0059*60 + 0.0080*28 = 0.0791 units BMI Example 4: For a woman aged 60, who completed high school and has a BMI of 33; Annual weight gain = 0.2091 – 0.0059*60 + 0.0080*30 = 0.0951 units BMI Example 5: For a woman aged 60, who did not complete high school and has a BMI of 33; Annual weight gain = -0.3478 + 0.0187*30 + 0.0066*(33-30) = 0.233 units BMI

Page 30 of 41


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


6

Figure B: Predictions of equation for weight (BMI) gain among women for four synthetic cohorts

centred on ages 34, 44, 54 and 64 years. Brown circles = low SEP; Turquoise circles = high SEP; Each point represents annual BMI change in deciles of BMI. Grey lines = BMI change from regression equation independent of SEP; Brown lines = annual BMI change from regression equation for low SEP; Turquoise lines = annual BMI gain from regression equation for high SEP.

Page 31 of 41


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


7

1.2 Modelling annual mortality

The modelling of age- and SEP- specific mortality is based on the 2011/12 Australian life table (4), a published meta-analysis of the association of BMI and all-cause mortality (5), and the published relative risk of mortality in lower and higher educated groups from a large Australian cohort study (6). Table A shows the age-specific association of BMI and SEP with mortality. Table A. Hazard ratios of increased mortality associated with BMI and socioeconomic position

Age at risk (years)

Hazard ratio per 5 kg/m2 increase in BMI between 25 and 50 kg/m2 (5)

Hazard ratio of low compared with high socioeconomic position (6)

20-34 1* 1.39 (95% CI 1.08 – 1.79) 35-59 1.37 (95% CI 1.31 – 1.42) 1.39 (95% CI 1.08 – 1.79) 60-69 1·32 (95% CI 1·27–1·36) 1.39 (95% CI 1.08 – 1.79) 70-79 1·27 (95% CI 1·23–1·32) 1.39 (95% CI 1.08 – 1.79) 80+ 1·16 (95% CI 1·10–1·23) 1.39 (95% CI 1.08 – 1.79)

* No association was found between BMI and mortality for those less than 35 years of age (5).

The model accounts for an increase in mortality for individuals in higher weight categories, compared with healthy weight for adults aged 35 years and over. This was based on a large meta-analysis and estimated different hazard ratios for different age groups (5). The model also includes an increase in mortality for individuals with low SEP, compared to individuals with high SEP at any age. This was informed by published data (6) from the Australian Diabetes Obesity and Lifestyle (AusDiab) study, a national population based survey of 11,247 adults aged 25 years or older in Australia. The measure of SEP was secondary school education, which matched our study’s measure of SEP. Deriving qxs Conditional probabilities of death (qx) for men and women in single years of age (from the lifetable) were adjusted by SEP and weight status. For each year of age, we took into account the prevalence of 6 weight status and 2 socioeconomic groups. The calculations apportion the conditional probability of death for the entire population of men age x years, into 12 qxs, using the method described in (7). For example, considering just the two SEP groups,

qx = qxl * Pl + qxh* Ph ;

where qx = conditional probability of death at age x for the whole male population qxl = conditional probability of death at age x for the low SEP male subgroup; qxh = conditional probability of death at age x for the high SEP male subgroup; Pl = prevalence of low SEP among men Ph = prevalence of high SEP among men

Since qx, Pl and Ph are known, and we also know that qxl = 1.39 * qxh (6) it is possible to solve for qxh. Example: For example, for a 40 year old man, the qx from the 2011/12 life table is 0.00134. This was firstly partitioned into 6 qxs representing healthy, overweight and obese I-IV categories, taking into account the prevalence of each BMI class for this age using data from the National Health Survey 2011/12. Then the qxs each of the 6 BMI are apportioned to high and low SEP (see following table) shows the 12 qxs derived.

Page 32 of 41


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


8

Table B. qxs for men aged 40 years for different BMI classes and SEP, 2011/12.

Healthy BMI<25

Overweight BMI 25-

25.99

Obese class I BMI 30-

34.99

Obese class 2 BMI 35-

39.99


44.99

Obese class 4 BMI>45

HR for mortality cf healthy weight

1 1.37 1.372 1.373 1.37 4 1.37 5

Prevalence (2011/12)

24.6% 45.2% 20.7% 6.6% 2.5% 0.4%

qx 0. 00088 0. 00121 0. 00165 0. 00226 0. 00310 0. 00425 HR for mortality cf

high SEP 1.39 1.39 1.39 1.39 1.39 1.39

Proportion high SEP (2011/12)

52.6% 56.7% 47.7% 45.0% 59.2% 59.2%

qx high SEP 0.00074 0.00103 0.00136 0.00185 0.00266 0.00365 qx low SEP 0.00103 0.00143 0.00190 0.00258 0.00370 0.00507

The following graphs, show qxs for men and women by age and SEP for selected weight status groups.

Figure A. qxs by high and low SEP groups and weight status Healthy weight (BMI<25); overweight (25<BMI<30); obesity (30<BMI<35); brown circles = low SEP; turquoise circles = high SEP.

Page 33 of 41


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


9

Simulation of mortality In each year of simulation, probability of dying is determined by the qxs for individual years of age and sex, by SEP and weight status. The number of people alive at any time is calculated from the number alive at the start of the year minus the number who have died since the start of the year. Thus:

"# =% ('(# − *'(# ∗ ,-(.(#)0)(12

(13

where Xt= Number of people alive at the end of time t for the whole population '(#= survey weight for ith individual in the simulated data at time t, representing the number of similar people alive at a population level ,-(.(#)= Probability of death for ith person at time t, conditional upon age, sex, BMI and SEP

The total number of people dying each year is determined from the sum across all simulated individuals of the annual probability of dying multiplied by the survey weights. Individual survey weights are adjusted at each time step of the simulation to reflect the number still alive at a population level.

Page 34 of 41


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


10

1.3 Sensitivity analysis

We carried one-way sensitivity analysis of major model parameters by changing to their upper and lower 95% confidence limits and observing the change in the projected prevalence of mean BMI, overall obesity and severe obesity at age 60 years, when compared with the base model. These sensitivity analyses were carried out for men and women of high and low SEP, for 4 different age and birth cohorts, centred around: 1940, 1950, 1960 and 1970.

