Caroline K. Kramer,1,2 Balakumar Swaminathan,1 Anthony J. Hanley,1,2,3 Philip W. Connelly,2,4

Mathew Sermer,5 Bernard Zinman,1,2,6 and Ravi Retnakaran1,2,6

Prospective Associations ofVitamin D Status With b-CellFunction, Insulin Sensitivity, andGlycemia: The Impact ofParathyroid Hormone StatusDiabetes 2014;63:3868–3879 | DOI: 10.2337/db14-0489

Previous studies have yielded conflicting findings onthe relationship between low vitamin D (25-OH-D) andimpaired glucose homeostasis. In this context, wehypothesized that combined assessment of 25-OH-Dwith its regulator parathyroid hormone (PTH) may berequired for optimal evaluation of the impact of vitaminD status on glucose metabolism. Thus, we evaluated theprospective associations of 25-OH-D and PTH at 3 monthspostpartum with b-cell function (Insulin Secretion-Sensitivity Index-2 [ISSI-2]), insulin sensitivity (Matsudaindex), and glycemia at 12 months postpartum in 494women undergoing serial metabolic characterization.Notably, 32% of those with prediabetes/diabetes mel-litus at 12 months postpartum had both vitamin D de-ficiency and PTH in the highest tertile at 3 monthspostpartum. On multiple-adjusted linear regressionanalyses, vitamin D deficiency/insufficiency with PTHin the highest tertile at 3 months independently pre-dicted poorer b-cell function (P = 0.03) and insulin sen-sitivity (P = 0.01) and increased fasting (P = 0.03) and 2-hglucose (P = 0.002) at 12 months postpartum. In con-trast, vitamin D deficiency/insufficiency with lowerPTH did not predict these outcomes. In conclusion, onlyvitamin D deficiency/insufficiency with increased PTHis an independent predictor of b-cell dysfunction, insu-lin resistance, and glycemia, highlighting the need for

consideration of the PTH/25-OH-D axis when studyingthe impact of vitamin D status on glucose homeostasis.

In recent years, a growing body of evidence has demon-strated extraskeletal associations of both vitamin D (25-OH-D) (1–3) and parathyroid hormone (PTH) (4–6), withparticular focus on their metabolic implications. Severalstudies have suggested that low levels of vitamin D mayplay a role in the development of type 2 diabetes mellitus(T2DM) (7–9). Indeed, in a meta-analysis of 21 studies,the circulating level of 25-OH-D was inversely associatedwith the risk of future T2DM (7). Furthermore, deterio-ration in b-cell function has been suggested as a patho-physiologic mechanism through which lower vitamin Dmay increase the risk of T2DM (10,11). Previous studieshave also reported that increased PTH is associated withinsulin resistance (12) and metabolic syndrome (4,13),and deterioration of insulin sensitivity and b-cell functionhas been described in hyperparathyroid states (14). Con-versely, however, many investigators have questionedthe association of vitamin D and PTH with glucose me-tabolism, particularly in light of several observationalstudies (13,15–18) and interventional trials (1,19–21)that either have been negative or showed only modest

1Leadership Sinai Centre for Diabetes, Mount Sinai Hospital, Toronto, Ontario,Canada2Division of Endocrinology, University of Toronto, Toronto, Ontario, Canada3Department of Nutritional Sciences, University of Toronto, Toronto, Ontario,Canada4Keenan Research Centre for Biomedical Science of St. Michael’s Hospital,Toronto, Ontario, Canada5Division of Obstetrics and Gynecology, University of Toronto, Toronto, Ontario,Canada

6Lunenfeld-Tanenbaum Research Institute, Mount Sinai Hospital, Toronto,Ontario, Canada

Corresponding author: Ravi Retnakaran, rretnakaran@mtsinai.on.ca.

Received 25 March 2014 and accepted 21 May 2014.

© 2014 by the American Diabetes Association. Readers may use this article aslong as the work is properly cited, the use is educational and not for profit, andthe work is not altered.

See accompanying article, p. 3593.

3868 Diabetes Volume 63, November 2014

PATHOPHYSIO

LOGY

beneficial effects of vitamin D supplementation on glu-cose homeostasis.

A possible explanation for this conflicting evidence isthat previous studies have generally evaluated the re-spective metabolic implications of vitamin D and PTH inisolation rather than considering both hormones togetheras a reflection of the status of the PTH–vitamin D axis.Indeed, for the comprehensive assessment of many endo-crine axes (e.g., thyroid), it is often necessary to considerboth upstream regulators (e.g., thyroid-stimulating hor-mone) and downstream effector hormones (e.g., thyrox-ine) in conjunction with one another. Moreover, in thecase of vitamin D, the 25-OH-D concentration that pro-vides maximal PTH suppression is widely variable, sug-gesting that there is an individual threshold for theserum 25-OH-D concentration below which PTH rises(22). In this context, we hypothesized that combined as-sessment of PTH and 25-OH-D together may be neededfor optimal evaluation of the impact of vitamin D statuson glucose metabolism. Specifically, it may be that glucosemetabolism is only adversely affected when circulating25-OH-D falls to a level that causes PTH to rise, reflectingtrue functional vitamin D inadequacy. Thus, our objectivein this study was to collectively evaluate vitamin D andPTH in relation to changes over time in b-cell function,insulin sensitivity, and glycemia in a cohort of subjectsreflecting a broad range of diabetic risk.

RESEARCH DESIGN AND METHODS

ParticipantsThis study was performed in the setting of a prospectiveobservational cohort consisting of women representingthe full spectrum of glucose tolerance in a recent preg-nancy (from normal to gestational diabetes mellitus[GDM]), who thereby have a broad range of risk for thefuture development of prediabetes and T2DM in the yearsafter delivery (23–25). This cohort provided a model forstudying the longitudinal relationship between vitaminD/PTH and glucose metabolism in the first year postpar-tum for two reasons: 1) lactating women in the first yearpostpartum in Toronto, Ontario, Canada (latitude 43°429N) are a population at risk for vitamin D deficiency/insufficiency and 2) the range of future diabetic riskwithin this cohort has been shown to manifest in changesin b-cell function, insulin sensitivity, and glycemia be-tween 3 and 12 months postpartum (23,26).

As previously described (23–25), the women compris-ing this cohort were recruited at the time of antepartumscreening for GDM in the late 2nd trimester and under-went metabolic characterization at recruitment and atboth 3 months and 12 months postpartum. At our in-stitution, women are screened for GDM by a 50-g glucosechallenge test (GCT) in the late 2nd trimester, followed byreferral for a diagnostic oral glucose tolerance test (OGTT)if the GCT is abnormal. In this cohort study, women arerecruited either before or after the GCT, and all partici-pants undergo a 3-h 100-g OGTT for determination of

GDM status (regardless of the GCT result) (27). The re-sultant cohort thus reflects the full spectrum of glucosetolerance in pregnancy from normal to GDM, whichtranslates to a gradient of future risk for postpartum pro-gression to prediabetes and T2DM. For this cohort study,participants return to the clinical investigation unit atboth 3 and 12 months postpartum to undergo repeatmetabolic characterization, including evaluation of glu-cose tolerance by 2-h 75-g OGTT. The protocol has beenapproved by the Mount Sinai Hospital Research EthicsBoard, and all women have provided written informedconsent for their participation. The current study wasperformed in 494 women who have completed their 12-month postpartum visit and had vitamin D and PTHmeasured at 3 months postpartum, thereby enabling assess-ment of the longitudinal relationships between vitaminD/PTH status and metabolic outcomes (b-cell function,insulin sensitivity, glycemia) 9 months later.

