Ravi Retnakaran, Ying Qi, Philip W. Connelly, Mathew Sermer, Anthony J. Hanley, and Bernard Zinman

Transcription

1 ORIGINAL ARTICLE Endocrine Care The Graded Relationship between Glucose Tolerance Status in Pregnancy and Postpartum Levels of Low-Density-Lipoprotein Cholesterol and Apolipoprotein B in Young Women: Implications for Future Cardiovascular Risk Ravi Retnakaran, Ying Qi, Philip W. Connelly, Mathew Sermer, Anthony J. Hanley, and Bernard Zinman Leadership Sinai Centre for Diabetes (R.R., Y.Q., A.J.H., B.Z.) and Division of Obstetrics and Gynecology (M.S.), Mount Sinai Hospital, Toronto, Canada M5T 3L9; Division of Endocrinology (R.R., P.W.C., A.J.H., B.Z.) and Department of Nutritional Sciences (A.J.H.), University of Toronto, Toronto, Canada M5S 3E2; Keenan Research Centre in the Li Ka Shing Knowledge Institute (P.W.C.), St. Michael s Hospital, Toronto, Canada M5B 1W8 Context and Objective: Both gestational diabetes mellitus (GDM) and gestational impaired glucose tolerance (GIGT) identify women at risk of future cardiovascular disease, although the mediators of this risk are unknown. Because lipid factors can contribute to cardiovascular risk, we sought to characterize the relationship between gestational glucose tolerance status and lipid profile in pregnancy and the postpartum. Design, Setting, and Participants: Fasting lipids were measured in 482 women in pregnancy and at 3 months postpartum. Antepartum glucose challenge test (GCT) and oral glucose tolerance test (OGTT) defined four gestational glucose tolerance groups: GDM (n 136), GIGT (n 89), abnormal GCT with normal glucose tolerance (NGT) on OGTT (abnormal GCT NGT) (n 170), and normal GCT with NGT on OGTT (normal GCT NGT) (n 87). Results: In pregnancy, there were no significant differences between the groups in total and low-density lipoprotein (LDL) cholesterol, triglycerides, total cholesterol to high-density lipoprotein (HDL) cholesterol ratio, apolipoprotein B (apob), apolipoprotein A1 (apoa1), and apob to apoa1 ratio. At 3 months postpartum, however, each of the following lipid parameters progressively increased from normal GCT NGT to abnormal GCT NGT to GIGT to GDM: total cholesterol (P ), LDL (P ), triglycerides (P ), total cholesterol to HDL ratio (P ), apob (P ), and apob to apoa1 ratio (P ). Furthermore, on multiple linear regression analyses, GDM emerged as an independent predictor of postpartum total cholesterol (t 3.09, P ), LDL (t 3.81, P ), triglycerides (t 3.38, P ), total cholesterol to HDL ratio (t 3.76,P ), apob (t 4.12, P ), and apob to apoa1 ratio (t 3.07, P ). GIGT was an independent predictor of postpartum total cholesterol to HDL ratio (t 2.27, P ), apob (t 2.04, P ), and apob to apoa1 ratio (t 1.97, P 0.049). Conclusions: Compared with their peers, women with GDM and GIGT have a more atherogenic lipid profile by 3 months postpartum, characterized by increased LDL and apob. (J Clin Endocrinol Metab 95: , 2010) ISSN Print X ISSN Online Printed in U.S.A. Copyright 2010 by The Endocrine Society doi: /jc Received February 17, Accepted May 28, First Published Online July 14, 2010 Abbreviations: apoa1, Apolipoprotein A1; apob, apolipoprotein B; AUC gluc, area under the glucose curve; BMI, body mass index; CVD, cardiovascular disease; GCT, glucose challenge test; GDM, gestational diabetes mellitus; GIGT, gestational impaired glucose tolerance; HDL, high-density lipoprotein; LDL, low-density lipoprotein; NGT, normal glucose tolerance; OGTT, oral glucose tolerance test; T2DM, type 2 diabetes mellitus. J Clin Endocrinol Metab, September 2010, 95(9): jcem.endojournals.org 4345

2 4346 Retnakaran et al. Gestational Glucose Intolerance and Lipids J Clin Endocrinol Metab, September 2010, 95(9): It has long been recognized that the diagnosis of gestational diabetes mellitus (GDM) identifies a population of young women who are at high risk of developing type 2 diabetes mellitus (T2DM) in the future (1 3). It is now known that, despite their relative youth, women with GDM also have a significantly increased risk of cardiovascular disease (CVD) compared with their peers that manifests within a median 11 yr after the index pregnancy (adjusted hazard ratio 1.71) (4). Recently it has emerged that, by 12 yr postpartum, even women with mild glucose intolerance in pregnancy have an increased risk of CVD, less than that of women with GDM but significantly higher than that of their peers who maintain normal glucose tolerance in pregnancy (5). Thus, as it does for long-term diabetic risk (6 9), the spectrum of gestational glucose tolerance (i.e. from normal to mild glucose intolerance to GDM) appears to stratify young women with respect to their future risk of CVD (5, 10). CVD represents the clinical manifestation of a chronic pathological process (atherosclerosis) that develops over decades. As such, although gestational dysglycemia predicts future T2DM, it is unlikely that the postpartum development of diabetes accounts for CVD detected yr after a pregnancy complicated by GDM or gestational glucose intolerance (particularly considering the long period of time required for the development of macrovascular disease in patients with T2DM) (5). Indeed, considering the length of time over which atherosclerotic disease develops, the vascular risk factors contributing to the gradient of future CVD risk associated with gestational dysglycemia are likely to be longstanding and present at the time of the pregnancy or shortly thereafter. In this context, we hypothesized that unrecognized differences in lipid factors between gestational glucose tolerance groups may be relevant to this risk gradient. Thus, in this study, we sought to systematically characterize the relationship between gestational dysglycemia and lipid profile, in both pregnancy and the postpartum, in a cohort of women representing the full spectrum of gestational glucose tolerance and hence a broad range of future cardiovascular risk. Subjects and Methods This analysis was conducted in the setting of an ongoing prospective observational study in which a cohort of women recruited at the time of antepartum GDM screening was undergoing longitudinal cardiometabolic characterization in pregnancy and the postpartum. The study protocol has been described in detail previously (6, 7, 11). In brief, at our institution, all pregnant women are screened for GDM by 50 g glucose challenge test (GCT) in the late second trimester, followed by referral for a diagnostic oral glucose tolerance test (OGTT) if the screening test is abnormal (1 h plasma glucose 7.8 mmol/liter). In the current study, regardless of the screening result, all participants underwent a 3-h, 100-g OGTT for ascertainment of gestational glucose tolerance status. Participants were recruited either before or after the glucose challenge test but always before the OGTT. The recruitment of women after an abnormal GCT served to enrich the study population for women with varying degrees of antepartum glucose intolerance, as previously noted (6). At 3 months postpartum, participants returned to the clinical investigation unit for assessment. The study protocol was approved by the Mount Sinai Hospital Research Ethics Board and all participants gave written informed consent. Study protocol On the morning of the OGTT in pregnancy, data regarding medical and family history were collected by an intervieweradministered questionnaire. As described previously (6, 7), the antepartum GCT and OGTT stratified subjects into the following four gestational glucose tolerance groups: 1) GDM (defined by exceeding two or more National Diabetes Data Group (12) glycemic thresholds on the OGTT); 2) gestational impaired glucose tolerance (GIGT) (defined by exceeding only one National Diabetes Data Group threshold); 3) an abnormal GCT with normal glucose tolerance on the OGTT [abnormal GCT normal glucose tolerance (NGT)]; and 4) normal GCT NGT. Lipids in pregnancy were measured from fasting serum obtained during the antepartum OGTT in the late second or early third trimester, as described below. At 3 months postpartum, participants returned for reassessment, which included measurement of weight, waist, and blood pressure [measured twice 5 min apart by automatic sphygmomanometer (Dinamap Pro ; GE Healthcare, Waukesha, WI)]. Total cholesterol (CHOL2 reagent; Roche, Laval, Canada), high-density lipoprotein (HDL) cholesterol (Roche direct HDL reagent, HDLC3), and triglycerides (Roche TRIGL reagent) were measured from fasting serum collected at both the antepartum and postpartumstudyvisits, usingtherochecobas6000c501analyzer (Roche Diagnostics, Laval, Canada). Lipid measurements were standardized by the Centers for Disease Control and Prevention Lipid Standardization Program (Atlanta, GA). Low-density lipoprotein (LDL) cholesterol (calculated by Friedwald formula) and total cholesterol to HDL ratio were determined because current guidelines of the Canadian Cardiovascular Society