Parameters investigated in the sensitivity analysis were: a. changing constants in the weight gain equations by upper and lower 95% confidence limits b. changing the hazard ratio for mortality (1.39 (95% CI 1.08 to 1.79) of low compared to high


Sensitivity analysis of annual weight gain Details of the sensitivity analysis of weight gain equations are shown graphically. Changing the constants by upper and lower CI has the result of increasing or decreasing annual weight gain, but not impacting on the slope of the relationship with baseline BMI.

Example: For young men aged 35 the graphs below show the base model prediction for annual weight gain for men of different BMI, and the dashed lines show the upper and lower CI of those predictions, used in the sensitivity analysis. Men aged 35 (brown = low SEP; blue = high SEP)

Men aged 55 (brown = low SEP; blue = high SEP)

Sensitivity analysis of mortality In this sensitivity analysis we investigated changing HR of mortality by low cf high SEP by its upper and lower limits (1.79 & 1.08) – this increases or decreases the risk of mortality of low SEP compared high SEP at all ages, and BMI classes.

Results of the one-way sensitivity analyses in tables A and B, for men and women of 4 birth cohorts. Sensitivity analysis of upper and lower CI of annual weight change has major impacts on BMI, obesity and severe obesity at age 60 and these impacts are more pronounced for the youngest cohort.

2 0 2 5 3 0 3 5 4 0

0 . 0

0 . 1

0 . 2

0 . 3

0 . 4

B M I

An

nu

al

ch

an

ge

in

BM

I d a t a

m o d e l

2 0 2 5 3 0 3 5 4 0

0 . 0

0 . 1

0 . 2

0 . 3

0 . 4

B M I

An

nu

al

ch

an

ge

in

BM

I d a t a

m o d e l

2 0 2 5 3 0 3 5 4 0

- 0 . 2

- 0 . 1

0 . 0

0 . 1

0 . 2

0 . 3

0 . 4

B M I

An

nu

al

ch

an

ge

in

BM

I

m o d e l

d a t a

2 0 2 5 3 0 3 5 4 0

- 0 . 2

- 0 . 1

0 . 0

0 . 1

0 . 2

0 . 3

0 . 4

B M I

An

nu

al

ch

an

ge

in

BM

I

m o d e l

d a t a

Page 35 of 41


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


11

Conversely, changing hazard of mortality by SEP to upper and lower 95% CI had little or no effect on projected mean BMI, obesity and severe obesity at age 60 years. The sensitivity analyses did not affect the pattern of obesity being higher with each successive generation and the conclusion that the youngest 3 cohorts would have much higher socioeconomic inequality at age 60, when compared with the 1940 birth cohort.

Page 36 of 41


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


12

Table A. Sensitivity analysis for males, showing simulated outcomes and absolute difference (inequality) in outcomes between lower and higher SEP groups

Mean BMI at age 60 Obesity prevalence at age 60 (%) Predicted severe obesity at age 60 (%)

SEP Low High

Inequality

(High

minus Low)

Low High

Inequality

(High

minus Low)

Low High

Inequality

(High

minus Low)

Base model

Birth cohort 1940 27.6 27.7 0.1 24.5 25.6 1.1 1.4 2.6 1.2

1950 28.9 27.6 -1.3 36.3 23.1 -13.2 8.3 7.3 -1.0

1960 30.7 28.5 -2.2 47.9 33.6 -14.3 20.3 9.4 -10.9

1970 31.4 29.7 -1.7 50.7 41.1 -9.6 24.6 13.9 -10.7

a) lower 95% CI estimate of the constants in all weight gain equations

Birth cohort 1940 27.2 27.5 0.2 20.5 23.2 2.8 1.4 2.1 0.7

1950 27.9 27.0 -0.8 27.4 19.2 -8.2 5.2 6.5 1.3

1960 28.8 27.2 -1.6 36.8 22.6 -14.2 12.6 4.9 -7.8

1970 28.7 27.5 -1.2 37.2 21.7 -15.5 15.3 9.2 -6.1

b) upper 95% CI estimate of the constants in all weight gain equations

Birth cohort 1940 27.9 28.0 0.1 26.3 28.6 2.3 2.1 3.7 1.6

1950 30.1 28.7 -1.4 46.1 35.0 -11.1 12.5 9.7 -2.8

1960 32.6 30.4 -2.2 63.7 48.0 -15.7 28.1 16.6 -11.5

1970 34.2 32.4 -1.8 73.5 66.0 -7.5 39.5 24.4 -15.1

c) lower 95% CI estimate of the hazard ratio of mortality by SEP

Birth cohort 1940 27.6 27.7 0.1 24.5 25.6 1.1 1.4 2.6 1.1

1950 28.9 27.6 -1.3 36.3 23.0 -13.3 8.3 7.2 -1.1

1960 30.7 28.5 -2.2 47.9 33.4 -14.5 20.3 9.4 -10.9

1970 31.4 29.7 -1.7 50.7 40.9 -9.8 24.7 13.8 -10.9

d) upper 95% CI estimate of the hazard ratio of mortality by SEP

Birth cohort 1940 27.6 27.7 0.1 24.5 25.6 1.2 1.4 2.6 1.2

1950 28.9 27.6 -1.3 36.3 23.1 -13.1 8.3 7.3 -0.9

1960 30.7 28.5 -2.1 47.9 33.7 -14.2 20.2 9.5 -10.7

1970 31.4 29.7 -1.7 50.6 41.2 -9.4 24.6 14.0 -10.6

Page 37 of 41


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


13

Table B. Sensitivity analysis for females, showing simulated outcomes and absolute difference (inequality) in outcomes between lower and higher SEP groups


SEP Low High

Inequality

(High

minus Low)

Low High

Inequality

(High

minus Low)

Low High

Inequality

(High

minus Low)