Study Visits at 3 and 12 Months PostpartumAt the study visits at 3 and 12 months postpartum,interviewer-administered questionnaires were completedincluding assessment of current medications/supplements,duration of breast-feeding, and physical activity. As pre-viously described (25), physical activity was assessed bythe validated Baecke questionnaire. This instrument mea-sures total physical activity and its component domains ofsport-related physical activity (sport index), nonsportleisure-time activity (leisure-time index), and occupation-associated activity (work index). Work index is not mea-sured at 3 months postpartum, as most women wouldnot yet have returned to their usual occupation at thattime. At each study visit, physical examination was per-formed including measurement of weight and waistcircumference.

Laboratory Measurements and Physiologic IndicesVitamin D status was assessed with measurement ofserum 25-OH-D by competitive electrochemiluminescentimmunoassay on the Roche Modular E170 (cat. no.05894913190; Roche Diagnostics, Laval, Canada). Thisassay has a lower reporting limit of 8 nmol/L. Serum PTHwas measured using an electrochemiluminescence immu-noassay on the Roche Modular E170 Analyzer (cat. no.11972103122; Roche Diagnostics), which has a detectionrange from 0.6 to 530 pmol/L.

All OGTTs at 3 and 12 months postpartum wereperformed in the morning after an overnight fast. Venousblood samples were drawn for the measurement ofglucose and specific insulin at fasting and at 30, 60, and120 min after the ingestion of the glucose load, aspreviously described (23,24).

At each OGTT, current glucose tolerance status wasdetermined according to Canadian Diabetes Associationguidelines (28). Dysglycemia refers to prediabetes (im-paired glucose tolerance, impaired fasting glucose, orboth) or T2DM. Area under the insulin curve (AUCins)and area under the glucose curve (AUCgluc) during the

diabetes.diabetesjournals.org Kramer and Associates 3869

OGTT were calculated using the trapezoidal rule. Insulinsensitivity was measured using the Matsuda index, anestablished measure of whole-body insulin sensitivity thathas been validated against the euglycemic-hyperinsulinemicclamp (29). b-Cell function was assessed on each OGTTwith the Insulin Secretion-Sensitivity Index-2 (ISSI-2).ISSI-2 is a validated measure of b-cell function that isanalogous to the disposition index obtained from the in-travenous glucose tolerance test (30,31). ISSI-2 has beendirectly validated against the disposition index from theintravenous glucose tolerance test, with which it exhibitsstronger correlation than do other OGTT-derived measuresof b-cell function (31), and has been used to measure b-cellfunction in several previous studies, including both clinicaltrials and observational studies (10,23,32–35). ISSI-2 isdefined as the product of 1) insulin secretion measuredby the ratio of AUCins to AUCgluc and 2) insulin sensitivitymeasured by Matsuda index (30,31).

Statistical AnalysesAll analyses were conducted using SAS 9.1 (SAS Institute,Cary, NC). Continuous variables were tested for normalityof distribution, and natural log transformations of skewedvariables were used, where necessary, in subsequentanalyses.

Participants were initially stratified into groups accord-ing to 1) vitamin D status at 3 months postpartum and 2)tertiles of PTH at 3 months postpartum, respectively.Vitamin D status was classified as per Endocrine Societyguidelines as vitamin D deficient (25-OH-D ,50 nmol/L)(n = 161), vitamin D insufficient (25-OH-D $50 nmol/Land ,75 nmol/L) (n = 178), or vitamin D sufficient (25-OH-D $75 nmol/L) (n = 155) (2,36). The tertiles of PTHwere defined as 1st tertile (PTH #2.6 pmol/L) (n = 167),2nd tertile (PTH .2.6 pmol/L and #3.8 pmol/L) (n =174), and 3rd tertile (PTH .3.8 pmol/L) (n = 153). Uni-variate differences across both the vitamin D groupsand the tertiles of PTH were assessed by ANOVA andWilcoxon rank sum test for continuous variables andthe x2 test for categorical variables (Table 1).

For evaluation of whether these vitamin D and PTHgroups at 3 months postpartum predict insulin sensitivity,b-cell function, and glycemia at 12 months postpartum,multiple linear regression models were constructed withmetabolic outcomes of insulin sensitivity (Matsuda index),b-cell function (ISSI-2), and glycemia (fasting glucose and 2-hglucose) at 12 months postpartum as dependent variables(Table 2). In each case, model 1 was adjusted for riskfactors for diabetes mellitus (age, ethnicity, family historyof diabetes mellitus, previous GDM, BMI at 3 months)and baseline levels of the outcome. The subsequent mod-els were further adjusted for possible confounders thatcould impact the association of vitamin D/PTH and themetabolic outcomes: duration of breast-feeding (model 2)and total physical activity and season of blood collection(model 3). The models were tested for collinearity of covariatesusing the variance inflation factor, which confirmed no

significant collinearity. Sensitivity analyses were per-formed with adjustment for use of calcium and vitaminD supplements, smoking status, and change in BMI be-tween 3 and 12 months postpartum. In addition, all anal-yses were repeated after excluding the 19 participantswith PTH above the upper limit of the laboratory normalrange (6.5 pmol/L).

To evaluate the combined impact of vitamin D statusand PTH tertile at 3 months on future glucose tolerance,we evaluated the prevalence of dysglycemia (prediabetesor diabetes mellitus) at 12 months postpartum in groupsdefined by both vitamin D status and PTH using the x2

test (Fig. 1). For further evaluation of the combined im-pact of vitamin D and PTH status on metabolic outcomes,participants were stratified into four groups as follows: 1)vitamin D sufficient and PTH in the 1st/2nd tertile (ref-erence group,; 2) vitamin D sufficient and PTH in the 3rdtertile, 3) vitamin D deficient/insufficient and PTH in the1st/2nd tertile, and 4) vitamin D deficient/insufficientand PTH in the 3rd tertile. We then tested for a biologicalinteraction between vitamin D deficiency/insufficiencyand PTH in the 3rd tertile. Using the equation describedby Rothman (37) and others (38–40), we evaluatedwhether the combined impact of vitamin D deficiency/insufficiency with PTH in the 3rd tertile on the outcomesof dysglycemia, lowest tertile of Matsuda index, and low-est tertile of ISSI-2 at 12 months postpartum exceededthe sum of the individual effects of these conditions alone(Fig. 2). With this approach (38–40), the following threemeasures of biological interaction are calculated to quan-tify the amount of interaction on a multiplicative scale:the relative excess risk due to interaction (RERI), theattributable proportion due to interaction (AP), and thesynergy index (S). RERI can be interpreted as the risk thatis additional to that which is to be expected on the basisof addition of the odds ratios (OR) under exposure, cal-culated as the difference between the expected risk andthe observed risk (RER1 = OR12 – OR1 – OR2 + 1). AP canbe interpreted as the proportion of disease that is due tointeraction among persons with both exposures (AP =RER1/OR12). S can be interpreted as the excess riskfrom both exposures in the setting of interaction, relativeto the risk from exposure without interaction: S = [OR12 21]/[(OR1 2 1) + (OR2 – 1)]. As previously described(37,38), RERI = 0, AP = 0, and S = 1 indicate the absenceof biological interaction.