recommend the assessment of these two measures for the targeting of treatment for dyslipidemia (13). Apolipoprotein B (apob) and apolipoprotein A1 (apoa1) were measured using the Siemens Healthcare Diagnostics BN ProSpec (Siemens Healthcare Diagnostics, Mississauga, Ontario, Canada). The apob to apoa1 ratio was determined because it has recently been proposed as the lipid measure with the highest population-attributable risk for myocardial infarction (14). Statistical analyses All analyses were conducted using the Statistical Analysis System (SAS, version 9.1; SAS Institute, Cary, NC). Continuous variables were tested for normality of distribution, and natural log transformations of skewed variables were used, where necessary, in subsequent analyses. Univariate differences across the four gestational glucose tolerance groups were assessed at the time of the OGTT in pregnancy and at 3 months postpartum using the Kruskal-Wallis test for continuous variables and the 2

3 J Clin Endocrinol Metab, September 2010, 95(9): jcem.endojournals.org 4347 TABLE 1. Characteristics of study subjects (n 482) stratified by glucose tolerance status in pregnancy test for categorical variables (Table 1 and Fig. 1). Univariate relationships between the area under the glucose curve (AUC gluc ) on the antepartum OGTT (calculated using the trapezoidal rule) and lipid parameters were assessed by Spearman correlation analysis (Table 2). Multiple linear regression analyses (Table 3) were performed to evaluate the relationships between glucose tolerance status in pregnancy and the following lipid parameters at 3 months postpartum (dependent variables for the respective models): total cholesterol (model A); LDL cholesterol (model B); HDL cholesterol (model C); triglycerides (model D); total cholesterol to HDL ratio (model E); apob (model F); and apob to apoa1 ratio (model G). Covariates that were included in all of these models were age, ethnicity, family history of diabetes, breast-feeding, waist circumference at 3 months postpartum, and gestational glucose tolerance status (with normal GCT NGT as the reference group). The models were also repeated with inclusion of body mass index (BMI) in place of waist circumference. Results Normal GCT NGT (n 87) Characteristics of study population in pregnancy and the postpartum Table 1 shows the characteristics of the study population (n 482) at baseline in pregnancy and at 3 months postpartum, stratified into the following four gestational glucose tolerance groups: 1) normal GCT NGT (n 87), 2) abnormal GCT NGT (n 170), 3) GIGT (n 89), and 4) GDM (n 136). There were no significant differences between the groups in age and ethnicity. The antepartum OGTT was performed slightly later in the normal GCT Abnormal GCT NGT (n 170) GIGT (n 89) GDM (n 136) P At OGTT in pregnancy Age (yr) 34.0 (4.4) 33.8 (4.3) 34.2 (4.3) 34.4 (4.3) Gestation (wk) 32 (31 34) 29 (28 30) 29 (28 31) 29 (28 31) Ethnicity Caucasian (%) Asian (%) Other (%) Family history of DM (%) Prepregnancy BMI (kg/m 2 ) 22.9 ( ) 23.5 ( ) 23.5 ( ) 25.0 ( ) Prepregnancy BMI category Normal (%) Overweight (%) Obese (%) At 3 months postpartum Postpartum (months) 3.2 ( ) 3.3 ( ) 3.3 ( ) 3.1 ( ) Current breast-feeding (%) Systolic blood pressure (mm Hg) 108 ( ) 109 ( ) 110 ( ) 111 ( ) Diastolic blood pressure (mm Hg) 67 (60 70) 64 (60 70) 64 (60 70) 66 (60 73) BMI (kg/m 2 ) 24.6 ( ) 25.8 ( ) 26.1 ( ) 26.6 ( ) Waist circumference (cm) 86.0 ( ) 85.2 ( ) 87.0 ( ) 88.6 ( ) Continuous variables are presented as median followed by interquartile range in parentheses (with the exception of age, which is presented as mean followed by SD in parentheses) and categorical variables are presented as percentages. P values refer to overall differences across groups as determined by the Kruskal-Wallis test for continuous variables or 2 test for categorical variables. For prepregnancy BMI categories, overweight is defined as BMI from 25 to kg/m 2 inclusive and obese is defined as BMI 30.0 kg/m 2 or greater. DM, Diabetes mellitus. NGT group (median 32 wk gestation) than in the other three groups (all median 29 wk) (overall P ). As gestational glucose tolerance status worsened, both prepregnancy BMI and the prevalence of a family history of diabetes increased (P and P , respectively), as expected for these known risk factors for GDM. At 3 months postpartum, the prevalence of breast-feeding was universally high and similar across the groups (P ). Systolic blood pressure