Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2023 Mar 1.
Published in final edited form as: Diabetes Res Clin Pract. 2022 Feb 19;185:109790. doi: 10.1016/j.diabres.2022.109790

Associations of Glycemia and Lipid Levels in Pregnancy with Dyslipidemia 10–14 Years Later: the HAPO Follow-Up Study

Lynn P Lowe 1,*, Amanda M Perak 1,2,*, Alan Kuang 1, Donald M Lloyd-Jones 1, David A Sacks 3, Chaicharn Deerochanawong 4, Michael Maresh 5, Ronald C Ma 6, William L Lowe Jr 1, Boyd E Metzger 1, Denise M Scholtens 1; HAPO and HAPO Follow-Up Study Cooperative Research Groups7
PMCID: PMC9210945  NIHMSID: NIHMS1786147  PMID: 35192911

Abstract

Aims:

To examine associations of pregnancy glycemia with future dyslipidemia.

Methods:

We analyzed data from Hyperglycemia and Adverse Pregnancy Outcome Follow-Up Study participants. We examined associations of gestational diabetes (GDM), sum of fasting, 1-hour, and 2-hour glucose z-scores after 75-g load, insulin sensitivity, and lipid levels at 24–32 weeks’ gestation with dyslipidemia 10–14 years postpartum.

Results:

Among 4,693 women, 14.3% had GDM. At follow-up, mean (SD) age was 41.7 (5.7) years, 32.3% had total cholesterol (TC)≥5.17, 27.2% had HDL cholesterol<1.29, 22.4% had LDL cholesterol (LDL-C)≥3.36, 10.9% had triglycerides≥1.69 mmol/L, and 2.9% had type 2 diabetes. After covariate adjustment, pregnancy glycemic measures were associated with all follow-up dyslipidemias. After additional adjustment for pregnancy lipids, GDM remained associated with TC≥5.17 mmol/L (odds ratio [95% CI], 1.63 [1.22–2.18]) and LDL-C≥3.36 mmol/L (1.63 [1.20–2.22]), even in the absence of type 2 diabetes development (1.55 [1.15–2.10] and 1.56 [1.13–2.16], respectively). Continuous glycemic measures in pregnancy were significantly associated with all follow-up dyslipidemias, independent of pregnancy lipids and type 2 diabetes.

Conclusions:

Pregnancy glycemia was associated with dyslipidemia 10–14 years later, independent of pregnancy lipid levels and in the absence of type 2 diabetes development. Lipid screening after GDM deserves special consideration.

Keywords: pregnancy, diabetes, lipids

1. INTRODUCTION

Gestational diabetes mellitus (GDM)1 and mild glucose intolerance in pregnancy2 are associated with maternal risk for cardiovascular disease (CVD). However, knowledge of underlying pathways to CVD remains incomplete.1,3 Although GDM is strongly associated with higher incidence of type 2 diabetes mellitus (T2DM),4 CVD risk is higher even among individuals who do not develop T2DM.1 GDM-related CVD risk could be due to other risk factors that develop before, during, or after pregnancy, and several studies have examined postpartum dyslipidemia. Results have been mixed, with some5–13 but not all14–16 studies reporting an association of GDM with one or more lipids postpartum. Methodologic challenges have included sample size, GDM ascertainment, length of follow-up, and adjustment for confounders such as antepartum lipid levels and postpartum T2DM development.

We utilized data from the multinational Hyperglycemia and Adverse Pregnancy Outcome (HAPO) Follow-Up Study (FUS) to examine whether: (1) GDM and other glycemic indicators in pregnancy were associated with dyslipidemia 10–14 years later, independent of pregnancy covariates such as body mass index (BMI) and hypertension; (2) lipid levels during pregnancy were associated with later dyslipidemia and attenuated GDM-dyslipidemia associations; and (3) associations between GDM and later dyslipidemia persisted in the absence of T2DM development.

2. MATERIALS AND METHODS

2.1. Study Design and Participants

The HAPO Study was an international, population-based cohort study designed to examine associations of glycemia with pregnancy outcomes.17,18 Pregnant females from 9 countries were examined, including 75-g oral glucose tolerance testing (OGTT), between 24–32 weeks’ gestation. Individuals with pre-existing diabetes were ineligible. OGTT results were unblinded only for plasma glucose levels: fasting >5.83 mmol/L, 2-hour >11.1 mmol/L, random ≥8.88 mmol/L, or any <2.5 mmol/L; any individuals with unblinded glucose values were excluded from subsequent data collection and were not part of the final HAPO dataset. HAPO demonstrated linear associations of glucose levels less than overt diabetes with risks of adverse pregnancy outcomes.17 These results led the International Association of Diabetes in Pregnancy Study Groups (IADSPG) to develop new, less restrictive criteria for GDM diagnosis,19 which were adopted by the World Health Organization (WHO).20

The HAPO Follow-Up Study (FUS) re-examined HAPO mothers 10–14 years later (in 2013–2016) to examine associations of pregnancy glycemia with long-term metabolic outcomes.21 HAPO FUS eligibility required having remained blinded to OGTT results during HAPO (i.e., did not have glucose levels meeting criteria for unblinding, listed above), delivery at ≥37 weeks’ gestation, and no fetal/neonatal death or major malformations. HAPO FUS examined 4,747 mothers; characteristics of HAPO participants who did versus did not participate in HAPO FUS are published.21 After exclusions (bariatric surgery: 49, type 1 diabetes [by autoantibody screening]: 1, cancer treatment: 1), 4,693 individuals comprised the analytic sample. A random ~50% subset with distribution across self-reported race and ethnicity groups (Hispanic, non-Hispanic Asian, non-Hispanic Black, non-Hispanic White) was selected for an ancillary study of gestational lipids; these 2,275 individuals are included in analyses utilizing pregnancy lipid data. See eMethods in the Supplement for details.

The HAPO and HAPO FUS protocols were approved by each center’s institutional review board. All participants gave written informed consent.

2.2. Exposures

Primary exposures were lipid levels and glycemic variables measured at the baseline (24–32 weeks’ gestation) HAPO examination. Venous blood was drawn after overnight fasting and 1 and 2 hours after a 75-g glucose load.

Fasting serum total cholesterol (TC), high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C) and triglycerides (TG) were measured in a central laboratory with direct enzymatic methods with quality control (Supplemental Methods). Non-HDL-C was calculated as TC minus HDL-C.