Using ANCOVA, we compared the percentage changefrom 3 to 12 months postpartum for insulin sensitivity(Matsuda index), b-cell function (ISSI-2), and glycemia(fasting glucose and 2-h glucose) between these four vita-min D/PTH groups, adjusted for age, ethnicity, family his-tory of diabetes mellitus, previous GDM, and BMI at 3months (Fig. 4). Finally, sequentially adjusted multiple linearregression models (Table 3) were constructed with Matsudaindex, ISSI-2, fasting glucose, and 2-h glucose at 12 monthspostpartum as the outcomes using the same approach tocovariate adjustment as described earlier for Table 2.

3870 Vitamin D/PTH Axis and Glucose Metabolism Diabetes Volume 63, November 2014

Tab

le1—Baseline

compariso

nsbetw

eengroup

sofvitam

inD

status(deficient/insuffi

cient/sufficient)at

3months

postp

artumand

betw

eentertiles

ofPTH

at3months

postp

artum

Vitam

inD

status

P

Tertilesof

PTH

P

Deficient

(25-OH-D

,50

nmol/L)

Insufficient(50

nmol/L

#

25-OH-D

,75

nmol/L)

Sufficient

(25-OH-D

$75

nmol/L)

1st(PTH

#2.6

pmol/L)

2nd(2.6

pmol/L

,PTH

#3.8

pmol/L)

3rd(PTH

.3.8

pmol/L)

n161

178155

167174

153

Months

postp

artum3(3–4)

3(3–4)

3(3–4)

0.103(3–4)

3(3–4)

3(3–4)

0.05

Age

(years)34.5

64.2

34.96

4.535.0

64.1

0.4734.8

64.4

34.86

4.534.6

63.9

0.88

Ethnicity

(%)

,0.001

0.06White

60.073.6

81.971.4

75.967.3

Asian

14.311.2

9.014.3

11.09.1

Other

25.515.2

9.014.3

13.223.5

Family

historyof

T2DM

(%)

57.054.0

43.20.04

47.650.3

56.90.25

Current

smoking

(%)

8.13.3

1.30.008

3.04.6

5.20.58

Season

ofblood

sample

collection(%

)0.47

0.19Winter

27.920.2

20.718.4

25.924.2

Spring

27.928.0

23.923.2

27.629.4

Sum

mer

17.421.3

23.927.4

18.417.0

Fall26.7

30.331.6

30.128.2

29.4

Totalphysicalactivity

4.76

1.05.0

61.0

5.16

1.00.001

5.06

0.95.0

61.1

4.86

1.00.06

Sport

index

1.8(1.5

–2.3)2.0

(1.5–2.5)

2.0(1.8

–2.5)0.03

2.0(1.5

–2.5)2.0

(1.5–2.5)

1.8(1.5

–2.6)0.32

Leisure-timeind

ex2.8

(2.5–3.3)

3.0(2.5

–3.3)3.0

(2.5–3.5)

0.0023.0

(2.8–3.5)

3.0(2.5

–3.2)2.8

(2.5–3.3)

0.03

Duration

ofbreast-feed

ing(m

onths)3.0

(2.0–4.0)

3.0(2.5

–3.0)3.0

(3.0–3.5)

0.303.0

(3.0–3.0)

3.0(3.0

–4.0)3.0

(2.0–4.0)

0.75

25-OH-D

(nmol/L)

35.76

10.264.4

67.4

91.96

12.5,0.001

72.66

21.864.1

624.1

50.16

23.3,0.001

PTH

(pmol/L)

4.0(3.2

–4.8)2.9

(2.2–3.7)

2.7(2.1

–3.5),0.001

2.1(1.8

–2.4)3.2

(2.9–3.5)

4.8(4.3

–5.6),0.001

BMI(kg/m

2)27.8

(23.9–32.3)

25.3(23.4

–28.9)24.5

(22.1–27.6)

,0.001

24.5(22.3

–27.6)26.0

(23.2–29.2)

27.6(24.1

–32.1),0.001

Waist

circumference

(cm)

90.0(83

–100)88.0

(82–96)

86.0(81

–93)0.03

85.5(80

–92)88.5

(81–97)

91.0(83

–100),0.001

Matsud

aind

ex7.7

(5.0–11.6)

11.6(7.8

–15.5)14.0

(9.0–17.6)

,0.001

12.2(8.4

–16.6)11.1

(6.9–15.8)

8.8(5.8

–12.9),0.001

ISSI-2

706(543

–933)733

(625–994)

797(602

–1,045)0.04

810(622

–1,009)721

(581–958)

741(565

–955)0.06

Fastingglucose

onOGTT

(mmol/L)

4.76

0.64.6

60.4

4.56

0.4,0.001

4.56

0.44.6

60.5

4.76

0.50.004

2-hglucose

onOGTT

(mmol/L)

6.3(5.3

–7.5)6.0

(5.1–7.3)

5.7(4.8

–6.7)0.002

5.8(5.1

–7.0)6.1

(5.2–7.4)

6.0(4.9

–7.2)0.16

Continuous

data

arepresented

asmean

6SD

(ifnorm

allydistrib

uted)or

med

ian(25th

–75th)ifskew

ed.

diabetes.diabetesjournals.org Kramer and Associates 3871

RESULTS

The study population consisted of 494 women aged34.8 6 4.3 years. There were no women with renal dis-ease or other serious medical comorbidities. We first eval-uated the respective metabolic implications of theirvitamin D status and their PTH status, in turn, beforeconsidering both vitamin D and PTH together.

Vitamin D Status: Cross-sectional AssociationsTable 1 shows baseline (3 months postpartum) charac-teristics of the participants stratified into the followingthree groups based on their vitamin D status: vitamin Ddeficient (n = 161; 33% of study population), insuffi-cient (n = 178; 36%), and sufficient (n = 155; 31%). Asanticipated, the groups differed in ethnicity (P ,0.001), with the vitamin D–deficient group having the

greatest proportion of nonwhite ethnicity (39.8%). Inaddition, the vitamin D–deficient group had higherprevalence rates of family history of diabetes mellitus(P = 0.04) and current smoking (P , 0.008) and lowerlevels of physical activity at 3 months postpartum(all P , 0.03).