Base model

Birth cohort 1940 27.5 27.5 0.0 26.5 24.0 -2.5 9.4 10.4 1.0

1950 28.7 27.0 -1.7 38.3 22.1 -16.2 13.9 9.2 -4.7

1960 30.4 28.0 -2.4 43.4 30.4 -13.0 22.7 11.0 -11.7

1970 31.7 29.7 -2.0 53.7 42.3 -11.4 25.8 18.1 -7.7


Birth cohort 1940 27.2 27.1 -0.1 24.8 22.2 -2.6 9.4 8.2 -1.2

1950 27.7 25.9 -1.8 34.7 16.3 -18.4 11.2 6.6 -4.6

1960 28.9 26.5 -2.4 36.0 22.3 -13.7 18.3 8.2 -10.1

1970 29.6 27.5 -2.0 42.0 28.8 -13.3 21.6 12.2 -9.4


Birth cohort 1940 27.9 28.1 0.2 27.0 29.3 2.3 9.4 10.7 1.3

1950 29.7 28.4 -1.3 44.3 28.1 -16.3 15.4 11.9 -3.5

1960 31.9 29.9 -2.0 56.1 42.0 -14.0 29.3 16.7 -12.6

1970 33.8 32.1 -1.7 69.2 57.9 -11.3 36.7 25.2 -11.6


Birth cohort 1940 27.5 27.5 0.0 26.5 24.0 -2.5 9.4 10.3 1.0

1950 28.7 27.0 -1.7 38.4 22.0 -16.4 13.9 9.1 -4.8

1960 30.4 28.0 -2.3 43.4 30.3 -13.1 22.8 11.0 -11.8

1970 31.7 29.7 -2.0 53.8 42.2 -11.6 25.8 18.0 -7.9


Birth cohort 1940 27.5 27.5 0.0 26.5 24.0 -2.4 9.4 10.4 1.0

1950 28.7 27.0 -1.7 38.3 22.1 -16.2 13.9 9.2 -4.7

1960 30.4 28.0 -2.3 43.4 30.5 -12.9 22.7 11.1 -11.6

1970 31.7 29.7 -2.0 53.7 42.4 -11.3 25.8 18.2 -7.6

Page 38 of 41


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


14

Section 2: Supplement Figures Supplement Figure 1. Simulated compared with actual weight status, for men by birth cohort and SEP

Lines = simulated prevalence; circles = prevalence from NHS; turquoise = high SEP; brown = low SEP. Healthy and underweight prevalence BMI<25;

overweight 30>BMI>25; obesity BMI>30; severe obesity BMI>35.

Page 39 of 41


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


15

Supplement Figure 2. Simulated compared with actual weight status, for women, by birth cohort and SEP

Lines = simulated prevalence; circles = prevalence from NHS; turquoise = high SEP; brown = low SEP. Healthy and underweight

prevalence BMI<25; overweight 30>BMI>25; obesity BMI>30; severe obesity BMI>35.

Page 40 of 41


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


16

References 1. Hayes AJ, Lung TWC, Bauman A, Howard K. Modelling obesity trends in Australia:

unravelling the past and predicting the future. Int J Obes 2017; 41: 178-185. 2. Hayes A, Gearon E, Backholer K, Bauman A, Peeters A. Age-specific changes in BMI and BMI

distribution among Australian adults using cross-sectional surveys from 1980 to 2008. Int J Obes 2015; 39: 1209-1216.

3. Cameron AJ, Welborn TA, Zimmet PZ, et al. Overweight and obesity in Australia: the 1999-2000 Australian Diabetes, Obesity and Lifestyle Study (AusDiab). The Medical Journal of Australia 2003; 178: 427-432.

4. Australian Government. Australian Life Tables 2010-12. Canberra: Commonwealth of Australia; 2012. Available from: http://www.aga.gov.au/publications/life_table_2010-12/.

5. Prospective Studies Collaboration. Body-mass index and cause-specific mortality in 900 000 adults: Collaborative analyses of 57 prospective studies. Lancet 2009; 373: 1083–1096.

6. Bihan H, Backholer K, Peeters A, et al. Socioeconomic position and premature mortality in the AusDiab cohort of Australian adults. Am J Public Health 2016; 106: 470–477.

7. Olshansky SJ, Passaro DJ, Hershow RC, et al. A potential decline in life expectancy in the United States in the 21st century. N Engl J Med 2005; 352: 1138–1145.

Page 41 of 41


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


trends in Australia

Journal: BMJ Open

Manuscript ID bmjopen-2018-026525.R1


Date Submitted by the Author: 21-Jan-2019


<b>Primary Subject Heading</b>: Health economics

Secondary Subject Heading: Epidemiology



BMJ Open


1









Thomas Lung, Health Economics Senior Research Fellow, The George Institute for Global Health,

University of New South Wales, NSW 2042




NSW 2006 Australia

[email protected]







Page 1 of 43


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

mailto:[email protected]


2











Word Count: 3376

Page 2 of 43


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


3

Abstract


(BMI) and obesity.




Setting: Australia.




1970.

Results: Simulations projected persistent or widening socioeconomic inequality in BMI and obesity

over the adult life course, for all birth cohorts. Recent birth cohorts were predicted to have greater

socioeconomic inequality by middle age, compared with earlier cohorts. For example, among men,



projected to be 51% [95% CI 43 to 58] and 41% [95% CI 36 to 46] for the low and high SEP groups,


inequality in severe obesity (BMI>35kg/m2) at age 60.




provides a framework for further research to investigate which interventions will be most effective

in narrowing the gap in socioeconomic disparities in obesity in adulthood.


Page 3 of 43


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


4

Article Summary


This is an innovative study and the first to use micro-simulation to increase our

understanding of trends in socioeconomic disparities in BMI and obesity among adults in

Australia.

The model combines best evidence pertaining to obesity progression and mortality in



The model has been presented transparently and externally validated using the most


A limitation is the use of only one indicator of SEP based on educational attainment, which


Page 4 of 43


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


5

What is already known on this subject

In high income countries, a higher burden of obesity is seen in lower socioeconomic groups

There is conflicting evidence of whether inequalities are widening in high income countries

Models are regarded as powerful tools for informing policy decisions in obesity treatment

and prevention, yet few models account for obesity progression within different

socioeconomic strata

What this study adds

Our model predicted that inequalities in obesity among adults in Australia will grow and

there will be a shift towards inequality in severe obesity

Recent generations will experience unprecedented levels of obesity and severe obesity by

the time they reach middle age, with greater socioeconomic inequality, compared with

previous generations

Our model provides a tool to for future investigatation of which interventions will be most

effective in narrowing the gap in socioeconomic disparities in obesity in adulthood