increased from normal GCT NGT to abnormal GCT NGT to GIGT to GDM (P ). Both BMI and waist circumference at 3 months postpartum showed a similar pattern, although at borderline significance in both cases (P and P , respectively). Only 34 women were using oral contraceptives and just two women were receiving depomedroxyprogesterone acetate injections. The distribution of these 36 women was as follows: six in the normal GCT NGT group (7.2% of group), 18 in the abnormal GCT NGT group (10.8% of group), five in the GIGT group (5.8% of group), and seven in the GDM group (5.3% of group). Lipid profiles in pregnancy and in the postpartum by gestational glucose tolerance group Figure 1 shows the median and interquartile range for each component of the lipid profile within the four gestational glucose tolerance groups both in pregnancy (plot i of each panel) and at 3 months postpartum (plot

4 4348 Retnakaran et al. Gestational Glucose Intolerance and Lipids J Clin Endocrinol Metab, September 2010, 95(9): FIG. 1. Comparison of the following lipid parameters between gestational tolerance groups in pregnancy (plot i of each panel) and at 3 months postpartum (plot ii of each panel): total cholesterol (A); LDL (B); HDL (C); triglycerides (D); total cholesterol to HDL ratio (E); apob (F); apoa1 (G); and apob to apoa1 ratio (H). Median and interquartile range within each group are shown. ii of each panel) (these data are also provided in tabular form in the Supplemental Table 1, published on The Endocrine Society s Journals Online web site at jcem.endojournals.org). In pregnancy, there were no significant differences between the groups with respect to total cholesterol (P ), LDL cholesterol (P ), triglycerides (P ), total cholesterol to HDL ratio (P ), apob (P ), and apob to apoa1 ratio (P ) (left plots of Fig. 1, A, B, D, E, F, and H, respectively). Remarkably, however, each of these lipid parameters showed a very different pattern at 3 months postpartum.

5 J Clin Endocrinol Metab, September 2010, 95(9): jcem.endojournals.org 4349 FIG. 1. Continued. Indeed, the postpartum levels of each of the following measures progressively increased across the gestational glucose tolerance groups from normal GCT NGT to abnormal GCT NGT to GIGT to GDM: total cholesterol (overall P , linear trend P ), LDL cholesterol (overall P , linear trend P ), triglycerides (overall P , linear trend P ), total cholesterol to HDL ratio (overall P , linear trend P ), apob (overall P , linear trend P ), and apob to apoa1 ratio (overall P , linear trend P ) (right plots of Fig. 1, A, B, D, E, F, and H, respectively). Only HDL cholesterol (Fig. 1C) and apoa1 (Fig. 1G) exhibited similar patterns in pregnancy and at 3 months postpartum,

6 4350 Retnakaran et al. Gestational Glucose Intolerance and Lipids J Clin Endocrinol Metab, September 2010, 95(9): TABLE 2. Spearman correlations between AUC gluc in pregnancy and lipid values in pregnancy and at 3-months postpartum (i) unadjusted and (ii) adjusted for age, ethnicity, and family history of diabetes Unadjusted Adjusted r P r P Lipids in pregnancy Total cholesterol LDL HDL Triglycerides Total cholesterol to HDL ratio apob apoa apob to apoa1 ratio Lipids at 3 months postpartum Total cholesterol LDL 0.23 < < HDL 0.21 < < Triglycerides 0.26 < < Total cholesterol to 0.27 < < HDL ratio apob 0.23 < < apoa apob to apoa1 ratio 0.21 < < Bold values indicate P with HDL decreasing with worsening gestational glucose tolerance on both occasions (P and P , respectively) and apoa1 similar across the groups on both assessments (P and P , respectively). Thus, although HDL was the only lipid measure that differed between the groups in pregnancy, several differences emerged by 3 months postpartum, such that women with gestational dysglycemia now had a more atherogenic lipid profile than their peers. Correlation and multiple linear regression analyses To further study the relationships between gestational glucose tolerance status and lipid profile, we evaluated the univariate correlations between AUC gluc on the antepartum OGTT (i.e. a continuous variable corresponding to categorical glucose tolerance) and lipid measures (Table 2). This analysis revealed modest correlations between antepartum AUC gluc and both HDL (r 0.14, P 0.002) and triglycerides (r 0.11, P ) in pregnancy. Once again, however, several stronger correlations were noted between AUC gluc in pregnancy and postpartum lipid measures, including positive associations with