Metabolically, BMI (P , 0.001) and waist circumfer-ence (P = 0.03) differed across the groups, with both beinghighest in the deficient group. Consistent with these dif-ferences in adiposity, the vitamin D groups also differedwith respect to insulin sensitivity, b-cell function, andglycemia at 3 months postpartum. Specifically, therewas a stepwise decrease in Matsuda index (P , 0.001)and ISSI-2 (P = 0.04) from the sufficient to insufficient todeficient group, coupled with an analogous progressiveincrease in fasting glucose (P , 0.001) and 2-h glucose(P = 0.002) across these groups.

Table 2—Multiple linear regression models showing adjusted estimates for vitamin D status at 3 months postpartum and PTHtertile at 3 months postpartum in predicting glucose/metabolic outcomes at 12 months postpartum adjusted for the indicatedcovariates in each model

Outcomes at 12 months

Vitamin D status at 3 months Tertiles of PTH at 3 months

Deficient(25-OH-D ,50

nmol/L)

Insufficient(50 nmol/L #

25-OH-D ,75 nmol/L)

2nd(2.6 pmol/L ,

PTH # 3.8 pmol/L)

3rd(PTH .3.8pmol/L)

Estimate P Estimate P Estimate P Estimate P

Matsuda indexModel 1: age, ethnicity, family history of

T2DM, previous GDM, BMI, Matsudaat 3 months 20.105 0.06 20.014 0.77 20.027 0.59 20.135 0.01

Model 2: model 1 plus duration ofbreast-feeding 20.107 0.06 20.013 0.79 20.032 0.52 20.141 0.008

Model 3: model 2 plus physical activityand season 20.099 0.08 20.013 0.79 20.032 0.53 20.143 0.008

ISSI-2Model 1: age, ethnicity, family history

of T2DM, previous GDM, BMI, ISSI-2at 3 months 20.051 0.23 20.116 0.004 20.041 0.30 20.077 0.07

Model 2: model 1 plus duration ofbreast-feeding 20.051 0.23 20.116 0.004 20.040 0.31 20.076 0.07

Model 3: model 2 plus physical activityand season 20.062 0.16 20.111 0.006 20.042 0.30 20.075 0.08

Fasting glucoseModel 1: age, ethnicity, family history

of T2DM, previous GDM, BMI, fastingglucose at 3 months 0.023 0.02 0.011 0.20 0.003 0.69 0.009 0.34

Model 2: model 1 and duration ofbreast-feeding 0.023 0.02 0.011 0.20 0.004 0.66 0.009 0.33

Model 3: model 2 plus physical activityand season 0.026 0.008 0.011 0.20 0.004 0.67 0.010 0.25

2-h glucoseModel 1: age, ethnicity, family history

of T2DM, previous GDM, BMI, 2-hglucose at 3 months 0.069 0.009 0.054 0.03 0.032 0.18 0.061 0.02

Model 2: model 1 plus duration ofbreast-feeding 0.069 0.009 0.053 0.03 0.034 0.16 0.063 0.01

Model 3: model 2 plus physical activityand season 0.070 0.01 0.051 0.04 0.042 0.09 0.071 0.007

For the vitamin D groups, the reference group is vitamin D sufficiency at 3 months postpartum. For PTH tertiles, the reference group isthe 1st tertile at 3 months postpartum. Boldface type indicates P , 0.05. Outcome variables are log transformed.

3872 Vitamin D/PTH Axis and Glucose Metabolism Diabetes Volume 63, November 2014

Vitamin D Status: Longitudinal Covariate-AdjustedAnalysesOnmultiple linear regression analyses (Table 2), vitamin Ddeficiency at 3 months postpartum independently predicted

increased fasting glucose (P = 0.008) and 2-h glucose (P =0.01) at 12 months postpartum compared with the vitaminD–sufficient (reference) group. However, by itself, vitaminD deficiency did not independently predict poorer insulinsensitivity (Matsuda index) or b-cell function (ISSI-2). Vi-tamin D insufficiency at 3 months predicted decreasedb-cell function (P = 0.006) and higher 2-h glucose (P =0.04) at 12 months postpartum compared with the refer-ence group.

PTH Status: Cross-sectional AssociationsAs shown in Table 1, the study population was also strat-ified into tertiles based on serum PTH at 3 months post-partum. In contrast to the vitamin D groups, these PTHgroups did not differ in ethnicity, family history of di-abetes mellitus, and current smoking. Leisure-time indexwas the only physical activity measure that differed be-tween the PTH tertiles and was lowest in the 3rd tertile(P = 0.03). As expected, 25-OH-D differed across thegroups (P , 0.001).

In the same way as observed across the vitamin Dgroups, BMI and waist differed across the PTH tertiles,being highest in the 3rd tertile (both P , 0.001). Inaddition, insulin sensitivity progressively decreased (P ,0.001) and fasting glucose increased (P = 0.004) from the

Figure 1—Prevalence of dysglycemia (prediabetes or diabetes mel-litus) at 12 months postpartum within each strata of vitamin D statusand PTH tertile at 3 months postpartum. P value refers to overallcomparison across the groups.

Figure 2—Excess risk for dysglycemia (A), lowest tertile of Matsuda index (B), and lowest tertile of ISSI-2 (C) at 12 months postpartumattributed to vitamin (Vit) D insufficiency (insuf.)/deficiency (def.), PTH in 3rd tertile, and their combined effect compared with vitamin Dsufficiency with PTH in 1st/2nd tertile (reference group).

diabetes.diabetesjournals.org Kramer and Associates 3873

1st to 2nd to 3rd tertile. ISSI-2 was lowest in the 3rdtertile, but the overall comparison across the groups didnot reach significance (P = 0.06).

PTH Status: Longitudinal Covariate-Adjusted AnalysesOn multiple linear regression analyses (Table 2), the 2ndtertile of PTH at 3 months postpartum did not independentlypredict insulin sensitivity, b-cell function, fasting glucose, or2-h glucose at 12 months compared with the (reference) 1sttertile. However, the 3rd tertile of PTH at 3 months indepen-dently predicted lower insulin sensitivity (P = 0.008) andhigher 2-h glucose (P = 0.007) 9 months later.

Vitamin D and PTH Status CombinedRecognizing that evaluation of vitamin D in conjunctionwith PTH might better reflect the status of the vitaminD/PTH axis than would consideration of either hormone

in isolation, we next assessed the prevalence of dysglyce-mia (prediabetes or diabetes mellitus) at 12 months inrelation to vitamin D status and PTH tertile combined.Indeed, as shown in Fig. 1, there was a striking differencein the prevalence of dysglycemia at 12 months postpar-tum across the nine groups defined by vitamin D status(sufficient, insufficient, and deficient) and PTH tertile at 3months postpartum (P = 0.003), with the highest preva-lence in the group with vitamin D deficiency and PTH inthe 3rd tertile (30%). Overall, of the 79 participants withprediabetes/diabetes mellitus at 12 months postpartum,32% were in the group with vitamin D deficiency and PTHin the 3rd tertile at 3 months.