Page 5 of 43


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


6

INTRODUCTION







with BMI above 35.4



disease outcomes for which overweight and obesity is a risk factor.6 However, the current literature

pertaining to high income countries is conflicting on whether inequalities in obesity are persistent,7

widening8 or narrowing.9,10

People generally accrue BMI during their life course11 and, as a result, there has been great interest

in identifying BMI trajectories through longitudinal studies or modelling BMI growth trajectories12 to

understand the epidemiology of disease and to identify at–risk populations. Beyond their value in

epidemiological studies, models are regarded as powerful tools for informing policy decisions, 13 yet

current models of obesity rarely take account of socioeconomic position (SEP), thus overlooking a

key policy-relevant determinant of obesity. There are currently few analytical tools to evaluate which

interventions are most effective in reducing inequalities.14 Simulation models may provide insights

in this context that are not possible with traditional statistical methods,15 but their use is just

beginning.16

Page 6 of 43


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


7

In this study we present a new version of a micro-simulation model for Australia adults17 which

projects obesity in different socioeconomic sub groups, based on educational attainment. In micro-

simulation, individuals are modelled separately, and these methods are well-established in health

economics as a way of predicting outcomes based on individual characteristics.18 An important

consideration before gaining insights from a model is that it predicts consistently across SEP

groups.14 Accordingly, in this study, we validate our model projections against observed BMI

trajectories and obesity progression, stratified by socio-economic position (SEP), and then use the

model to determine future trends in socioeconomic inequalities in obesity – both within and

between generations.

METHODS

Study populations


1995 and 2015, including the 1995 National Nutrition Survey (NNS), and National Health Surveys in

2007/8, 2011/12 and 2014/15.3,19-21 Height and weight were objectively measured by trained

interviewers and participation was voluntary. A stadiometer was used to measure height to a

maximum of 210 cm and correct to one decimal point. Weight was measured using digital scales

measuring to a maximum of 139.9 kg (1995 survey) and to a maximum of 150 kg (later surveys) and

reported correct to one decimal point. Body mass index (BMI) was determined from weight in kg

divided by height in metres squared (m2). Socioeconomic position was based on completion of high

school and derived from responses to the survey question “whether completed secondary school” in

NNS 1995 and “highest year of school completed” for subsequent health surveys. (2007/08 and

beyond). Data from the 1995 and 2011/12 surveys were used to derive equations for age-related

annual weight gain, and all health surveys beyond 1995 were used in validation of model predictions

until 2014/15.

Page 7 of 43


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


8











Methods 1.1) are based on Australian national data, using a synthetic cohort technique.22 The


2011/12 Australian life table,23 a published meta-analysis of the association of BMI and all-cause

mortality,24 and the published relative risk of mortality by SEP from a large Australian cohort study.25










with health survey data from 2007/8, 2011/12 and 2014/15, matched to the same birth cohorts. As

only the 1995 and 2011/12 data were used in the derivation of model equations, this represents

Page 8 of 43


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


9

both internal and external validation.27 Survey estimation (svy command in STATA) was used

throughout the analysis of health survey data and when preparing summary results of simulated

data. These weights take account of the multi-stage sampling, and summary data such as mean BMI

or prevalence of obesity can be inferred at a population level.






this is the age at which obesity related chronic disease starts to become apparent.28 For the two


determined inequality, calculated as differences between high and low SEP in mean BMI, prevalence

of obesity (BMI>30 kg/m2) and prevalence of severe obesity (BMI>35 kg/m2) at age 60 years.



in driving model outputs.29 We changed major model parameters by their upper and lower 95%








Page 9 of 43


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


10



Further details are provided in Supplement Methods 1.3.




the Census Act.

RESULTS





showed good internal and external validation as health survey data were within the simulated 95%

interval.



was corroborated by survey data for each birth cohort (Figure 2). Validation graphs of more weight






Page 10 of 43


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


11



















there was negligible inequality in obesity (30kg/m2>BMI<35kg/m2) between high and low SEP groups


above 30kg/m2 is due to inequality at the extreme upper bound (BMI>35kg/m2).


Page 11 of 43


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


12











Page 12 of 43


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


13

Table 1. Simulated outcomes at age 60 for different birth cohorts of men and women, and difference (inequality) in outcomes between lower and higher SEP groups (High minus Low)

Mean (95%CI) BMI at age 60(kg/m2)

Obesity (BMI >30 kg/m2)prevalence (%) and 95% CI

Severe obesity (BMI>35 kg/m2)Prevalence (%) and 95% CI

Low SEP High SEP Difference Low SEP High SEP Difference Low SEP High SEP Difference

MEN

Birth Cohort 1940 27.6

(26.6, 28.5)27.7

(27.1, 28.3)0.1

(-1.0, 1.2)24.5

(15.5, 33.4)25.6

(18.2, 33.0)1.1

(0.7, 1.5)1.4

(0.0, 2.9)2.6

(0.3, 4.8)1.2

(0.9, 1.5)

1950 28.9(28.2, 29.6)

27.6(27.1, 28.1)

-1.3(-2.2, -0.5)

36.3(29.3, 43.2)

23.1(18.3, 27.8)

-13.2(-13.5, -12.9)

8.3(4.6, 12.0)

7.3(4.3, 10.3)

-1.0(-1.3, -0.7)

1960 30.7(30.0, 31.4)

28.5(28.0, 29.0)

-2.2(-3.0, -1.3)

47.9(41.8, 54.0)

33.6(28.6, 38.5)

-14.3(-14.5, -14.1)

20.3(15.6, 24.9)

9.4(6.5, 12.4)

-10.9(-11.1, -10.7)

1970 31.4(30.5, 32.2)

29.7(29.0, 30.3)

-1.7(-2.8, -0.6)

50.7(43.4, 58.0)

41.1(35.8, 46.4)

-9.6(-9.8, -9.4)

24.6(18.4, 30.9)

13.9(10.4, 17.4)

-10.7(-11.0, -10.4)

WOMEN


(26.3, 28.7)27.5

(26.6, 28.4)0.0

(-1.5, 1.5)26.5

(17.5, 35.5)24.0

(16.7, 31.4)-2.5

(-2.9, -2.1)9.4

(2.3, 16.5)10.4

(4.8, 16.0)1.0

(0.6, 1.4)

1950 28.7(27.9, 29.5)

27.0(26.2, 27.7)

-1.7(-2.8, -0.6)

38.3(31.4, 45.3)

22.1(16.8, 27.3)

-16.2(-16.5, -15.9)

13.9(9.2, 18.6)

9.2(4.9, 13.5)

-4.7(-5.0, -4.4)

1960 30.4(29.5, 31.2)

28.0(27.4, 28.6)

-2.4(-3.4, -1.3)

43.4(37.2, 49.6)

30.4(25.6, 35.1)

-13.0(-13.2, -12.8)

22.7(17.5, 28.0)

11.0(7.7, 14.4)

-11.7(-11.9, -11.45)

1970 31.7(30.4, 33.0)

29.7(29.1, 30.3)

-2.0(-3.4, -0.6)

53.7(46.2, 61.3)

42.3(37.3, 47.2)

-11.4(-11.6, -11.2)

25.8(18.9, 32.7)

18.1(14.1, 22.0)

-7.7(-8.0, -7.4)

Abbreviations: SEP, Socioeconomic position; BMI, Body-mass index; CI, Confidence Interval. Numbers in brackets represent 95% confidence intervals.