Fasting, 1-hour and 2-hour plasma glucose were measured in a central laboratory with enzymatic methods with quality control.22 Fasting and 1-hour serum C-peptide were measured with fluoro-immunometric methods (Supplemental Methods).22,23 GDM was defined post hoc using IADPSG/WHO criteria,19,20 i.e., fasting plasma glucose ≥5.1 mmol/L, 1-hour ≥10.0 mmol/L, or 2-hour ≥8.5 mmol/L after 75-g OGTT.19 Glucose z-scores at fasting, 1-hour, and 2-hour OGTT (calculated among the entire HAPO cohort) were summed as a composite glycemia variable. Insulin sensitivity was calculated from fasting and 1-hour glucose and C-peptide values.24

2.3. Outcomes

Primary outcomes were maternal lipid levels at the HAPO FUS exam, 10–14 years postpartum. Venous blood was drawn after overnight fasting. Fasting serum TC, HDL-C, and TG were measured in a central laboratory with direct enzymatic methods (Supplemental Methods). Non-HDL-C was calculated as TC minus HDL-C. LDL-C was calculated using the Friedewald equation25 for samples with TG ≤4.52 mmol/L (otherwise, set missing [N=9]). Cholesterol-lowering medication use was determined from the question, “Are you taking medication for treatment of high cholesterol?”.

Dyslipidemia was defined using US clinical guidelines.26,27 Because guidelines vary and multiple adverse levels are defined (e.g., “borderline high” versus “high”), “primary” and “secondary” thresholds were used. Primary thresholds were more liberal: TC ≥5.17,26,28 HDL-C <1.29,29,30 non-HDL-C ≥4.14,26 LDL-C ≥3.36,26 and TG ≥1.69 mmol/L.27,30 Secondary thresholds were more stringent: TC ≥6.21,28,30 HDL-C <1.03,26 non-HDL-C ≥4.91,26 LDL-C ≥4.14,26,30 and TG ≥2.26 mmol/L.26 Cholesterol-lowering medication use was also considered dyslipidemia for TC, non-HDL-C, and LDL-C (regardless of lipid level).

T2DM was examined as a joint outcome. After fasting blood draw, a 75-g glucose load was administered and 2-hour venous blood samples were drawn. Fasting and 2-hour plasma glucose were measured with enzymatic methods at a central laboratory (Supplemental Methods). T2DM was defined as self-reported treated diabetes or fasting glucose ≥7.0 mmol/L or 2-hour glucose ≥11.1 mmol/L.31

2.4. Covariates

Covariates in all models (i.e., “base model”) included field center (each with high demographic homogeneity) and maternal variables during the HAPO pregnancy. Parity was abstracted from medical records. Standardized questionnaires provided maternal age, smoking, and alcohol use. Gestational age was determined with last menstrual period and ultrasound dating.18 Height, weight, and blood pressure were measured twice at the HAPO exam with calibrated instruments by trained study personnel; additional blood pressure measurements throughout pregnancy were abstracted from medical records. BMI was calculated as weight (kilograms) divided by height (meters) squared. Obesity (reported descriptively, although BMI was used for adjustment) was defined as BMI ≥33 kg/m2 during pregnancy, corresponding to pre-pregnancy BMI ≥30 kg/m2 (based on regression of pregnancy BMI on pre-pregnancy BMI32). Blood pressure was classified as normal, chronic hypertension, gestational hypertension, or preeclampsia using International Society for Study of Hypertension in Pregnancy guidelines.23,33

2.5. Statistical Analysis

Multiple linear and logistic regression were used to examine associations of exposures (pregnancy lipid levels and glycemia, singly and jointly) with outcomes (lipid levels and dyslipidemia at follow-up), after covariate adjustment. To determine whether associations between exposures and three key outcomes (TC, LDL-C, and HDL-C, selected for importance in cardiovascular risk34) occurred in the absence of T2DM development, we used multinomial logistic regression for a categorical joint outcome defined as: no dyslipidemia/no T2DM (referent), dyslipidemia/no T2DM, T2DM/no dyslipidemia, or dyslipidemia/T2DM. Dyslipidemia outcomes were examined using primary and secondary thresholds, described above. We performed two sensitivity analyses. First, we excluded participants taking cholesterol-lowering medication at follow-up. Second, to determine whether loss of statistical significance for glycemic predictors with addition of pregnancy lipids to the model was due to confounding or the smaller sample size with pregnancy lipid data, we repeated analyses for glycemic predictors (without pregnancy lipids in the model) using the smaller sample with pregnancy lipid data available; when statistical significance differed compared to the main results, this was interpreted as evidence for loss of power (rather than confounding).

All analyses were conducted in R (3.5.1); P-values <.05 were considered statistically significant.35 Model fit was confirmed with Hosmer-Lemeshow goodness-of-fit tests.36 Collinearity of model predictors was evaluated using pairwise correlations for continuous variables, chi-square tests for categorical variables, and ANOVA for categorical and continuous pairs. Pairwise correlations were 0.01–0.29, except BMI and insulin sensitivity during pregnancy; BMI was included as a covariate to adjust for its association with later dyslipidemia.

3. RESULTS

3.1. Participant Characteristics

A total of 4,693 participants—including 672 (14.3%) with GDM during the HAPO pregnancy—were included (Table 1). Mean ages were 30.1 (SD 5.6) years during pregnancy and 41.7 (5.7) years at follow-up. Obesity was present in 11.8% during pregnancy and 25.4% at follow-up. Lipid levels were substantially higher during pregnancy (as physiologically expected37) than at follow-up. At follow-up, 2.9% had T2DM, 1.9% used cholesterol-lowering medication, and 4.0%–32.3% had dyslipidemia. Participants with (vs without) GDM during the HAPO pregnancy more often self-identified as Hispanic or non-Hispanic Asian, were older, had higher BMIs, and had more dyslipidemia at follow-up (Table 1).

Table 1.