To further evaluate the combined impact of vitamin Dand PTH on metabolic parameters, we stratified partic-ipants into the following four groups based on vitamin D

Table 3—Multiple linear regression models showing adjusted estimates for vitamin D/PTH groups at 3 months postpartum inpredicting glucose/metabolic outcomes at 12 months postpartum, adjusted for the indicated covariates in each model

Outcomes at 12 months

Vitamin D/PTH groups

Vitamin D sufficient Vitamin D deficient/insufficient

PTH 1st/2nd tertile(25-OH-D $75nmol/L and PTH#3.8 pmol/L)

PTH 3rd tertile(25-OH-D $75nmol/L and PTH.3.8 pmol/L)

PTH 1st/2ndtertile (25-OH-D,75 nmol/L andPTH #3.8 pmol/L)

PTH 3rd tertile(25-OH-D ,75nmol/L and PTH.3.8 pmol/L)

Estimate P Estimate P Estimate P Estimate P

Matsuda indexModel 1: age, ethnicity, family history

of T2DM, previous GDM, BMI,Matsuda index at 3 months Ref. — 20.105 0.29 20.0289 0.58 20.151 0.01

Model 2: model 1 plus duration ofbreast-feeding Ref. — 20.104 0.30 20.0280 0.59 20.154 0.01

Model 3: model 2 plus physicalactivity and season Ref. — 20.089 0.38 20.0188 0.72 20.156 0.01

ISSI-2Model 1: age, ethnicity, family history

of T2DM, previous GDM, BMI,ISSI-2 at 3 months Ref. — 0.003 0.96 20.069 0.09 20.126 0.008

Model 2: model 1 plus duration ofbreast-feeding Ref. — 0.003 0.96 20.070 0.09 20.126 0.008

Model 3: model 2 plus physicalactivity and season Ref. — 0.003 0.96 20.073 0.08 20.128 0.008

Fasting glucoseModel 1: age, ethnicity, family history

of T2DM, previous GDM, BMI,fasting glucose at 3 months Ref. — 0.0030 0.86 0.015 0.10 0.020 0.05

Model 2: model 1 plus duration ofbreast-feeding Ref. — 0.0031 0.85 0.015 0.10 0.020 0.05

Model 3: model 2 plus physicalactivity and season Ref. — 0.0034 0.84 0.015 0.09 0.023 0.03

2-h glucoseModel 1: age, ethnicity, family history

of T2DM, previous GDM, BMI, 2-hglucose at 3 months Ref. — 20.026 0.58 0.038 0.12 0.091 0.002

Model 2: model 1 plus duration ofbreast-feeding Ref. — 20.026 0.58 0.038 0.13 0.091 0.002

Model 3: model 2 plus physicalactivity and season Ref. — 20.027 0.58 0.035 0.17 0.094 0.002

The reference group is vitamin D sufficient with PTH in 1st/2nd tertile at 3 months postpartum. Boldface type indicates P , 0.05.Outcome variables are log transformed.

3874 Vitamin D/PTH Axis and Glucose Metabolism Diabetes Volume 63, November 2014

and PTH status at 3 months postpartum: 1) vitamin Dsufficient and PTH in the 1st/2nd tertile (referencegroup) (n = 130), 2) vitamin D sufficient and PTH inthe 3rd tertile (n = 25), 3) vitamin D deficient/insufficientand PTH in the 1st/2nd tertile (n = 212), and 4) vitamin Ddeficient/insufficient and PTH in the 3rd tertile (n = 127).To test for a biological interaction, we investigated whetherthe combined impact of vitamin D deficiency/insufficiencyand PTH in the 3rd tertile on the outcomes of 1) dysgly-cemia, 2) lowest tertile of Matsuda index, and 3) lowesttertile of ISSI-2 at 12 months postpartum exceeded thesum of the individual effects of these conditions.

With respect to dysglycemia at 12 months postpartum,the presence of vitamin D deficiency/insufficiency andPTH in the 3rd tertile alone conferred increments in therisk for dysglycemia of 48.4% and 71.4%, respectively,compared with the reference group (vitamin D sufficiencyand PTH in the 1st/2nd tertile). Interestingly, when bothconditions were present, there was a 203.2% increasedrisk for dysglycemia, which resulted in RERI = 83.4%,AP = 27.5% and S = 1.69 (Fig. 2A). In other words, thecombined effect of vitamin D deficiency/insufficiency andPTH in the 3rd tertile conferred an excess risk of 83.4%beyond the sum of the individual effects. Furthermore, asindicated by the AP parameter, the effect of the interac-tion accounts for 27.5% of the dysglycemia among indi-viduals with both vitamin D deficiency/insufficiency andPTH in the 3rd tertile.

Similarly, for the outcome of being in the lowest tertileof Matsuda index at 12 months postpartum, the presenceof vitamin D deficiency/insufficiency and PTH in the 3rdtertile alone conferred excess risks of 109.8% and 21.3%,respectively, while the combined effect conferred anincreased risk of 360.9% (RERI = 229.8%, AP = 49.8%,and S = 2.75) (Fig. 2B). Lastly, for the lowest tertile ofISSI-2, the presence of vitamin D deficiency/insufficiencyand PTH in the 3rd tertile alone conferred risk incrementsof 90.8% and 25%, respectively, while the combined ef-fect conferred an increased risk of 152% (RERI = 66.2%,AP = 26.3%, and S = 1.77) (Fig. 2C).

Vitamin D and PTH Status Combined: LongitudinalCovariate-Adjusted AnalysesFigure 3 shows the adjusted estimates at 3 and 12 monthspostpartum for insulin sensitivity, b-cell function, fastingglucose, and 2-h glucose in the four vitamin D/PTHgroups after adjustment for the diabetes mellitus riskfactors age, ethnicity, family history of diabetes mellitus,previous GDM, and BMI. Notably, vitamin D deficiency/insufficiency with PTH in the 3rd tertile was associatedwith lower Matsuda index and ISSI-2 and increased fast-ing glucose and 2-h glucose. In addition, this groupexhibited a pronounced deterioration in ISSI-2 and 2-hglucose between 3 and 12 months postpartum that wasnot observed in the other three groups.

To further explore the differential changes in metabolicoutcomes between the four groups defined by vitamin D

and PTH together, we evaluated the adjusted percentagechange in insulin sensitivity, b-cell function, and glycemiain each of these groups between 3 and 12 months post-partum. As shown in Fig. 4, only the group with vitamin Ddeficiency/insufficiency and PTH in the 3rd tertile hada significant decline in ISSI-2 (24.7 6 3.4% vs. 7.6 64.4%, P = 0.04) and an increase in 2-h glucose (6.6 62.6% vs. 20.4 6 2.8%, P = 0.05) compared with thereference group. Of note, none of the other vitaminD/PTH groups showed changes in these metabolic out-comes compared with the reference group (all P . 0.15).