Page 13 of 43


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


14

DISCUSSION







related disease.





Another strength is the validation of model projections using the most recently available national


future. We have adhered to good reporting practices for modelling27 and the modelling is






weight gain has been stable, or even slowed.22,31 Another assumption is that there are no changes

over time in the association between BMI and mortality among SEP groups. Nonetheless, in

sensitivity analysis we have investigated the scenarios of annual weight gain and the hazard ratio for

Page 14 of 43


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


15

mortality being higher or lower, and the major conclusions pertaining to the projected widening

inequality in obesity and severe obesity prevalence still hold.



indicator to use in an adult life-course framework.32 There is evidence of education being an

important predictor of weight gain33,34 and the use of an individual level characteristic, is also



meta-analysis35 using occupational status as a marker of SEP, reported a very similar hazard ratio, for


Other socio-economic determinants such as household income, which may change over the life

course were not accounted for in the modelling. Investigations of inequalities in obesity progression

using other measures of socioeconomic position will be an important avenue for future research.

Finally, there were some missing data on education status, in our baseline population in 1995,

particularly for the oldest birth cohort, which could lead to bias. However, our results show good

internal and external validation, suggesting that any bias did not have major impact on the overall

results.




growing. Our finding of widening inequalities in obesity corroborates with existing studies in the US,7

UK,8,36 Australia30 and Europe,37 whilst other developed (OECD) countries report stable inequalities,38


The majority of these studies used traditional statistical analysis. In contrast, our dynamic model

Page 15 of 43


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


16

which accounts for the association of weight gain and mortality with SEP, has allowed us to model

into the future and hence to compare, side by side, 4 different birth cohorts of different SEP, born 10

years apart.



have also found disparities in weight gain by educational or occupational class.40,41 The mediators of


low SEP groups in Australia,42 and poorer health behaviours (smoking and physical activity)43 may






physical activity.









Page 16 of 43


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


17







of the impact of interventions on inequalities in health.16 We hope future research using this model






Page 17 of 43


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


18

References


2004;94(9):1477.












7. Walsemann KM, Ailshire JA. BMI trajectories during the transition to older adulthood:

persistent, widening, or diminishing disparities by ethnicity and education? Res Aging

2011;33:286–311.

8. Bann D, Johnson W, Li L, Kuh D, Hardy R. Socioeconomic inequalities in body mass index

across adulthood: coordinated analyses of individual participant data from three British

birth cohort studies initiated in 1946, 1958 and 1970. PLoS Med 2017;14:1–20.

9. Zhu J, Coombs N, Stamatakis E. Temporal trends in socioeconomic inequalities in obesity

prevalence among economically-active working-age adults in Scotland between 1995 and

2011: a population-based repeated cross-sectional study. BMJ Open 2015;5:1–10.

10. Zhang Q, Wang Y. Trends in the association between obesity and socioeconomic status in

U.S. adults: 1971 to 2000. Obes Res 2004;12:1622–1632.

Page 18 of 43


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


19










15. Speybroeck N, Van Malderen C, Harper S, Müller B, Devleesschauwer B. Simulation

models for socioeconomic inequalities in health: a systematic review. Int J Env Res Pub He

2013;10:5750-5780.





18. Weinstein MC. Recent developments in decision-analytic modelling for economic

evaluation. Pharmacoeconomics 2006;24(11):1043–1053.





2008.



Page 19 of 43


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


20



2015;39:1209-1216.



12/.









2012;15:796–803.








63.



2014;4:e003667.

Page 20 of 43


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


21


Br Med Bull 2007;81–82(1):21–37.



2003;6(03):269–280.


Med 2005;60(9):1987–2010.


















Int J Epidemiol 2009;38(2):499–509.

Page 21 of 43


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


22






Diabetologia 2010;53:2538–2545.

Page 22 of 43


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


23

Figure legends

Figure 1 Simulated compared with actual BMI trajectories for 4 birth cohorts stratified by SEP(A) Birth cohort 1966-75 for men, (B) birth cohort 1966-75 for women, (C) birth cohort 1956-65 for men, (D) birth cohort 1956-65 for women, (E) birth cohort 1946-55 for men, (F) birth cohort 1946-55 for women, (G) birth cohort 1936-45 for men, (H) birth cohort 1936-45 for women. Lines = simulated BMI trajectory and 95% confidence interval; Circles = observed mean (95% CI) BMI from national health surveys; turquoise = high SEP; brown = low SEP.

Figure 2. Simulated compared with actual obesity (BMI>30 kg/m2) prevalence for 4 birth cohorts stratified by SEP(A) Birth cohort 1966-75 for men, (B) birth cohort 1966-75 for women, (C) birth cohort 1956-65 for men, (D) birth cohort 1956-65 for women, (E) birth cohort 1946-55 for men, (F) birth cohort 1946-55 for women, (G) birth cohort 1936-45 for men, (H) birth cohort 1936-45 for women. Lines = simulated obesity prevalence and 95% confidence interval; Circles = observed obesity prevalence (95% CI) from national health surveys; turquoise = high SEP; brown = low SEP.

Figure 3. Simulated BMI distributions in 1995, 2015 and 2035 for men, 1966-75 birth cohort(A) High SEP (B) Low SEP. Light grey = 1995; dark grey= 2015; black= 2035. Dotted lines represent obesity and severe obesity cut-points.

Figure 4. Simulated prevalence of obesity and severe obesity at age 60 for different birth cohorts, men and womenBrown = obesity (30<BMI<35); red =severe obesity (BMI>35); solid bars= high SEP; hatched bars= low SEP.