Participant characteristics at pregnancy and follow-up examinations, overall and by gestational diabetes mellitus status during the study pregnancy: The HAPO Follow-Up Study

During Pregnancy At Follow-Up
Overall GDM No GDM Overall GDM No GDM
N (%) 4693 672 (14.3) 4021 (85.7) 4693 672 (14.3) 4021 (85.7)
Race and ethnicity, N (%)
 Hispanic 488 (10.4) 108 (16.1) 380 (9.5) --- --- ---
 Non-Hispanic Asian 1173 (25) 196 (29.2) 977 (24.3) --- --- ---
 Non-Hispanic Black 733 (15.6) 82 (12.2) 651 (16.2) --- --- ---
 Non-Hispanic White 2212 (47.1) 270 (40.2) 1942 (48.3) --- --- ---
 Othera 87 (1.9) 16 (2.4) 71 (1.8) --- --- ---
Age at exam, mean (SD), years 30.1 (5.6) 31.9 (5.3) 29.8 (5.6) 41.7 (5.7) 43.6 (5.4) 41.4 (5.7)
Time since delivery, mean (SD), years --- --- --- 11.4 (1.2) 11.5 (1.2) 11.4 (1.2)
Gestational age at exam, mean (SD), weeks 27.7 (1.7) 27.9 (1.7) 27.7 (1.7) --- --- ---
Parity ≥1 at pregnancy HAPO exam, N (%) 2420 (51.6) 383 (57) 2037 (50.7) --- --- ---
Parity ≥1 since HAPO delivery, N (%) --- --- --- 2560 (54.7) 303 (45.4) 2257 (56.2)
Smoking, N (%) 243 (5.2) 42 (6.2) 201 (5) 368 (7.9) 75 (11.2) 293 (7.3)
Alcohol consumption, N (%) 402 (8.6) 56 (8.3) 346 (8.6) 1527 (32.6) 195 (29.1) 1332 (33.2)
Body mass index, mean (SD), kg/m2 27.4 (4.8) 29.7 (5.2) 27 (4.6) 27 (6.1) 28.9 (6.5) 26.7 (6)
Obesity, N (%) b 554 (11.8) 141 (21) 413 (10.3) 1178 (25.4) 246 (37) 932 (23.5)
Preeclampsia, N (%) c 205 (4.4) 57 (8.5) 148 (3.7) --- --- ---
Gestational Hypertension, N (%) d 295 (6.3) 63 (9.4) 232 (5.8) --- --- ---
Fasting glucose, mean (SD), mmol/L 4.5 (0.4) 4.9 (0.4) 4.4 (0.3) 5.1 (0.7) 5.5 (1.2) 5.1 (0.5)
1-hour glucose, mean (SD), mmol/L 7.4 (1.7) 9.6 (1.6) 7 (1.4) --- --- ---
2-hour glucose, mean (SD), mmol/L 6.1 (1.3) 7.6 (1.5) 5.9 (1.1) 6.3 (2) 7.5 (2.9) 6.1 (1.7)
Fasting insulin, mean (SD), pmol/L --- --- --- 64.6 (185.8) 64.6 (185.8) 64.6 (185.8)
Fasting c-peptide, mean (SD), nmol/L 0.6 (0.3) 0.9 (0.4) 0.6 (0.2) 0.7 (0.3) 0.8 (0.3) 0.7 (0.3)
Type 2 Diabetes, N (%) --- --- --- 134 (2.9) 71 (10.7) 63 (1.6)
Total cholesterol, mean (SD), mmol/L e 6.4 (1.2) 6.3 (1.2) 6.4 (1.2) 4.8 (0.9) 5 (0.9) 4.8 (0.9)
Total cholesterol ≥5.17 mmol/L, N (%) e 1913 (85) 274 (82.5) 1639 (85.4) 1477 (32.3) 285 (44) 1192 (30.4)
Total cholesterol ≥6.21 mmol/L, N (%) e 1217 (54.1) 174 (52.4) 1043 (54.4) 369 (8.1) 81 (12.5) 288 (7.3)
HDL cholesterol, mean (SD), mmol/L e 1.9 (0.4) 1.8 (0.4) 1.9 (0.4) 1.5 (0.4) 1.5 (0.4) 1.5 (0.4)
HDL cholesterol <1.29 mmol/L, N (%) e 114 (5.1) 20 (6) 94 (4.9) 1239 (27.2) 220 (34.5) 1019 (26)
HDL cholesterol <1.03 mmol/L, N (%) e 17 (0.8) 6 (1.8) 11 (0.6) 222 (4.9) 35 (5.5) 187 (4.8)
Non-HDL cholesterol, mean (SD), mmol/L e 4.5 (1.1) 4.4 (1.1) 4.5 (1.1) 3.3 (0.8) 3.5 (0.9) 3.2 (0.8)
Non-HDL cholesterol ≥4.14 mmol/L, N (%) e 1349 (60) 199 (59.9) 1150 (60) 747 (16.5) 164 (25.3) 583 (15)
Non-HDL cholesterol ≥4.91 mmol/L, N (%) e 727 (32.3) 98 (29.5) 629 (32.8) 261 (5.8) 64 (9.9) 197 (5.1)
LDL cholesterol, mean (SD), mmol/L e,f 3.6 (1.1) 3.5 (1) 3.6 (1.1) 2.8 (0.8) 3 (0.8) 2.8 (0.7)
LDL cholesterol ≥3.36 mmol/L, N (%) e,f 1280 (56.9) 183 (55.1) 1097 (57.2) 1023 (22.4) 203 (31.5) 820 (20.9)
LDL cholesterol ≥4.14 mmol/L, N (%) e,f 657 (29.2) 79 (23.8) 578 (30.1) 309 (6.8) 74 (11.5) 235 (6)
Triglycerides, median (IQR), mmol/L e,g 2.2 (1.8–2.8) 2.5 (2.1–3.2) 2.2 (1.7–2.7) 0.9 (0.7–1.3) 1.1 (0.8–1.5) 0.9 (0.7–1.2)
Triglycerides ≥1.69 mmol/L, N (%) e 1779 (78.4) 298 (88.7) 1481 (76.6) 496 (10.9) 113 (17.7) 383 (9.8)
Triglycerides ≥2.26 mmol/L, N (%) e 1075 (47.4) 214 (63.7) 861 (44.5) 184 (4) 41 (6.4) 143 (3.7)
Taking cholesterol-lowering medication, N (%) --- --- --- 87 (1.9) 27 (4) 60 (1.5)
Open in a new tab
a

Included Native American/Alaskan Native at US field centers. First Nation in Toronto, Ontario, Canada, and self-identified as other at all field centers.

b

Defined during pregnancy as body mass index ≥33 kg/m2, based on previously reported regression of pregnancy body mass index on pre-pregnancy body mass index, and selected to correspond to pre-pregnancy body mass index ≥30 kg/m2.

c

Defined as a systolic blood pressure of ≥140 mm Hg or diastolic blood pressure of ≥90 mm Hg on ≥2 occasions which are ≥6 hours apart, plus proteinuria of ≥1+ on a dipstick or ≥300 mg/24 hours.

d

If the criteria elevated blood pressure (as for preeclampsiac) were first met after 20 weeks of gestation, but those for proteinuria were not met, the hypertension was classified as gestational hypertension.

e

Lipid measurements were available in a subset of 2275 women during pregnancy.

f

Directly measured during pregnancy; calculated using the Friedewald equation (in women with TG<4.52 mmol/L) at follow-up.

g

Because the distributions of triglycerides at baseline and follow-up were skewed, median and interquartile ranges are reported, and continuous triglyceride levels were log-transformed in subsequent analyses.