Finally, we performed multiple linear regression anal-yses to evaluate the longitudinal associations betweencombined vitamin D/PTH status at 3 months postpartumand metabolic outcomes after full covariate adjustment.As shown in Table 3, vitamin D deficiency/insufficiencyaccompanied by PTH in the 3rd tertile was the onlygroup that independently predicted lower insulin sensi-tivity (P = 0.01) and b-cell function (P = 0.008) andincreased fasting glucose (P = 0.03) and 2-h glucose(P = 0.002) at 12 months postpartum compared withthe reference group (Table 3). Of note, vitamin D de-ficiency/insufficiency with PTH in the 1st/2nd tertiledid not independently predict any of these metabolicoutcomes. In sensitivity analyses, these results were un-changed with further adjustment for use of vitamin D/calcium supplements, smoking status, and change inBMI between 3 and 12 months postpartum (data notshown). In addition, none of the results changed afterexcluding the 19 participants with PTH above the labo-ratory normal range, with the exception of fasting glu-cose with which vitamin D deficiency/insufficiency andPTH in the 3rd tertile were now associated at borderlinesignificance (P = 0.06) (data not shown).

DISCUSSION

In this study, we show that the combination of vitamin Ddeficiency/insufficiency and increased PTH is an indepen-dent predictor of deterioration in insulin sensitivity,b-cell function, and glycemia in a cohort of women inthe 1st year postpartum. Notably, vitamin D deficiency/insufficiency with higher PTH conferred an excess risk forfuture dysglycemia, decreased insulin sensitivity, andpoorer b-cell function that exceeded the sum of the in-dividual risks associated with these conditions. Whileboth vitamin D and PTH alone were independently asso-ciated with some of the outcomes, only the coupling ofvitamin D deficiency/insufficiency with higher PTH wasconsistently associated with declining insulin sensitivityand b-cell function and rising glycemia over time.

Previous studies have suggested an association of inte-grated assessment of vitamin D/PTH with glucose me-tabolism (41,42). Specifically, in a cross-sectional analysisevaluating 15 obese girls and 15 matched control subjects,Stanley et al. (41) demonstrated that the ratio of PTH tovitamin D was associated with insulin sensitivity andhsCRP. These results were confirmed by another study

diabetes.diabetesjournals.org Kramer and Associates 3875

of 133 obese adolescents that showed an association ofthe PTH–to–vitamin D ratio with presence of metabolicsyndrome (42). Our study further extends this concept bydemonstrating an interaction between vitamin D defi-ciency/insufficiency and PTH on longitudinal changes inglucose metabolism in a prospective cohort. This novelanalysis has two key implications. First, it supports aneffect of vitamin D status on glucose homeostasis bydemonstrating an independent association of the PTH–vitamin D axis with the development of hyperglycemiathrough the worsening of insulin sensitivity and b-cellfunction. Second and most importantly, these data high-light the need for assessment of the entire PTH–vitaminD axis when studying the effect of vitamin D on glucosemetabolism. Specifically, by showing that vitamin D de-ficiency/insufficiency has a differential association withglucose metabolism and its underlying physiology depend-ing on the concurrent level of PTH, our study suggestsa new perspective for future clinical trials of vitamin Dsupplementation.

Previous studies evaluating the association of vitaminD and glucose metabolism have yielded inconsistentresults. Although lower 25-OH-D has been associatedwith incident T2DM (7–9,43) and decline in b-cell func-tion over time (10,11) in several observational studies,other investigators have noted conflicting results(13,15,16,18). In addition, interventional studies aimingto evaluate the effect of vitamin D supplementation onglucose metabolism have yielded inconsistent conclusions(19–21,44,45). Physiologic studies have found either noimpact of vitamin D supplementation on clamp-derivedinsulin secretion (21) or an effect restricted to first-phaseinsulin secretion (44). In the same way, the results ofclinical trials have not been consistent or robust in sup-porting a role for vitamin D supplementation in the pre-vention of T2DM (46). In a randomized trial of 71 obesemen, vitamin D supplementation improved postprandialinsulin sensitivity but had no effect on insulin secretionor hepatic insulin resistance (20). Moreover, in a study of92 adults at risk for diabetes mellitus, supplementation

Figure 3—Adjusted mean levels for Matsuda index (A), ISSI-2 (B), fasting glucose (C), and 2-h glucose (D) in each of the four vitamin D/PTHgroups at 3 and 12 months postpartum, adjusted for age, ethnicity, family history of diabetes mellitus, previous GDM, and BMI at 3 months.○, vitamin D sufficiency and PTH 1st/2nd tertile;■, vitamin D sufficiency and PTH 3rd tertile;◇, vitamin D deficiency/insufficiency and PTH1st/2nd tertile; ▼, vitamin D deficiency/insufficiency and PTH 3rd tertile.

3876 Vitamin D/PTH Axis and Glucose Metabolism Diabetes Volume 63, November 2014

with vitamin D yielded a slight improvement in b-cell func-tion and a marginal effect on glycemic control (45). Thereare several possible reasons for these inconsistencies suchas differences in study populations, diverse methodologiesin the assessment of b-cell function and insulin sensitivity,and the potential impact of confounders (such as obesityand outdoor physical activity). However, the results of thecurrent study suggest that the lack of careful documenta-tion of the PTH–vitamin D axis may in part explain thecontradictory findings of these previous reports.

Our analyses demonstrate that vitamin D deficiency/insufficiency of sufficient biologic impact to result in anincrease in PTH was independently associated withdysglycemia, declining insulin sensitivity, and b-cell dys-function. These data raise the possibility that previousrandomized controlled trials failed to detect a robusteffect of vitamin D supplementation on glucose metab-olism because they were not performed in the specificpatient population that would most benefit from thisintervention.

Our findings are supported by biological factors sug-gesting a link between vitamin D status and glucosehomeostasis. First, vitamin D receptors are expressed inpancreatic b-cells and target tissues for insulin action suchas skeletal muscle and adipose tissue (9). Second, vitaminD receptor polymorphisms impact insulin secretion and

sensitivity in humans (47,48). Finally, the active form ofvitamin D, calcitriol, has an effect in modulating calciuminflux and gene expression in b-cells (49). Most notably, ithas been reported that calcitriol does not impact insulinrelease when pancreatic islets are under normal condi-tions, instead requiring a stressed environment such asexposure to pathologic cytokines or vitamin D deficiencyfor the detection of its effect (49). Thus, it would appearthat low 25-OH-D coupled with increased PTH may be abetter indicator reflecting the state of vitamin D deficiency/insufficiency that leads to dysregulation of glucosehomeostasis.