Page 23 of 43


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960





125x235mm (300 x 300 DPI)

Page 24 of 43


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960






140x242mm (300 x 300 DPI)

Page 25 of 43


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




183x272mm (300 x 300 DPI)

Page 26 of 43


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




173x85mm (300 x 300 DPI)

Page 27 of 43


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


1

SUPPLEMENT MATERIAL







Table of Contents

Section 1: Supplement Methods ...........................................................................................................2

1.1 Prediction equations for annual weight (BMI) gain, by age, sex and SEP ................................2

1.2 Modelling annual mortality ...................................................................................................7

1.3 Sensitivity analysis...............................................................................................................10

Section 2: Supplement Figures ............................................................................................................14

Supplement Figure 1. Simulated compared with actual weight status, for men by birth cohort and SEP .....................................................................................................................................14

Supplement Figure 2. Simulated compared with actual weight status, for women, by birth cohort and SEP ..........................................................................................................................15

References............................................................................................................................................16

Page 28 of 43


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


2

Section 1: Supplement Methods1.1 Prediction equations for annual weight (BMI) gain, by age, sex and SEP



Table A. Characteristics of birth cohorts used to derive weight (BMI) gain equations

Mean (95%CI) BMINNS 1995 NHS 2011/12

Birth cohort Low SEP High SEP Missing

Education Low SEP High SEP Missing Education

1936-45 26.9(26.2 – 27.6)

27.3(26.8 – 27.8) 56.3% 29.1

(28.6 - 29.5)27.5

(26.86 -28.17) 0%

1946-55 26.7(26.3 – 27.2)

26.2(25.8 – 26.6) 33.2% 29.6

(29.2 - 30.1)27.7

(27.2 - 28.1) 0%

1956-65 26.5(26.1 – 26.9)

25.4(25.1 – 25.7) 22.0% 29.0

(28.5 - 29.4)27.7

(27.3 - 28.1) 0%

1966-75 25.11(24.6 – 25.6)

24.4(24.0 – 24.7) 13.6% 28.7

(28.3 - 29.2)27.2

(26.9 - 27.5) 0%




∆ BMI it = c + β1x1 + β2x2 + β3x3 +є


Page 29 of 43


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


3




Coefficients and 95% CI of the weight gain equations for men and women are shown in Tables A and B.

Table B. Weight gain equations for men


Coefficient (95% CI) p-value Coefficient (95% CI) p-value

Age -0.0065 (-0.0078, -0.0054) <0.001 - -

BMI 0.0118 (0.0104, 0.0132) <0.001 0.0151 (0.0013, 0.0176) <0.001


Constant 0.0909 (0.0317,0.150) 0.004 -0.2987 (-0.3701, -0.2273) <0.001Adjusted R2 0.91 0.85

Some examples of annual weight gain by SEP Example 1: For a man aged 25, who completed high school and has BMI of 30; Annual weight gain = 0.0909 – 0.0065*25 + 0.0118*30 – 0.0129 = 0.2695units BMIExample 2: For a man aged 25, who did not complete high school and has a BMI of 30;Annual weight gain = 0.0909 – 0.0065*25 + 0.0118*30 = 0.2824 units BMIExample 3: For a man aged 60, who completed high school and has a BMI of 35;Annual weight gain = -0.2987 + 0.0151*35 – 0.0731 = 0.1567 units BMIExample 4: For a man aged 60, who did not complete high school and has a BMI of 35;Annual weight gain = -0.2987 + 0.0151*35 = 0.2298 units BMI

Page 30 of 43


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


4

Figure A: Annual weight gain in 4 synthetic cohorts centred around age 34 (synthetic cohort 1), age 44 (synthetic cohort 2), age 54 (synthetic cohort 3), and age 64 (synthetic cohort 4). Brown circles = low SEP group; Turquoise circles = high SEP group; Each point represents annual BMI change in deciles of BMI. Brown lines = annual BMI change from regression equation(s) for low SEP; Turquoise lines = annual BMI change from regression equation(s) for low SEP

Page 31 of 43


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


5


As there was no significant difference in BMI gain between the high and low SEP groups for younger females (p=0.58), an equation already derived and not stratified by SEP (1) was used to predict annual change in BMI for young women. For older females, equations for high and low SEP groups were derived separately. Polynomial splines (curves that are defined by two or more points) were used to account for the plateauing of BMI gain for people in higher BMI range the upper part of the BMI spectrum.

Table C. Weight gain equations for women

Women under 50(High and low SEP)

Women over 50(High SEP group)

Women over 50(Low SEP group)

Coefficient(95% CI) p-value Coefficient

(95% CI) p-value Coefficient(95% CI) p-value

Age -0.0050(-0.0055,-0.0046) <0.001 -0.0059

(-0.0076, -0.0042) <0.001 - -

BMI under 30

0.0185(0.0170, 0.0200) <0.001 0.0080

(0.0055, 0.0105) <0.001 0.0187(0.0158, 0.0216) <0.001

BMI over 30 - - - - 0.0066

(0.0032, 0.010) 0.001

Constant -0.0861(-0.1268, -0.0454) <0.001 0.2091

(0.0932, 0.3251) 0.001 -0.3478(-0.4231, -0.2725) <0.001

Adjusted R2 0.92 0.82 0.95

Some examples of annual weight gain by SEP Example 1: For a woman aged 25, and has BMI of 28; Annual weight gain = -0.0861 – 0.0050*25 + 0.0185*28 = 0.3069 units BMI (regardless of SEP status)Example 2: For a woman aged 25, who has a BMI of 35;Annual weight gain = -0.0861 – 0.0050*25 + 0.0185*30 = 0.3439 units BMI (regardless of SEP status)Example 3: For a woman aged 60, who completed high school and has a BMI of 28;Annual weight gain = 0.2091 – 0.0059*60 + 0.0080*28 = 0.0791 units BMIExample 4: For a woman aged 60, who completed high school and has a BMI of 33;Annual weight gain = 0.2091 – 0.0059*60 + 0.0080*30 = 0.0951 units BMIExample 5: For a woman aged 60, who did not complete high school and has a BMI of 33;Annual weight gain = -0.3478 + 0.0187*30 + 0.0066*(33-30) = 0.233 units BMI

Page 32 of 43


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


6

Figure B: Predictions of equation for weight (BMI) gain among women for four synthetic cohorts centred on ages 34, 44, 54 and 64 years. Brown circles = low SEP; Turquoise circles = high SEP; Each point represents annual BMI change in deciles of BMI. Grey lines = BMI change from regression equation independent of SEP; Brown lines = annual BMI change from regression equation for low SEP; Turquoise lines = annual BMI gain from regression equation for high SEP.

Page 33 of 43


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


7


The modelling of age- and SEP- specific mortality is based on the 2011/12 Australian life table (4), a published meta-analysis of the association of BMI and all-cause mortality (5), and the published relative risk of mortality in lower and higher educated groups from a large Australian cohort study (6). Table A shows the age-specific association of BMI and SEP with mortality.

Table A. Hazard ratios of increased mortality associated with BMI and socioeconomic position

Age at risk (years)



20-34 1* 1.39 (95% CI 1.08 – 1.79)35-59 1.37 (95% CI 1.31 – 1.42) 1.39 (95% CI 1.08 – 1.79)60-69 1·32 (95% CI 1·27–1·36) 1.39 (95% CI 1.08 – 1.79)70-79 1·27 (95% CI 1·23–1·32) 1.39 (95% CI 1.08 – 1.79)80+ 1·16 (95% CI 1·10–1·23) 1.39 (95% CI 1.08 – 1.79)


The model accounts for an increase in mortality for individuals in higher weight categories, compared with healthy weight for adults aged 35 years and over. This was based on a large meta-analysis and estimated different hazard ratios for different age groups (5). The model also includes an increase in mortality for individuals with low SEP, compared to individuals with high SEP at any age. This was informed by published data (6) from the Australian Diabetes Obesity and Lifestyle (AusDiab) study, a national population based survey of 11,247 adults aged 25 years or older in Australia. The measure of SEP was secondary school education, which matched our study’s measure of SEP.

Deriving qxsConditional probabilities of death (qx) for men and women in single years of age (from the lifetable) were adjusted by SEP and weight status. For each year of age, we took into account the prevalence of 6 weight status and 2 socioeconomic groups. The calculations apportion the conditional probability of death for the entire population of men age x years, into 12 qxs, using the method described in (7). For example, considering just the two SEP groups,


where qx = conditional probability of death at age x for the whole male populationqxl = conditional probability of death at age x for the low SEP male subgroup;qxh = conditional probability of death at age x for the high SEP male subgroup;Pl = prevalence of low SEP among menPh = prevalence of high SEP among men

Since qx, Pl and Ph are known, and we also know that qxl = 1.39 * qxh (6) it is possible to solve for qxh.

Example: For example, for a 40 year old man, the qx from the 2011/12 life table is 0.00134. This was firstly partitioned into 6 qxs representing healthy, overweight and obese I-IV categories, taking into account the prevalence of each BMI class for this age using data from the National Health Survey 2011/12. Then the qxs each of the 6 BMI are apportioned to high and low SEP (see following table) shows the 12 qxs derived.

Page 34 of 43


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


8

Table B. qxs for men aged 40 years for different BMI classes and SEP, 2011/12.

HealthyBMI<25

OverweightBMI 25-

25.99

Obese class IBMI 30-

34.99

Obese class 2BMI 35-

39.99

Obese class 3BMI 40-

44.99

Obese class 4BMI>45


1 1.37 1.372 1.373 1.37 4 1.37 5

Prevalence(2011/12)

24.6% 45.2% 20.7% 6.6% 2.5% 0.4%

qx 0. 00088 0. 00121 0. 00165 0. 00226 0. 00310 0. 00425HR for mortality cf

high SEP1.39 1.39 1.39 1.39 1.39 1.39


52.6% 56.7% 47.7% 45.0% 59.2% 59.2%

qx high SEP 0.00074 0.00103 0.00136 0.00185 0.00266 0.00365qx low SEP 0.00103 0.00143 0.00190 0.00258 0.00370 0.00507


Figure A. qxs by high and low SEP groups and weight statusHealthy weight (BMI<25); overweight (25<BMI<30); obesity (30<BMI<35); brown circles = low SEP; turquoise circles = high SEP.

Page 35 of 43


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


9

Simulation of mortalityIn each year of simulation, probability of dying is determined by the qxs for individual years of age and sex, by SEP and weight status. The number of people alive at any time is calculated from the number alive at the start of the year minus the number who have died since the start of the year. Thus:

𝑋𝑡 = ∑𝑖 = 𝑛

𝑖 = 1(𝑥𝑖𝑡 ― (𝑥𝑖𝑡 ∗ 𝑝𝑟(𝑑𝑖𝑡)))

where Xt= Number of people alive at the end of time t for the whole population= survey weight for ith individual in the simulated data at time t, representing the number 𝑥𝑖𝑡

of similar people alive at a population level= Probability of death for ith person at time t, conditional upon age, sex, BMI and SEP𝑝𝑟(𝑑𝑖𝑡)


Page 36 of 43


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


10





Sensitivity analysis of annual weight gainDetails of the sensitivity analysis of weight gain equations are shown graphically. Changing the constants by upper and lower CI has the result of increasing or decreasing annual weight gain, but not impacting on the slope of the relationship with baseline BMI.

Example: For young men aged 35 the graphs below show the base model prediction for annual weight gain for men of different BMI, and the dashed lines show the upper and lower CI of those predictions, used in the sensitivity analysis.


20 25 30 35 400.0

0.1

0.2

0.3

0.4

BMI

Ann

ual c

hang

e in

BM

I data

model

20 25 30 35 400.0

0.1

0.2

0.3

0.4

BMI

Ann

ual c

hang

e in

BM

I datamodel


20 25 30 35 40

-0.2

-0.1

0.0

0.1

0.2

0.3

0.4

BMIAnn

ual c

hang

e in

BM

I

model

data

20 25 30 35 40

-0.2

-0.1

0.0

0.1

0.2

0.3

0.4

BMIAnn

ual c

hang

e in

BM

I

model

data


Results of the one-way sensitivity analyses in tables A and B, for men and women of 4 birth cohorts. Sensitivity analysis of upper and lower CI of annual weight change has major impacts on BMI, obesity and severe obesity at age 60 and these impacts are more pronounced for the youngest cohort.