GDM, gestational diabetes mellitus; HAPO, Hyperglycemia and Adverse Pregnancy Outcome; HDL, high-density lipoprotein; IQR, interquartile range; LDL, low-density lipoprotein; SD, standard deviation.

3.2. Associations of lipid levels and glycemia during pregnancy with continuous lipid levels 10–14 years later

In base (covariate-adjusted) models and models including pregnancy glycemic indicators, the level of each lipid during pregnancy was significantly associated with the level of the same lipid at follow-up (e.g., TC at follow-up: β, 0.47 mmol/L [95% CI, 0.44–0.51] per 1-SD [0.88 mmol/L] higher TC in pregnancy, adjusted for GDM; Figure 1, first panel, blue circle). In base models, GDM was significantly associated with more adverse levels of all lipids at follow-up (e.g., higher LDL-C, lower HDL-C). After additional adjustment for pregnancy lipid levels, GDM remained significantly associated with levels of TC (β, 0.14 mmol/L [0.05–0.23]), non-HDL-C (β, 0.16 mmol/L [0.08–0.25]), and LDL-C (β, 0.16 mmol/L [0.08–0.24]), but not HDL-C or log-TG (Figure 1, pink squares). Sum of glucose z-scores and insulin sensitivity during pregnancy were significantly associated with all lipid levels at follow-up in base models; most associations remained significant after adjustment for respective lipid levels during pregnancy (Figure 1, pink squares).

Figure 1. Adjusted individual and joint associations of glycemic variables and lipid levels during pregnancy with continuous lipid levels at follow-up: The HAPO Follow-Up Study.

Figure 1.

Open in a new tab

The base model included field center and the following maternal variables during the HAPO pregnancy: age, gestational age, parity (0, 1+), smoking (yes/no), drinking (yes/no), body mass index, height, chronic hypertension, gestational hypertension, and preeclampsia. Additional covariate adjustments for lipid level (for glycemic predictors) or glycemic indicators (for lipid predictors) are indicated by the shape in the forest plot, as shown in the legend. For all models including pregnancy lipid levels as the predictor or covariate, the same lipid examined at follow-up was included as the pregnancy lipid (i.e., total cholesterol during pregnancy for the total cholesterol at follow-up outcome), and the beta for association is per +1 standard deviation in pregnancy lipid level. Standard deviations for pregnancy lipid levels were as follows: total cholesterol 1.18 mmol/L, non-HDL cholesterol 1.09 mmol/L, HDL cholesterol 0.39 mmol/L, LDL cholesterol 1.05 mmol/L, and log-Triglycerides 0.34. Sum of glucose refers to sum of glucose z-scores and includes z-scores for fasting, 1-hour, and 2-hour plasma glucose levels, calculated among the entire original HAPO cohort; associations are per +1 standard deviation in sum of glucose z-scores. Insulin sensitivity was calculated from fasting and 1-hour glucose and C-peptide values among the entire original HAPO cohort; associations are per +1 standard deviation in insulin sensitivity z-score. For total, non-HDL, and LDL cholesterol outcomes, women taking cholesterol medications at follow-up were excluded. Sample sizes in the figure range from 4464–4554 for models without pregnancy lipid levels and from 2175–2215 for models including pregnancy lipid levels. For the insulin sensitivity predictor with total and LDL cholesterol outcomes, the loss of significance with addition of pregnancy lipids to the model may be due to the decreased sample size with pregnancy lipid data available, as indicated by lack of statistical significance for the insulin sensitivity predictor in the smaller sample even when pregnancy lipids were not included in the model (data not shown). CI, confidence interval; GDM, gestational diabetes mellitus; HAPO, Hyperglycemia and Adverse Pregnancy Outcome; HDL, high-density lipoprotein; LDL, low-density lipoprotein; SD, standard deviation.

In sensitivity analyses excluding participants taking cholesterol-lowering medication at follow-up, results were similar (Supplemental Table S1).

3.3. Associations of lipid levels and glycemia during pregnancy with dyslipidemia 10–14 years later

In base (covariate-adjusted) models and models including pregnancy glycemic indicators, the level of each lipid during pregnancy was significantly associated with the corresponding dyslipidemia at follow-up (e.g., for TC ≥5.17 mmol/L or cholesterol-lowering medication at follow-up: OR 3.18 [95% CI, 2.80–3.63] per 1-SD [1.18 mmol/L] higher TC in pregnancy, adjusted for GDM; Figure 2, first panel, blue circle). In base models, GDM was significantly associated with all dyslipidemias. After additional adjustment for pregnancy lipid levels, GDM remained significantly associated with TC ≥5.17 mmol/L (OR 1.63 [1.22–2.18]) and LDL-C ≥3.36 mmol/L (OR 1.63 [1.20–2.22]), but not other dyslipidemias (Figure 2, pink squares); however, sensitivity analyses suggested lack of significance for the latter may be due to lower power in the smaller sample. Sum of glucose z-scores and insulin sensitivity during pregnancy were associated with nearly all dyslipidemias at follow-up, even after adjustment for the respective lipid level during pregnancy (Figure 2, pink squares).

Figure 2. Adjusted individual and joint associations of glycemic variables and lipid levels during pregnancy with dyslipidemia at follow-up: The HAPO Follow-Up Study.

Figure 2.