Our study is robust, as it was performed in a well-characterized cohort undergoing serial metabolic evalu-ation at a time (first year postpartum when women maybe breast-feeding and often indoors to limit infant sunexposure) and geographic location (northern latitude)that may contribute to an increased likelihood of vitaminD deficiency/insufficiency. Accordingly, this cohort pro-vides a unique opportunity to study the impact of PTH–vitamin D status on glucose metabolism. Furthermore,to our knowledge, this is the first study to prospectivelyevaluate both vitamin D and PTH together in relation toglucose homeostasis and its underlying physiology. Apossible limitation of this study is that we did not mea-sure 1,25-dihydroxyvitamin D and vitamin D–binding

Figure 4—Percentage change in Matsuda index (A), ISSI-2 (B), fasting glucose (C), and 2-h glucose (D) in each of the four vitamin D/PTHgroups between 3 and 12 months postpartum, adjusted for age, ethnicity, family history of diabetes mellitus, previous GDM, and BMI at 3months. **P # 0.05 for comparison with reference group (vitamin D sufficient with PTH in 1st/2nd tertile). Def./insuf., deficiency/insufficiency;Suf., sufficiency.

diabetes.diabetesjournals.org Kramer and Associates 3877

protein, which could provide a more comprehensive eval-uation of this pathway. However, 25-OH-D and PTH arethe standard clinical measures for assessment of vitaminD status in practice. In addition, because vitamin D andPTH are associated with adiposity, it is reasonable toconsider the possibility that the impact of the vitaminD/PTH axis on glucose metabolism is due to its associa-tion with obesity. However, it should be noted that allmodels were adjusted for BMI and that change in BMIwas further evaluated in sensitivity analyses, which sup-ported an independent association between the vitaminD/PTH axis and glucose homeostasis. Another consider-ation is that the study population consisted of youngwomen in the postpartum period, such that the resultsshould be confirmed in large population-based studiesoutside this setting. Lastly, b-cell function and insulinsensitivity were assessed with surrogate indices, ratherthan more time-consuming and cumbersome clamp stud-ies, which would have been difficult to implement in 494new mothers on two occasions in the first year afterdelivery. Moreover, ISSI-2 and Matsuda index are vali-dated measures that have been widely used in previousstudies (10,11,23,32–35).

In conclusion, we demonstrate that PTH status needsto be considered when evaluating the associations ofvitamin D status with glucose homeostasis and itsunderlying determinants. Specifically, vitamin D defi-ciency/insufficiency with increased PTH is independentlyassociated with deterioration in insulin sensitivity,b-cell function, and glycemia over time, which was notobserved in women with vitamin D deficiency/insuffi-ciency in conjunction with lower PTH. This concept pro-vides novel pathophysiologic insight relevant to theimpact of vitamin D status on T2DM and warrants fur-ther evaluation in future clinical trials of vitamin Dsupplementation.

Funding. This study was supported by operating grants MOP-84206 andMHC-115442 from the Canadian Institutes of Health Research and OG-3-11-3300-RR from the Canadian Diabetes Association (CDA). C.K.K. holds a CDAPostdoctoral Fellowship Award. R.R. holds an Ontario Ministry of Research andInnovation Early Researcher Award.Duality of Interest. No potential conflicts of interest relevant to this articlewere reported.Author Contributions. C.K.K. wrote the first draft of the manuscript,contributed to statistical analysis, contributed to analysis and interpretation ofdata and revision of the manuscript for intellectual content, and gave approval forsubmission. B.S. contributed to statistical analysis, contributed to analysis andinterpretation of data and revision of the manuscript for intellectual content, andgave approval for submission. A.J.H., P.W.C., M.S., and B.Z. contributed to studyconception and design, contributed to analysis and interpretation of data andrevision of the manuscript for intellectual content, and gave approval for sub-mission. R.R. contributed to study conception and design, contributed to statis-tical analysis, contributed to analysis and interpretation of data and revision of themanuscript for intellectual content, and gave approval for submission. R.R. is theguarantor of this work and, as such, had full access to all the data in the study andtakes responsibility for the integrity of the data and the accuracy of the dataanalysis.

References1. Pittas AG, Chung M, Trikalinos T, et al. Systematic review: Vitamin D andcardiometabolic outcomes. Ann Intern Med 2010;152:307–3142. Holick MF. Vitamin D deficiency. N Engl J Med 2007;357:266–2813. Wang TJ, Pencina MJ, Booth SL, et al. Vitamin D deficiency and risk ofcardiovascular disease. Circulation 2008;117:503–5114. Kayaniyil S, Vieth R, Harris SB, et al. Association of 25(OH)D and PTH withmetabolic syndrome and its traditional and nontraditional components. J ClinEndocrinol Metab 2011;96:168–1755. Ahlström T, Hagström E, Larsson A, Rudberg C, Lind L, Hellman P. Cor-relation between plasma calcium, parathyroid hormone (PTH) and the metabolicsyndrome (MetS) in a community-based cohort of men and women. Clin Endo-crinol (Oxf) 2009;71:673–6786. Bhatt SP, Nigam P, Misra A, Guleria R, Qadar Pasha MA. Independentassociations of low 25 hydroxy vitamin D and high parathyroid hormonal levelswith nonalcoholic fatty liver disease in Asian Indians residing in north India.Atherosclerosis 2013;230:157–1637. Song Y, Wang L, Pittas AG, et al. Blood 25-hydroxy vitamin D levels andincident type 2 diabetes: a meta-analysis of prospective studies. Diabetes Care2013;36:1422–14288. Mitri J, Muraru MD, Pittas AG. Vitamin D and type 2 diabetes: a systematicreview. Eur J Clin Nutr 2011;65:1005–10159. Stivelman E, Retnakaran R. Role of vitamin D in the pathophysiology andtreatment of type 2 diabetes. Curr Diabetes Rev 2012;8:42–4710. Kayaniyil S, Retnakaran R, Harris SB, et al. Prospective associations ofvitamin D with b-cell function and glycemia: the PROspective Metabolism andISlet cell Evaluation (PROMISE) cohort study. Diabetes 2011;60:2947–295311. Kayaniyil S, Vieth R, Retnakaran R, et al. Association of vitamin D withinsulin resistance and beta-cell dysfunction in subjects at risk for type 2 diabetes.Diabetes Care 2010;33:1379–138112. Bellia A, Marinoni G, D’Adamo M, et al. Parathyroid hormone and insulinresistance in distinct phenotypes of severe obesity: a cross-sectional analysis inmiddle-aged men and premenopausal women. J Clin Endocrinol Metab 2012;97:4724–473213. Reis JP, von Mühlen D, Kritz-Silverstein D, Wingard DL, Barrett-Connor E.Vitamin D, parathyroid hormone levels, and the prevalence of metabolic syn-drome in community-dwelling older adults. Diabetes Care 2007;30:1549–155514. Procopio M, Borretta G. Derangement of glucose metabolism in hyper-parathyroidism. J Endocrinol Invest 2003;26:1136–114215. Mattila C, Knekt P, Männistö S, et al. Serum 25-hydroxyvitamin D concen-tration and subsequent risk of type 2 diabetes. Diabetes Care 2007;30:2569–257016. Grimnes G, Emaus N, Joakimsen RM, et al. Baseline serum 25-hydroxyvitamin D concentrations in the Tromsø Study 1994-95 and risk ofdeveloping type 2 diabetes mellitus during 11 years of follow-up. Diabet Med2010;27:1107–111517. Schafer AL, Napoli N, Lui L, Schwartz AV, Black DM; Study of OsteoporoticFractures. Serum 25-hydroxyvitamin D concentration does not independentlypredict incident diabetes in older women. Diabet Med 2014;31:564–56918. de las Heras J, Rajakumar K, Lee S, Bacha F, Holick MF, Arslanian SA. 25-Hydroxyvitamin D in obese youth across the spectrum of glucose tolerance fromnormal to prediabetes to type 2 diabetes. Diabetes Care 2013;36:2048–205319. Jorde R, Sneve M, Torjesen P, Figenschau Y. No improvement in cardio-vascular risk factors in overweight and obese subjects after supplementation withvitamin D3 for 1 year. J Intern Med 2010;267:462–47220. Nagpal J, Pande JN, Bhartia A. A double-blind, randomized, placebo-controlled trial of the short-term effect of vitamin D3 supplementation on insulinsensitivity in apparently healthy, middle-aged, centrally obese men. Diabet Med2009;26:19–2721. Lind L, Pollare T, Hvarfner A, Lithell H, Sørensen OH, Ljunghall S. Long-termtreatment with active vitamin D (alphacalcidol) in middle-aged men with impairedglucose tolerance. Effects on insulin secretion and sensitivity, glucose toleranceand blood pressure. Diabetes Res 1989;11:141–147