LDL, triglycerides, total cholesterol to HDL ratio, apob, and apob to apoa1 ratio (all r 0.21, P ) and an inverse relationship with HDL (r 0.21, P ). Furthermore, all of these relationships were unchanged with adjustment for age, ethnicity, and a family history of diabetes (Table 2). We next performed multiple linear regression analyses to determine whether the relationships between categorical glucose tolerance status in pregnancy and lipid profile at 3 months postpartum were independent of covariates (Table 3). Indeed, after adjustment for age, ethnicity, family history of diabetes, breast-feeding, and waist circumference at 3 months postpartum, GDM emerged as an independent predictor of log-transformed total cholesterol (t 3.09, P ) (model A), LDL cholesterol (t 3.81, P ) (model B), triglycerides (t 3.38, P ) (model D), total cholesterol to HDL ratio (t 3.76, P ) (model E), apob (t 4.12, P ) (model F), and apob to apoa1 ratio (t 3.07, P ) (model G). Furthermore, GIGT also emerged as an independent predictor of log-transformed total cholesterol to HDL ratio (t 2.27, P ) (model E), apob (t 2.04, P ) (model F), and apob to apoa1 ratio (t 1.97, P 0.049) (model G). Both GDM (t 2.10, P ) and GIGT (t 2.14, P ) were negative independent predictors of log-transformed HDL cholesterol at 3 months postpartum (model C). Similar results were observed when all of these models were repeated with adjustment for BMI at 3 months postpartum in place of waist circumference (data not shown). In addition, the results were unchanged when all of the models were repeated with exclusion of the 34 women using oral contraceptives and the two women receiving depomedroxyprogesterone acetate injections (Supplemental Table 2). Thus, it emerges that GDM and GIGT both consistently predict a more atherogenic lipid profile at 3 months postpartum. Discussion In this report, we demonstrate that, at the time of GDM screening in late pregnancy, gestational glucose tolerance groups exhibit little difference in lipid profile, limited only to HDL cholesterol concentration. By 3 months postpartum, however, there are significant differences in total cholesterol, LDL, triglycerides, total cholesterol to HDL ratio, apob, and apob to apoa1 ratio, with each progressively increasing from the normal GCT NGT group to abnormal GCT NGT to GIGT to GDM. Furthermore, GDM emerged as an independent positive predictor of each of these lipid measures, whereas GIGT predicted increased total cholesterol to HDL, apob, and apob to apoa1 ratio. These data suggest that, compared with their peers, women with GDM and GIGT have a more atherogenic lipid profile by 3 months postpartum that may contribute to their increased risk of future CVD.

7 J Clin Endocrinol Metab, September 2010, 95(9): jcem.endojournals.org 4351 TABLE 3. Multiple linear regression analyses of the relationships between (exposure) gestational glucose tolerance status and (dependent variable) lipid parameters at 3 months postpartum Models and dependent variables Beta SE t P Model A (dependent variable: log total cholesterol at 3 months postpartum) GDM GIGT Abnormal GCT NGT Model B (dependent variable: log LDL cholesterol at 3 months postpartum) GDM GIGT Abnormal GCT NGT Model C (dependent variable: log HDL cholesterol at 3 months postpartum) GDM GIGT Abnormal GCT NGT Model D (dependent variable: log triglycerides at 3 months postpartum) GDM GIGT Abnormal GCT NGT Model E (dependent variable: log total cholesterol to HDL ratio at 3 months postpartum) GDM GIGT Abnormal GCT NGT Model F (dependent variable: log apob at 3 months postpartum) GDM < GIGT Abnormal GCT NGT Model G (dependent variable: log apob to apoa1 at 3 months postpartum) GDM GIGT Abnormal GCT NGT The dependent variables are as follows: model A, log total cholesterol; model B, log LDL; model C, log HDL; model D, log triglycerides; model E, log total cholesterol to HDL ratio; model F, log apob; and model G, log apob to apoa1. Each model is adjusted for age, ethnicity, family history of diabetes, breast-feeding, and waist circumference at 3 months postpartum. Bold values indicate P Previous studies reported conflicting findings regarding the lipid profile during pregnancy of women with GDM, with some reporting no differences compared with women with normal glucose tolerance (15 18) and others suggesting various abnormalities, including increased triglyceride, decreased HDL, and lower LDL concentrations (19 21). A consistent finding, however, observed