Open in a new tab

The base model included field center and the following maternal variables during the HAPO pregnancy: age, gestational age, parity (0, 1+), smoking (yes/no), drinking (yes/no), body mass index, height, chronic hypertension, gestational hypertension, and preeclampsia. Additional covariate adjustments for lipid level (for glycemic predictors) or glycemic indicators (for lipid predictors) are indicated by the shape in the forest plot, as shown in the legend. For all models including pregnancy lipid levels as the predictor or covariate, the same lipid examined at follow-up was included as the pregnancy lipid (i.e., total cholesterol during pregnancy for the total cholesterol at follow-up outcome), and the odds ratio for association is per +1 standard deviation in pregnancy lipid level. Standard deviations for pregnancy lipid levels were as follows: total cholesterol 1.18 mmol/L, non-HDL cholesterol 1.09 mmol/L, HDL cholesterol 0.39 mmol/L, LDL cholesterol 1.05 mmol/L, and log-Triglycerides 0.34. Sum of glucose refers to sum of glucose z-scores and includes z-scores for fasting, 1-hour, and 2-hour plasma glucose levels, calculated among the entire original HAPO cohort; associations are per +1 standard deviation in sum of glucose z-scores. Insulin sensitivity was calculated from fasting and 1-hour glucose and C-peptide values among the entire original HAPO cohort; associations are per +1 standard deviation in insulin sensitivity z-score. For total, non-HDL, and LDL cholesterol outcomes, use of cholesterol medication at follow-up was also considered dyslipidemia (regardless of lipid level). Sample sizes in the figure ranged from 4479–4569 for models without pregnancy lipid levels and from 2201–2215 for models including pregnancy lipid levels. For the GDM predictor with non-HDL and HDL cholesterol and triglycerides outcomes, the loss of significance with addition of pregnancy lipids to the model may be due to the decreased sample size with pregnancy lipid data available, as indicated by lack of statistical significance for the GDM predictor in the smaller sample even when pregnancy lipids were not included in the model (data not shown). CI, confidence interval; GDM, gestational diabetes mellitus; HAPO, Hyperglycemia and Adverse Pregnancy Outcome; HDL, high-density lipoprotein; LDL, low-density lipoprotein; SD, standard deviation.

In secondary analyses using more stringent dyslipidemia thresholds, findings were generally similar to primary analyses. Associations were less often statistically significant for HDL-C <1.03 mmol/L and TG ≥2.26 mmol/L (Supplemental Table S2 versus Figure 2); these outcomes had the smallest sample sizes (among individuals with GDM, N=35 and N=41 respectively; Table 1).

In sensitivity analyses excluding individuals taking cholesterol-lowering medication at follow-up, associations were similar using primary dyslipidemia thresholds (Supplemental Table S1). They were less often statistically significant using secondary thresholds; analyses suggested this was partly due to loss of power.

3.4. Associations of lipid levels and glycemia during pregnancy with joint dyslipidemia and T2DM status 10–14 years later

In base (covariate-adjusted) models and models including pregnancy glycemic indicators, the level of each lipid (TC, LDL-C and HDL-C) during pregnancy was associated with risk for the corresponding dyslipidemia with or without concurrent T2DM at follow-up (Figure 3, blue symbols in columns labeled “No Diabetes, + Dyslipidemia” and “+Diabetes, + Dyslipidemia”). In base models, GDM was significantly associated with T2DM (with or without dyslipidemia) and all dyslipidemias (with or without T2DM) (Figure 3, pink circles). After additional adjustment for pregnancy lipid levels, GDM remained associated with dyslipidemia without T2DM for TC ≥5.17 mmol/L (OR 1.55 [1.15–2.10]) and LDL-C ≥3.36 mmol/L (OR 1.56 [1.13–2.16]), but not for HDL-C <1.29 mmol/L (Figure 3, pink squares in the first column); sensitivity analyses suggested the latter may have been related to loss of power. Sum of glucose z-scores and insulin sensitivity in pregnancy were each significantly associated with all dyslipidemias without T2DM at follow-up, even after adjustment for pregnancy lipid levels (Figure 3, pink squares in the first column).

Figure 3. Adjusted individual and joint associations of glycemic variables and lipid levels during pregnancy with combined status for dyslipidemia and type 2 diabetes mellitus at follow-up: The HAPO Follow-Up Study.

Figure 3.

Open in a new tab

The base model included field center and the following maternal variables during the HAPO pregnancy: age, gestational age, parity (0, 1+), smoking (yes/no), drinking (yes/no), body mass index, height, chronic hypertension, gestational hypertension, and preeclampsia. Additional covariate adjustments for lipid level (for glycemic predictors) or glycemic indicators (for lipid predictors) are indicated by the shape in the forest plot, as shown in the legend. For all models including pregnancy lipid levels as the predictor or covariate, the same lipid examined at follow-up was included as the pregnancy lipid (i.e., total cholesterol during pregnancy for the total cholesterol at follow-up outcome), and the odds ratio for association is per +1 standard deviation in pregnancy lipid level. Standard deviations for pregnancy lipid levels were as follows: total cholesterol 1.18 mmol/L, non-HDL cholesterol 1.09 mmol/L, HDL cholesterol 0.39 mmol/L, LDL cholesterol 1.05 mmol/L, and log-Triglycerides 0.34. Sum of glucose refers to sum of glucose z-scores and includes z-scores for fasting, 1-hour, and 2-hour plasma glucose levels, calculated among the entire original HAPO cohort; associations are per +1 standard deviation in sum of glucose z-scores. Insulin sensitivity was calculated from fasting and 1-hour glucose and C-peptide values among the entire original HAPO cohort; associations are per +1 standard deviation in insulin sensitivity z-score. For total, non-HDL, and LDL cholesterol outcomes, use of cholesterol medication at follow-up was also considered dyslipidemia (regardless of lipid level). Sample sizes in the figure ranged from 4495–4564 for models without pregnancy lipid levels and from 2190–2203 for models including pregnancy lipid levels. For the GDM predictor with the no type 2 diabetes, + dyslipidemia for HDL cholesterol outcome, the loss of significance with addition of pregnancy lipids to the model may be due to the decreased sample size with pregnancy lipid data available, as indicated by lack of statistical significance for the GDM predictor in the smaller sample even when pregnancy lipids were not included in the model (data not shown). CI, confidence interval; GDM, gestational diabetes mellitus; HAPO, Hyperglycemia and Adverse Pregnancy Outcome; HDL, high-density lipoprotein; HDL-C, high-density lipoprotein cholesterol; LDL, low-density lipoprotein; LDL-C, low-density lipoprotein cholesterol; SD, standard deviation; T2M, type 2 diabetes mellitus; TC, total cholesterol.

In secondary analyses using more stringent dyslipidemia thresholds, patterns and point estimates were similar, but statistical significance was often lost (Supplemental Table S3).