3878 Vitamin D/PTH Axis and Glucose Metabolism Diabetes Volume 63, November 2014

22. Saliba W, Barnett O, Rennert HS, Lavi I, Rennert G. The relationship betweenserum 25(OH)D and parathyroid hormone levels. Am J Med 2011;124:1165–117023. Retnakaran R, Qi Y, Sermer M, Connelly PW, Hanley AJ, Zinman B. Beta-cellfunction declines within the first year postpartum in women with recent glucoseintolerance in pregnancy. Diabetes Care 2010;33:1798–180424. Retnakaran R, Qi Y, Sermer M, Connelly PW, Hanley AJ, Zinman B. Glucoseintolerance in pregnancy and future risk of pre-diabetes or diabetes. DiabetesCare 2008;31:2026–203125. Kew S, Ye C, Hanley AJ, et al. Cardiometabolic implications of postpartumweight changes in the first year after delivery. Diabetes Care 2014;37:1998–200626. Retnakaran R, Qi Y, Ye C, et al. Hepatic insulin resistance is an early de-terminant of declining b-cell function in the first year postpartum after glucoseintolerance in pregnancy. Diabetes Care 2011;34:2431–243427. National Diabetes Data Group. Classification and diagnosis of diabetes mel-litus and other categories of glucose intolerance. Diabetes 1979;28:1039–105728. Canadian Diabetes Association Clinical Practice Guidelines Expert Com-mittee. Canadian Diabetes Association 2013 clinical practice guidelines for theprevention and management of diabetes in Canada. Can J Diabetes 2013;37(Suppl. 1):S8–S1129. Matsuda M, DeFronzo RA. Insulin sensitivity indices obtained from oralglucose tolerance testing: comparison with the euglycemic insulin clamp. Di-abetes Care 1999;22:1462–147030. Retnakaran R, Shen S, Hanley AJ, Vuksan V, Hamilton JK, Zinman B. Hy-perbolic relationship between insulin secretion and sensitivity on oral glucosetolerance test. Obesity (Silver Spring) 2008;16:1901–190731. Retnakaran R, Qi Y, Goran MI, Hamilton JK. Evaluation of proposed oraldisposition index measures in relation to the actual disposition index. Diabet Med2009;26:1198–120332. O’Gorman CS, Syme C, Lang J, Bradley TJ, Wells GD, Hamilton JK. Anevaluation of early cardiometabolic risk factors in children and adolescents withTurner syndrome. Clin Endocrinol (Oxf) 2013;78:907–91333. Zinman B, Harris SB, Neuman J, et al. Low-dose combination therapy withrosiglitazone and metformin to prevent type 2 diabetes mellitus (CANOE trial):a double-blind randomised controlled study. Lancet 2010;376:103–11134. Kramer CK, Choi H, Zinman B, Retnakaran R. Determinants of reversibility ofb-cell dysfunction in response to short-term intensive insulin therapy in patientswith early type 2 diabetes. Am J Physiol Endocrinol Metab 2013;305:E1398–E140735. Kramer CK, Choi H, Zinman B, Retnakaran R. Glycemic variability in patientswith early type 2 diabetes: the impact of improvement in b-cell function. Di-abetes Care 2014;37:1116–1123

36. Holick MF, Binkley NC, Bischoff-Ferrari HA, et al.; Endocrine Society.Evaluation, treatment, and prevention of vitamin D deficiency: an EndocrineSociety clinical practice guideline. J Clin Endocrinol Metab 2011;96:1911–193037. Rothman KJ. Epidemiology. An Introduction. New York, Oxford UniversityPress, 2002, p. 198–21038. Andersson T, Alfredsson L, Källberg H, Zdravkovic S, Ahlbom A. Calculatingmeasures of biological interaction. Eur J Epidemiol 2005;20:575–57939. Hallqvist J, Ahlbom A, Diderichsen F, Reuterwall C. How to evaluate in-teraction between causes: a review of practices in cardiovascular epidemiology. JIntern Med 1996;239:377–38240. Ahlbom A, Alfredsson L. Interaction: A word with two meanings createsconfusion. Eur J Epidemiol 2005;20:563–56441. Stanley T, Bredella MA, Pierce L, Misra M. The ratio of parathyroid hormoneto vitamin D is a determinant of cardiovascular risk and insulin sensitivity inadolescent girls. Metab Syndr Relat Disord 2013;11:56–6242. Alemzadeh R, Kichler J. Parathyroid hormone is associated with biomarkersof insulin resistance and inflammation, independent of vitamin D status, in obeseadolescents. Metab Syndr Relat Disord 2012;10:422–42943. Forouhi NG, Ye Z, Rickard AP, et al. Circulating 25-hydroxyvitamin D con-centration and the risk of type 2 diabetes: results from the European ProspectiveInvestigation into Cancer (EPIC)-Norfolk cohort and updated meta-analysis ofprospective studies. Diabetologia 2012;55:2173–218244. Borissova AM, Tankova T, Kirilov G, Dakovska L, Kovacheva R. The effect ofvitamin D3 on insulin secretion and peripheral insulin sensitivity in type 2 diabeticpatients. Int J Clin Pract 2003;57:258–26145. Mitri J, Dawson-Hughes B, Hu FB, Pittas AG. Effects of vitamin D andcalcium supplementation on pancreatic b cell function, insulin sensitivity, andglycemia in adults at high risk of diabetes: the Calcium and Vitamin D forDiabetes Mellitus (CaDDM) randomized controlled trial. Am J Clin Nutr 2011;94:486–49446. Shapses SA, Manson JE. Vitamin D and prevention of cardiovascular dis-ease and diabetes: why the evidence falls short. JAMA 2011;305:2565–256647. Hitman GA, Mannan N, McDermott MF, et al. Vitamin D receptor genepolymorphisms influence insulin secretion in Bangladeshi Asians. Diabetes 1998;47:688–69048. Oh JY, Barrett-Connor E. Association between vitamin D receptor poly-morphism and type 2 diabetes or metabolic syndrome in community-dwellingolder adults: the Rancho Bernardo Study. Metabolism 2002;51:356–35949. Wolden-Kirk H, Overbergh L, Gysemans C, et al. Unraveling the effects of1,25OH2D3 on global gene expression in pancreatic islets. J Steroid Biochem MolBiol 2013;136:68–79

diabetes.diabetesjournals.org Kramer and Associates 3879