Page 37 of 43


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


11


Page 38 of 43


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


12

Table A. Sensitivity analysis for males, showing simulated outcomes and absolute difference (inequality) in outcomes between lower and higher SEP groups


SEP Low HighInequality

(High minus Low)

Low HighInequality

(High minus Low)

Low HighInequality

(High minus Low)

Base modelBirth cohort 1940 27.6 27.7 0.1 24.5 25.6 1.1 1.4 2.6 1.2

1950 28.9 27.6 -1.3 36.3 23.1 -13.2 8.3 7.3 -1.01960 30.7 28.5 -2.2 47.9 33.6 -14.3 20.3 9.4 -10.91970 31.4 29.7 -1.7 50.7 41.1 -9.6 24.6 13.9 -10.7

a) lower 95% CI estimate of the constants in all weight gain equations Birth cohort 1940 27.2 27.5 0.2 20.5 23.2 2.8 1.4 2.1 0.7

1950 27.9 27.0 -0.8 27.4 19.2 -8.2 5.2 6.5 1.31960 28.8 27.2 -1.6 36.8 22.6 -14.2 12.6 4.9 -7.81970 28.7 27.5 -1.2 37.2 21.7 -15.5 15.3 9.2 -6.1

b) upper 95% CI estimate of the constants in all weight gain equations Birth cohort 1940 27.9 28.0 0.1 26.3 28.6 2.3 2.1 3.7 1.6

1950 30.1 28.7 -1.4 46.1 35.0 -11.1 12.5 9.7 -2.81960 32.6 30.4 -2.2 63.7 48.0 -15.7 28.1 16.6 -11.51970 34.2 32.4 -1.8 73.5 66.0 -7.5 39.5 24.4 -15.1

c) lower 95% CI estimate of the hazard ratio of mortality by SEPBirth cohort 1940 27.6 27.7 0.1 24.5 25.6 1.1 1.4 2.6 1.1

1950 28.9 27.6 -1.3 36.3 23.0 -13.3 8.3 7.2 -1.11960 30.7 28.5 -2.2 47.9 33.4 -14.5 20.3 9.4 -10.91970 31.4 29.7 -1.7 50.7 40.9 -9.8 24.7 13.8 -10.9

d) upper 95% CI estimate of the hazard ratio of mortality by SEP Birth cohort 1940 27.6 27.7 0.1 24.5 25.6 1.2 1.4 2.6 1.2

1950 28.9 27.6 -1.3 36.3 23.1 -13.1 8.3 7.3 -0.91960 30.7 28.5 -2.1 47.9 33.7 -14.2 20.2 9.5 -10.71970 31.4 29.7 -1.7 50.6 41.2 -9.4 24.6 14.0 -10.6

Page 39 of 43


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


13

Table B. Sensitivity analysis for females, showing simulated outcomes and absolute difference (inequality) in outcomes between lower and higher SEP groups


SEP Low HighInequality

(High minus Low)

Low HighInequality

(High minus Low)

Low HighInequality

(High minus Low)

Base modelBirth cohort 1940 27.5 27.5 0.0 26.5 24.0 -2.5 9.4 10.4 1.0

1950 28.7 27.0 -1.7 38.3 22.1 -16.2 13.9 9.2 -4.71960 30.4 28.0 -2.4 43.4 30.4 -13.0 22.7 11.0 -11.71970 31.7 29.7 -2.0 53.7 42.3 -11.4 25.8 18.1 -7.7

a) lower 95% CI estimate of the constants in all weight gain equationsBirth cohort 1940 27.2 27.1 -0.1 24.8 22.2 -2.6 9.4 8.2 -1.2

1950 27.7 25.9 -1.8 34.7 16.3 -18.4 11.2 6.6 -4.61960 28.9 26.5 -2.4 36.0 22.3 -13.7 18.3 8.2 -10.11970 29.6 27.5 -2.0 42.0 28.8 -13.3 21.6 12.2 -9.4

b) upper 95% CI estimate of the constants in all weight gain equationsBirth cohort 1940 27.9 28.1 0.2 27.0 29.3 2.3 9.4 10.7 1.3

1950 29.7 28.4 -1.3 44.3 28.1 -16.3 15.4 11.9 -3.51960 31.9 29.9 -2.0 56.1 42.0 -14.0 29.3 16.7 -12.61970 33.8 32.1 -1.7 69.2 57.9 -11.3 36.7 25.2 -11.6

c) lower 95% CI estimate of the hazard ratio of mortality by SEPBirth cohort 1940 27.5 27.5 0.0 26.5 24.0 -2.5 9.4 10.3 1.0

1950 28.7 27.0 -1.7 38.4 22.0 -16.4 13.9 9.1 -4.81960 30.4 28.0 -2.3 43.4 30.3 -13.1 22.8 11.0 -11.81970 31.7 29.7 -2.0 53.8 42.2 -11.6 25.8 18.0 -7.9

d) upper 95% CI estimate of the hazard ratio of mortality by SEPBirth cohort 1940 27.5 27.5 0.0 26.5 24.0 -2.4 9.4 10.4 1.0

1950 28.7 27.0 -1.7 38.3 22.1 -16.2 13.9 9.2 -4.71960 30.4 28.0 -2.3 43.4 30.5 -12.9 22.7 11.1 -11.61970 31.7 29.7 -2.0 53.7 42.4 -11.3 25.8 18.2 -7.6

Page 40 of 43


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


14

Section 2: Supplement FiguresSupplement Figure 1. Simulated compared with actual weight status, for men by birth cohort and SEP

Lines = simulated prevalence; circles = prevalence from NHS; turquoise = high SEP; brown = low SEP. Healthy and underweight prevalence BMI<25; overweight 30>BMI>25; obesity BMI>30; severe obesity BMI>35.

Page 41 of 43


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


15


Lines = simulated prevalence; circles = prevalence from NHS; turquoise = high SEP; brown = low SEP. Healthy and underweight prevalence BMI<25; overweight 30>BMI>25; obesity BMI>30; severe obesity BMI>35.

Page 42 of 43


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


16

References1. Hayes AJ, Lung TWC, Bauman A, Howard K. Modelling obesity trends in Australia:

unravelling the past and predicting the future. Int J Obes 2017; 41: 178-185.2. Hayes A, Gearon E, Backholer K, Bauman A, Peeters A. Age-specific changes in BMI and BMI







Page 43 of 43


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960


trends in Australia

Journal: BMJ Open

Manuscript ID bmjopen-2018-026525.R2


Date Submitted by the Author: 14-Feb-2019


<b>Primary Subject Heading</b>: Health economics

Secondary Subject Heading: Epidemiology



BMJ Open