again in this study, is that antepartum lipid levels (total cholesterol, LDL, HDL, triglycerides, apob, and apoa1) are increased compared with the postpartum in women with and without GDM, likely reflecting a physiological adaptation of lipid metabolism in the pregnant state (15 21). Because physiological up-regulation in pregnancy may mask underlying differences between the gestational glucose tolerance groups, the postpartum lipid assessment is more important in the current context, in which we seek to determine whether these groups exhibit differences in lipid profile that may be relevant to their gradient of future cardiovascular risk. Several studies have shown that women with a history of GDM have an increased risk of the metabolic syndrome within the first decade postpartum (22, 23), and we have shown that this risk exists as early as 3 months postpartum (11). As such, the independent associations observed in the current study between GDM and both decreased HDL and increased triglycerides at 3 months postpartum are not surprising (i.e. both of these lipid disorders being components of the metabolic syndrome). Of far greater interest, however, are the differences observed with respect to LDL and apob, which would not necessarily be anticipated on the basis of the relationship between GDM and the metabolic syndrome. Indeed, previous studies found no difference in LDL cholesterol between women with GDM and those with normal gestational glucose tolerance in the

8 4352 Retnakaran et al. Gestational Glucose Intolerance and Lipids J Clin Endocrinol Metab, September 2010, 95(9): first 3 yr postpartum (15, 21, 24), although increased LDL in women with a history of previous GDM has been reported at 5 6 yr postpartum (22, 25). Similarly, within the first year postpartum, apob has been reported to be similar between women with recent GDM and controls (although the largest of these studies involved only 20 women with GDM) (15, 21). Thus, in this context, the current study extends the literature by demonstrating that differences in each of LDL cholesterol concentration, apob, and apob to apoa1 ratio are indeed present as early as 3 months postpartum. Moreover, by evaluating these measures across the full spectrum of antepartum glucose tolerance, we demonstrate that early postpartum levels of LDL, apob, and apob to apoa1 ratio progressively increase with worsening gestational dysglycemia, thereby mirroring the gradient of future cardiovascular risk. As such, these data raise the possibility that LDL, apob, and apob to A1 ratio [all established predictors of CVD (13, 14)] may be relevant to future vascular risk in women with GDM and GIGT. The idea that LDL cholesterol may contribute to the risk of CVD in women with gestational dysglycemia is supported by certain other findings in GDM. Specifically, compared with their peers, women with GDM have reduced mean LDL particle size (19) and an increased preponderance of small, dense LDL particles, with an altered distribution of LDL subspecies characterized by upregulation of the very small LDL IVA and LDL IVB subclasses (18). Small dense LDL, a phenotypic feature associated with the atherogenic lipid profile of T2DM, is readily taken up by the vasculature and is particularly susceptible to oxidation in a pathological process that contributes to endothelial dysfunction and atherosclerosis (18). Indeed, it has been demonstrated that the lag phase in LDL oxidation is decreased in women with GDM during all three trimesters of pregnancy, indicative of increased LDL susceptibility to oxidation (26). These findings suggest that women with GDM have an atherogenic LDL particle phenotype. The current study now extends these data by demonstrating that, as early as 3 months postpartum (i.e. much earlier than previously reported), this patient population also has increased circulating levels of LDL cholesterol, thereby potentially amplifying the atherogenic capacity of its high-risk LDL phenotype over time. Furthermore, their presence at 3 months postpartum raises the possibility that the observed differences in LDL and apob may actually precede the pregnancy. In other words, gestational glucose intolerance potentially may enable the transient unmasking of a longstanding unrecognized gradient in LDL and apob concentration in young women. Whereas the absolute LDL concentrations observed in young women are not extremely elevated (as observed in Fig. 1B), long-term exposure to the combination