4. DISCUSSION

The principal findings of this analysis were that, first, GDM and other pregnancy glycemic indicators were associated with maternal dyslipidemia 10–14 years later for all lipids (TC, HDL-C, non-HDL-C, LDL-C, and TG), independent of pregnancy covariates such as BMI and hypertension. Second, pregnancy lipid levels were significantly associated with later dyslipidemia; however, even with pregnancy lipid levels in the model, GDM remained significantly associated with dyslipidemia for TC (≥5.17 mmol/L), non-HDL-C (≥4.91 mmol/L), and LDL-C (≥3.36 and ≥4.14 mmol/L), and sensitivity analyses suggested that lack of statistical significance for HDL-C (<1.29 mmol/L) may be related to low power. Third, in joint outcomes analyses, GDM was significantly associated with later dyslipidemia in the absence of T2DM, and associations remained significant for TC and LDL-C after adjustment for pregnancy lipid levels. In all analyses, continuous glycemic indicators during pregnancy were associated with a broader range of dyslipidemias (e.g., TG and HDL-C).

Prior studies have examined associations of GDM with subsequent levels of TC,8–15 HDL-C,6,8–16 LDL-C,7–16 TG,6–16 and less commonly, categorical dyslipidemia.5,8,9 Sample size was limited in most studies (GDM N=17–166 in all but two6,14), and some had very short follow-up (e.g., 0–3 months postpartum10,11,14). Some screened only selectively for GDM,6,15,16 and none used the IADSPG GDM criteria.19 Notably, the majority performed no covariate adjustment,5–9,11,12 none adjusted for antepartum lipid levels, and only two considered T2DM development in the analyses.14,15 Among two studies with follow-up durations (10–11 years) similar to the present study, one found significant unadjusted associations of GDM with TC, LDL-C, and TG levels and TG ≥1.69 mmol/L but not with HDL-C levels or HDL-C <1.29 mmol/L;8 the other found significant unadjusted associations of GDM with levels of TG and HDL-C (no other lipids were examined).6 The present study adds to the literature by demonstrating in a large, multinational sample that GDM is associated with dyslipidemia for TC, non-HDL-C, LDL-C, and HDL-C 10–14 years later independent of not only antepartum lipid levels and other covariates but also intercurrent T2DM development.

The associations of GDM and other glycemic indicators with dyslipidemias 10–14 years later after adjustment for antepartum lipid levels suggest that dysglycemia in pregnancy may presage sustained dyslipidemia. Further study will be needed to determine whether the GDM-dyslipidemia link is causal within the overall GDM-CVD association. However, probably more clinically useful is the observation that GDM was associated with later dyslipidemia even in the absence of T2DM development. These results suggest a need to specifically define postpartum lipid screening for individuals with GDM.

Taking the United States as an example, whereas the American Diabetes Association recommends that individuals with GDM should be screened for T2DM 4–12 weeks postpartum and at least every 3 years thereafter,31 recommendations about dyslipidemia screening after GDM are less clear. The 2019 American College of Cardiology (ACC)/American Heart Association (AHA) CVD prevention guidelines include a level IIa recommendation (i.e., benefit>>risk, reasonable to perform) for universal CVD risk factor screening—including lipids—in all adults every 4–6 years starting at age 20 years.34 However, the most recent United States Preventive Services Task Force recommended lipid screening for females only in the presence of certain risk conditions including T2DM but not GDM.38 This recommendation may need to be reconsidered.

Under current guidelines, lipid screening rates are about 70% for US adults aged ≥20 years overall39 but may be as low as 19%–39% among females of childbearing age.40 Females are also less likely than males to be offered statins (despite being eligible) or treated at guideline-recommended intensities.41 Awareness among primary care clinicians, including obstetrician/gynecologists, about increased rates of CVD risk factors and CVD events among females with GDM may help lessen these disparities. This is supported by recent ACC/AHA guidelines for CVD prevention34 and cholesterol30 highlighting the role of pregnancy histories in CVD risk assessment for females.

The present analyses benefited from a relatively large and diverse study sample, high-quality GDM ascertainment with participant and clinician blinding to pregnancy glycemia (therefore no treatment given to confound results), ≥10-year follow-up, and careful confounder control. However, there are limitations. First, although blinding and absence of GDM treatment (which were possible because HAPO preceded, and informed, the current GDM definition) enabled assessment of the natural history of lipid metabolism after GDM, associations with later dyslipidemia may be weaker in the setting of diagnosed GDM with treatment and lifestyle modification. Indeed, an important implication of our findings is the need for comprehensive CVD prevention efforts in females after GDM. Second, lipids were measured during pregnancy and 10–14 years postpartum, but no conclusions can be drawn about relationships between GDM and lipids measured pre-pregnancy or early postpartum. Thus, the optimal timing and frequency of lipid screening after GDM (e.g., every 4–6 years as for the general population versus more often) remains unknown. Third, despite relatively large sample size overall (N=4,693), power was more limited for analyses requiring pregnancy lipid data (N=2,275), especially analyses of stringent dyslipidemia thresholds and joint dyslipidemia/T2DM status. Fourth, consistent with the eligibility criteria for HAPO FUS, there were some differences between participants in the original HAPO study and those in the HAPO FUS; similarly, consistent with the design of the gestational lipids substudy, there were some differences between groups in the current analysis who did vs did not have lipid data from pregnancy (see eMethods in the Supplement). In both cases, eligibility criteria served to increase the proportions of individuals self-identifying as a racial and ethnic group other than non-Hispanic White, which lends generalizability to the study’s findings. In addition, HAPO participants who had severe GDM based on prior, more restrictive diagnostic criteria (who were therefore unblinded to their glucose results and not part of the final HAPO dataset) or had preterm delivery were not eligible for HAPO FUS; consistent with this, eligible (vs ineligible) individuals tended to have modestly healthier profiles and outcomes in the HAPO pregnancy, which would be expected to bias associations toward the null. Of note, although the timing of OGTT in HAPO (24–32 weeks’ gestation) is standard, sensitivity for mild dysglycemia is greater in the second half of pregnancy versus earlier testing with the same glucose thresholds, which would similarly create potential for bias toward the null. For the lipid substudy, which included a stratified random sample across self-reported racial and ethnic groups, differences between individuals with vs without lipid data from pregnancy were not consistent; whether differences might have impacted the analyses using those gestational lipid data is unknown. Fifth, as in prior HAPO analyses,21,42 we adjusted for field center as a sociodemographic indicator, as other variables such as race, ethnicity, education, and employment status are highly collinear with and confounded by field center in this multinational study. However, a degree of residual confounding by socioeconomic status cannot be excluded.