of this LDL concentration gradient and enhanced oxidative susceptibility may work together over many years in determining long-term cardiovascular risk. Of note, a pathophysiological role for LDL concentration in mediating vascular risk in this setting is supported by the observation that, in women with a history of GDM, LDL lowering can improve endothelium-dependent vasodilation (27), a measure of endothelial function that predicts incident CVD (28). This observation thus underscores the key clinical significance of the current findings. Specifically, unlike LDL phenotype, both apob and LDL concentration are readily amenable to clinical monitoring and therapy and hence may provide the means for modifying cardiovascular risk in women with a history of GDM or GIGT. At present, further study is warranted to determine whether LDL and/or apob contribute to the vascular risk associated with gestational dysglycemia and whether women with such a history can benefit from increased surveillance and treatment. A limitation of this study is that lipid profile was not assessed before pregnancy, which would indicate whether the postpartum lipid gradient associated with gestational glucose tolerance status is present pregravid. Second, lipid factors such as LDL and apob have not been directly linked to cardiovascular events in this study because many years of follow-up will be needed before the presentation of CVD in a cohort of young women. Third, although significant, our findings are not of immediate clinical importance in that they do not hold specific implications for clinical management at this time. Nevertheless, the current study represents an important advance by demonstrating the presence of differences in LDL and apob concentration between gestational glucose tolerance groups as early as 3 months postpartum and should lead to further studies. In summary, in contrast to findings in pregnancy, there are significant differences in lipid profile between gestational glucose tolerance groups by 3 months postpartum. In particular, GDM is an independent positive predictor of postpartum total cholesterol, LDL, triglycerides, total cholesterol to HDL ratio, apob, and apob to apoa1 ratio, whereas GIGT predicts increased total cholesterol to HDL ratio, apob, and apob to apoa1 ratio. Of particular interest are these early postpartum relationships with LDL, apob, and apob to apoa1 ratio, all of which predict CVD. It thus emerges that, compared with their peers, women with GDM and GIGT have a more atherogenic lipid profile by 3 months postpartum that ultimately may contribute to their increased risk of future CVD.

9 J Clin Endocrinol Metab, September 2010, 95(9): jcem.endojournals.org 4353 Acknowledgments We thank the Mount Sinai Hospital Department of Pathology and Laboratory Medicine and Patient Care Services. Address all correspondence and requests for reprints to: Dr. Ravi Retnakaran, Leadership Sinai Centre for Diabetes, 60 Murray Street, Suite L5-039, Mailbox 21, Toronto, Ontario, Canada M5T 3L9. This work was supported by Operating Grants MOP and MOP from the Canadian Institutes of Health Research (CIHR), Grant OG RR from the Canadian Diabetes Association, and Grant NA 6747 from the Heart and Stroke Foundation of Ontario. R.R. holds a CIHR Clinical Research Initiative New Investigator Award, Canadian Diabetes Association Clinician-Scientist incentive funding, and University of Toronto Banting and Best Diabetes Centre New Investigator funding. A.J.H. holds a Tier II Canada Research Chair in Diabetes Epidemiology. B.Z. holds the Sam and Judy Pencer Family Chair in Diabetes Research at Mount Sinai Hospital and the University of Toronto. Disclosure Summary: The authors have nothing to disclose. References 1. Buchanan TA, Xiang AH 2005 Gestational diabetes mellitus. J Clin Invest 115: Bellamy L, Casas JP, Hingorani AD, Williams D 2009 Type 2 diabetes after gestational diabetes a systematic review and meta-analysis. Lancet 373: Feig DS, Zinman B, Wang X, Hux JE 2008 Risk of development of diabetes mellitus after diagnosis of gestational diabetes. CMAJ 179: Shah BR, Retnakaran R, Booth GL 2008 Increased risk of cardiovascular disease in young women following gestational diabetes mellitus. Diabetes Care 31: Retnakaran R, Shah BR 2009 Mild glucose intolerance in pregnancy and risk of cardiovascular disease: a population-based cohort study. CMAJ 181: Retnakaran R, Qi Y, Sermer M, Connelly PW, Hanley AJ, Zinman B 2008 Glucose