CONCLUSION

GDM and pregnancy glycemic measures were associated with dyslipidemias 10–14 years later, independent of pregnancy lipid levels, and in most cases, even in the absence of intercurrent T2DM development. Special consideration should be given to lipid screening after GDM.

Supplementary Material

1

Highlights.

  • GDM was associated with dyslipidemia for all lipids 10–14 years later

  • Most GDM-dyslipidemia associations were independent of pregnancy lipid levels

  • GDM-dyslipidemia associations occurred without type 2 diabetes development

Acknowledgements

We gratefully acknowledge the mothers who participated in HAPO and HAPO FUS.

Funding

The HAPO Study was funded by grants R01HD34242 and R01HD34243 from the Eunice Kennedy Shriver National Institute of Child Health and Human Development, with additional HAPO ancillary study data obtained through grants R01DK095963 and R01DK117491 from the National Institute of Diabetes and Digestive and Kidney Diseases. The HAPO Follow-Up Study was funded by grant 1U01DK094830 from the National Institute of Diabetes and Digestive and Kidney Diseases. The ancillary lipid study was funded by a Dixon Translational Research Grant from the Northwestern University Clinical and Translational Sciences (NUCATS) Institute and the Northwestern Memorial Foundation, an Eleanor Wood-Prince Grant from the Woman’s Board of Northwestern Memorial Hospital, and a subcontract under grant 17SFRN33660752 from the American Heart Association. The research reported in this article was supported, in part, by grant UL1TR001422 from the National Center for Advancing Translational Sciences, National Institutes of Health. AMP’s work was supported by grant K23HL145101 from the National Heart, Lung, and Blood Institute and a Pediatric Physician-Scientist Research Award from the Northwestern University Feinberg School of Medicine Department of Pediatrics.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Declaration of Competing Interest: The authors report no conflict of interest relevant to this article.

REFERENCES

  • 1.Kramer CK, Campbell S, Retnakaran R. Gestational diabetes and the risk of cardiovascular disease in women: a systematic review and meta-analysis. Diabetologia. 2019;62(6):905–914. [DOI] [PubMed] [Google Scholar]
  • 2.Retnakaran R, Shah BR. Mild glucose intolerance in pregnancy and risk of cardiovascular disease: a population-based cohort study. Canadian Medical Association journal. 2009;181(6–7):371–376. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Rich-Edwards JW, Fraser A, Lawlor DA, Catov JM. Pregnancy characteristics and women’s future cardiovascular health: an underused opportunity to improve women’s health? Epidemiologic reviews. 2014;36:57–70. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Bellamy L, Casas JP, Hingorani AD, Williams D. Type 2 diabetes mellitus after gestational diabetes: a systematic review and meta-analysis. Lancet (London, England). 2009;373(9677):1773–1779. [DOI] [PubMed] [Google Scholar]
  • 5.Di Cianni G, Lencioni C, Volpe L, et al. C-reactive protein and metabolic syndrome in women with previous gestational diabetes. Diabetes Metab Res Rev. 2007;23(2):135–140. [DOI] [PubMed] [Google Scholar]
  • 6.Lauenborg J, Mathiesen E, Hansen T, et al. 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. 2005;90(7):4004–4010. [DOI] [PubMed] [Google Scholar]
  • 7.Ajala O, Jensen LA, Ryan E, Chik C. Women with a history of gestational diabetes on long-term follow up have normal vascular function despite more dysglycemia, dyslipidemia and adiposity. Diabetes Res Clin Pract. 2015;110(3):309–314. [DOI] [PubMed] [Google Scholar]
  • 8.Verma A, Boney CM, Tucker R, Vohr BR. Insulin resistance syndrome in women with prior history of gestational diabetes mellitus. J Clin Endocrinol Metab. 2002;87(7):3227–3235. [DOI] [PubMed] [Google Scholar]
  • 9.Meyers-Seifer CH, Vohr BR. Lipid levels in former gestational diabetic mothers. Diabetes Care. 1996;19(12):1351–1356. [DOI] [PubMed] [Google Scholar]
  • 10.Retnakaran R, Qi Y, Connelly PW, Sermer M, Hanley AJ, Zinman B. 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. J Clin Endocrinol Metab. 2010;95(9):4345–4353. [DOI] [PubMed] [Google Scholar]
  • 11.Tarim E, Yigit F, Kilicdag E, et al. Early onset of subclinical atherosclerosis in women with gestational diabetes mellitus. Ultrasound Obstet Gynecol. 2006;27(2):177–182. [DOI] [PubMed] [Google Scholar]
  • 12.Anastasiou E, Lekakis JP, Alevizaki M, et al. Impaired endothelium-dependent vasodilatation in women with previous gestational diabetes. Diabetes Care. 1998;21(12):2111–2115. [DOI] [PubMed] [Google Scholar]
  • 13.Heitritter SM, Solomon CG, Mitchell GF, Skali-Ounis N, Seely EW. Subclinical inflammation and vascular dysfunction in women with previous gestational diabetes mellitus. J Clin Endocrinol Metab. 2005;90(7):3983–3988. [DOI] [PubMed] [Google Scholar]
  • 14.Kjos SL, Buchanan TA, Montoro M, Coulson A, Mestman JH. Serum lipids within 36 mo of delivery in women with recent gestational diabetes. Diabetes. 1991;40 Suppl 2:142–146. [DOI] [PubMed] [Google Scholar]
  • 15.Kim C, Cheng YJ, Beckles GL. Cardiovascular disease risk profiles in women with histories of gestational diabetes but without current diabetes. Obstet Gynecol. 2008;112(4):875–883. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Fraser A, Nelson SM, Macdonald-Wallis C, et al. Associations of pregnancy complications with calculated cardiovascular disease risk and cardiovascular risk factors in middle age: the Avon Longitudinal Study of Parents and Children. Circulation. 2012;125(11):1367–1380. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.HAPO Study Cooperative Research Group. Hyperglycemia and adverse pregnancy outcomes. New England Journal of Medicine. 2008;358(19):1991–2002. [DOI] [PubMed] [Google Scholar]
  • 18.HAPO Study Cooperative Research Group. The Hyperglycemia and Adverse Pregnancy Outcome (HAPO) Study. Int J Gynaecol Obstet. 2002;78(1):69–77. [DOI] [PubMed] [Google Scholar]
  • 19.International Association of Diabetes Pregnancy Study Groups Consensus Panel, Metzger BE, Gabbe SG, et al. International association of diabetes and pregnancy study groups recommendations on the diagnosis and classification of hyperglycemia in pregnancy. Diabetes Care. 2010;33(3):676–682. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Diagnostic criteria and classification of hyperglycaemia first detected in pregnancy: a World Health Organization Guideline. Diabetes Res Clin Pract. 2014;103(3):341–363. [DOI] [PubMed] [Google Scholar]
  • 21.Lowe WL Jr., Scholtens DM, Lowe LP, et al. Association of Gestational Diabetes With Maternal Disorders of Glucose Metabolism and Childhood Adiposity. Jama. 2018;320(10):1005–1016. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.HAPO Study Cooperative Research Group, Nesbitt GS, Smye M, Sheridan B, Lappin TR, Trimble ER. Integration of local and central laboratory functions in a worldwide multicentre study: Experience from the Hyperglycemia and Adverse Pregnancy Outcome (HAPO) Study. Clin Trials. 2006;3(4):397–407. [DOI] [PubMed] [Google Scholar]
  • 23.Yogev, Chen, Hod, et al. Hyperglycemia and Adverse Pregnancy Outcome (HAPO) study: preeclampsia. Am J Obstet Gynecol. 2010;202(3):255 e251–257. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Radaelli T, Farrell KA, Huston-Presley L, et al. Estimates of insulin sensitivity using glucose and C-Peptide from the hyperglycemia and adverse pregnancy outcome glucose tolerance test. Diabetes Care. 2010;33(3):490–494. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Friedewald WT, Levy RI, Fredrickson DS. Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge. Clinical chemistry. 1972;18(6):499–502. [PubMed] [Google Scholar]
  • 26.Expert Panel on Detection Evaluation and Treatment of High Blood Cholesterol in Adults. Executive Summary of The Third Report of The National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol In Adults (Adult Treatment Panel III). Jama. 2001;285(19):2486–2497. [DOI] [PubMed] [Google Scholar]
  • 27.Stone NJ, Robinson JG, Lichtenstein AH, et al. 2013 ACC/AHA guideline on the treatment of blood cholesterol to reduce atherosclerotic cardiovascular risk in adults: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines. Circulation. 2014;129(25 Suppl 2):S1–45. [DOI] [PubMed] [Google Scholar]
  • 28.Lloyd-Jones DM, Hong Y, Labarthe D, et al. Defining and setting national goals for cardiovascular health promotion and disease reduction: the American Heart Association’s strategic Impact Goal through 2020 and beyond. Circulation. 2010;121(4):586–613. [DOI] [PubMed] [Google Scholar]
  • 29.Mosca L, Benjamin EJ, Berra K, et al. Effectiveness-based guidelines for the prevention of cardiovascular disease in women--2011 update: a guideline from the American Heart Association. Journal of the American College of Cardiology. 2011;57(12):1404–1423. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Grundy SM, Stone NJ, Bailey AL, et al. 2018 AHA/ACC/AACVPR/AAPA/ABC/ACPM/ADA/AGS/APhA/ASPC/NLA/PCNA Guideline on the Management of Blood Cholesterol: A Report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines. Circulation. 2019;139(25):e1082–e1143. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.American Diabetes Association. 2. Classification and Diagnosis of Diabetes: Standards of Medical Care in Diabetes-2020. Diabetes Care. 2020;43(Suppl 1):S14–S31. [DOI] [PubMed] [Google Scholar]
  • 32.HAPO Study Cooperative Research Group. Hyperglycaemia and Adverse Pregnancy Outcome (HAPO) Study: associations with maternal body mass index. BJOG. 2010;117(5):575–584. [DOI] [PubMed] [Google Scholar]
  • 33.Brown MA, Lindheimer MD, de Swiet M, Van Assche A, Moutquin JM. The classification and diagnosis of the hypertensive disorders of pregnancy: statement from the International Society for the Study of Hypertension in Pregnancy (ISSHP). Hypertens Pregnancy. 2001;20(1):IX–XIV. [DOI] [PubMed] [Google Scholar]
  • 34.Arnett DK, Blumenthal RS, Albert MA, et al. 2019 ACC/AHA Guideline on the Primary Prevention of Cardiovascular Disease: A Report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines. Journal of the American College of Cardiology. 2019;74(10):e177–e232. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.R Core Team. R: A Language and Environment for Statistical Computing. Vienna, Austria: R Foundation for Statistical Computing;2016. [Google Scholar]
  • 36.Hosmer DW, Lemeshow S. Applied Logistic Regression. New York: Wiley; 2013. [Google Scholar]
  • 37.Lu Y, Jia Z, Su S, et al. Establishment of trimester-specific reference intervals of serum lipids and the associations with pregnancy complications and adverse perinatal outcomes: a population-based prospective study. Ann Med. 2021;53(1):1632–1641. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.United States Preventive Services Task Force. Lipid Disorders in Adults (Cholesterol, Dyslipidemia): Screening. https://www.uspreventiveservicestaskforce.org/Page/Document/RecommendationStatementFinal/lid-disorders-in-adults-cholesterol-dyslipidemia-screening#. Published 2013. Accessed January 14, 2020.
  • 39.Carroll MD, Kit BK, Lacher DA, Yoon SS. Total and high-density lipoprotein cholesterol in adults: National Health and Nutrition Examination Survey, 2011–2012. NCHS Data Brief. 2013(132):1–8. [PubMed] [Google Scholar]
  • 40.Mszar R, Gopal DJ, Chowdary R, et al. Racial/Ethnic Disparities in Screening for and Awareness of High Cholesterol Among Pregnant Women Receiving Prenatal Care. J Am Heart Assoc. 2020:e017415. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Nanna MG, Wang TY, Xiang Q, et al. Sex Differences in the Use of Statins in Community Practice. Circ Cardiovasc Qual Outcomes. 2019;12(8):e005562. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Group HSCR, Metzger BE, Lowe LP, et al. Hyperglycemia and adverse pregnancy outcomes. The New England journal of medicine. 2008;358(19):1991–2002. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1
NIHMS1786147-supplement-1.docx (74.4KB, docx)

RESOURCES