intolerance in pregnancy and future risk of prediabetes or diabetes. Diabetes Care 31: Retnakaran R, Qi Y, Sermer M, Connelly PW, Zinman B, Hanley AJ 2008 Isolated hyperglycemia at 1-hour on oral glucose tolerance test in pregnancy resembles gestational diabetes in predicting postpartum metabolic dysfunction. Diabetes Care 31: Retnakaran R, Qi Y, Sermer M, Connelly PW, Zinman B, Hanley AJ 2009 An abnormal screening glucose challenge test in pregnancy predicts postpartum metabolic dysfunction, even when the antepartum oral glucose tolerance test is normal. Clin Endocrinol (Oxf) 71: Retnakaran R, Shah BR 2009 Abnormal screening glucose challenge test in pregnancy and future risk of diabetes in young women. Diabet Med 26: Bentley-Lewis R 2009 Late cardiovascular consequences of gestational diabetes. Semin Reprod Med 27: Retnakaran R, Qi Y, Connelly PW, Sermer M, Zinman B, Hanley AJ 2010 Glucose intolerance in pregnancy and postpartum risk of metabolic syndrome in young women. J Clin Endocrinol Metab 95: National Diabetes Data Group 1979 Classification and diagnosis of diabetes mellitus and other categories of glucose intolerance. Diabetes 28: McPherson R, Frohlich J, Fodor G, Genest J, Canadian Cardiovascular Society 2006 Canadian Cardiovascular Society Canadian Cardiovascular Society position statement recommendations for the diagnosis and treatment of dyslipidemia and prevention of cardiovascular disease. Can J Cardiol 22: McQueen MJ, Hawken S, Wang X, Ounpuu S, Sniderman A, Probstfield J, Steyn K, Sanderson JE, Hasani M, Volkova E, Kazmi K, Yusuf S 2008 INTERHEART Study Investigators. Lipids, lipoproteins, and apolipoproteins as risk markers of myocardial infarction in 52 countries (the INTERHEART study): a case-control study. Lancet 372: Montelongo A, Lasunción MA, Pallardo LF, Herrera E 1992 Longitudinal study of plasma lipoproteins and hormones during pregnancy in normal and diabetic women. Diabetes 41: Vitoratos N, Kassanos D, Salamalekis E, Sirisratidis Ch, Baimacou E, Creatsas G 2002 Maternal homocysteine levels and plasma lipids in gestational diabetes: is there any relationship? J Obstet Gynaecol 22: de Arcos F, Castelo-Branco C, Casals E, Sanllehy C, Cararach V 1998 Normal and gestational diabetic pregnancies. Lipids, lipoproteins and apolipoproteins. J Reprod Med 43: Rizzo M, Berneis K, Altinova AE, Toruner FB, Akturk M, Ayvaz G, Rini GB, Spinas GA, Arslan M 2008 Atherogenic lipoprotein phenotype and LDL size and subclasses in women with gestational diabetes. Diabet Med 25: Qiu C, Rudra C, Austin MA, Williams MA 2007 Association of gestational diabetes mellitus and low-density lipoprotein (LDL) particle size. Physiol Res 56: Enquobahrie DA, Williams MA, Qiu C, Luthy DA 2005 Early pregnancy lipid concentrations and the risk of gestational diabetes mellitus. Diabetes Res Clin Pract 70: Koukkou E, Watts GF, Lowy C 1996 Serum lipid, lipoprotein and apolipoprotein changes in gestational diabetes mellitus: a cross-sectional and prospective study. J Clin Pathol 49: Verma A, Boney CM, Tucker R, Vohr BR 2002 Insulin resistance syndrome in women with prior history of gestational diabetes mellitus. J Clin Endocrinol Metab 87: Lauenborg J, Mathiesen E, Hansen T, Glümer C, Jørgensen T, Borch-Johnsen K, Hornnes P, Pedersen O, Damm P 2005 The prevalence of the metabolic syndrome in a Danish population of women with previous gestational diabetes mellitus is three-fold higher than in the general population. J Clin Endocrinol Metab 90: Kjos SL, Buchanan TA, Montoro M, Coulson A, Mestman JH 1991 Serum lipids within 36 mo of delivery in women with recent gestational diabetes. Diabetes 40(Suppl 2): Meyers-Seifer CH, Vohr BR 1996 Lipid levels in former gestational diabetic mothers. Diabetes Care 19: Sánchez-Vera I, Bonet B, Viana M, Quintanar A, Martín MD, Blanco P, Donnay S, Albi M 2007 Changes in plasma lipids and increased low-density lipoprotein susceptibility to oxidation in pregnancies complicated by gestational diabetes: consequences of obesity. Metabolism 56: Bergholm R, Tiikkainen M, Vehkavaara S, Tamminen M, Teramo K, Rissanen A, Yki-Järvinen H 2003 Lowering of LDL cholesterol rather than moderate weight loss improves endothelium-dependent vasodilatation in obese women with previous gestational diabetes. Diabetes Care 26: Yeboah J, Crouse JR, Hsu FC, Burke GL, Herrington DM 2007 Brachial flow-mediated dilation predicts incident cardiovascular events in older adults: the Cardiovascular Health Study. Circulation 115: