Abstract
Background:
Metabolic syndrome (MetS) is associated with a history of gestational diabetes (GDM), hypertensive disorders of pregnancy (HDP), and preterm birth (PTB), but it is unclear whether this association is due to the pregnancy complication(s) or prepregnancy/early pregnancy confounders. The study examines the association of GDM, HDP, and PTB with MetS 2–7 years later, independent of early pregnancy factors.
Materials and Methods:
Large, diverse cohort of nulliparous pregnant people with singleton gestations enrolled during their first trimester and who attended a follow-up study visit 2–7 years after delivery. The longitudinal cohort was recruited from eight medical centers across the United States. Using standardized protocols, anthropometry, biospecimens, and surveys were collected at study visits and pregnancy outcomes were abstracted from medical records. We estimated the relative risk of prevalent MetS at the follow-up study visit for participants with GDM, HDP, or PTB (vs. no complications), adjusting for early pregnancy age, body mass index, self-reported race/ethnicity, insurance type, and smoking status.
Results:
Of 4,402 participants, 738 (16.8%) had MetS at follow-up: 13.1% (441/3,365) among those with no complications, and 27.9% (290/1,002) among those with complications. MetS occurred in 39.0% of GDM (73/187, adjusted relative risk [aRR] = 1.75; 95% confidence interval [CI] 1.42–2.16); 29.2% of HDP (176/603, aRR = 1.49; 95% CI 1.27–1.75); and 29.7% of PTB (113/380, aRR = 1.78; 95% CI 1.49–2.12). Those who had both HDP and PTB (n = 113) had an aRR = 1.95 (95% CI 1.50–2.54).
Conclusions:
People whose pregnancies were complicated by GDM, HDP, or PTB are at a higher risk of MetS within 2–7 years after delivery, independent of early pregnancy risk factors. The highest MetS risk follows pregnancies complicated by both HDP and PTB.
Introduction
Individual trajectories of cardiovascular risk evolve over the life course and may be shaped, in part, by life events, including pregnancy. 1,2 Adults with Metabolic Syndrome (MetS), as defined by standard criteria (waist circumference, dyslipidemia, glucose intolerance, and hypertension), have an increased risk of atherosclerotic cardiovascular disease, diabetes mellitus, and mortality. 3,4 MetS is increasingly common in the United States, where an estimated 20% of adults 20–39 years of age meet the criteria for this diagnosis. 3,5 While the prevalence of MetS increases with age, it is less common among those with a higher income, and its prevalence varies by educational attainment. 6 Recent analyses of data from the National Health and Nutrition Examination Survey (NHANES) show the prevalence of MetS varies by race and ethnicity, with the highest rates in the United States among adults who reported “other” race/ethnicity or Hispanic ethnicity. 5
Beyond demographic risk factors, epidemiological studies have suggested that MetS is more common among individuals with a history of gestational diabetes (GDM) and other adverse pregnancy outcomes (APOs), including hypertensive disorders of pregnancy (HDP) and preterm birth (PTB), 7 although data confirming this association from prospective cohort studies that account for potentially important confounding factors are lacking. Most studies are limited by a number of factors, including small sample size, retrospective design, imprecise classification of pregnancy outcomes, and a lack of ethnic/racial diversity in the study population, which might limit the generalizability of findings. 7 –16 Moreover, the inability of most of these studies to account for prepregnancy or early pregnancy measures (such as obesity and socioeconomic factors), which may be shared by APOs and MetS, raises additional concerns about confounding bias. 17 –19
The objective of this study was to examine the association of GDM, HDP, and PTB with the development of MetS 2–7 years after delivery. We used prospectively collected data from the Nulliparous Pregnancy Outcomes Study: Monitoring Mothers-to-be Heart Health Study (nuMoM2b-HHS), a large, contemporary, cohort of geographically and racially diverse participants in the United States, who were followed from their first pregnancies. 20 The nuMoM2b-HHS provides the rich set of data needed to better understand the association of pregnancy outcomes with the later development of MetS, independent of early pregnancy factors.
Materials and Methods
Study participants
We used data from 10,038 participants with singleton pregnancies, who were enrolled into the Nulliparous Pregnancy Outcomes Study: Monitoring Mothers-to-Be (nuMoM2b). Participants were enrolled during their first trimester of pregnancy, between 6 weeks and 0 days and 13 weeks and 6 days of gestation, from eight sites across the United States.
Pregnancy measures and survey data were collected at study visits during each trimester and at delivery. 21 Of the total cohort, 7,003 were recontacted for at least one follow-up interview as part of the nuMoM2b-HHS, and 4,508 returned for a follow-up study visit between 2 and 7 years after their delivery. Details of study methods for this visit have been published. 20 The demographic characteristics of nuMoM2b-HHS participants were representative of the original cohort. Rates of APOs in the nuMoM2b-HHS cohort were also similar to those in the original cohort. 22 This study was approved by the institutional review board at each participating recruitment site.
Figure 1 provides details of participant inclusion. In this analysis, we included nuMoM2b participants attending a follow-up study visit 2–7 years after delivery, who did not experience an index pregnancy loss before 20 weeks or termination (n = 4,484). 21 Henceforth, we will use the term “mothers” to align with terminology used in previous literature, but we recognize that “mothers” includes birthing people of all genders.
This flow diagram provides the details of participant inclusion.
Measures, exposures, and covariates
Anthropometric measures, biospecimens, and surveys were collected at study visits using standard protocols as previously described. 20 The primary exposures for this analysis were GDM, HDP, and PTB, as documented during nuMoM2b using a priori criteria and validated by chart abstraction. GDM was defined as occurring when an individual without preexisting diabetes mellitus had documented abnormal values during a glucose screen or glucose tolerance test. 23 HDP included cases of preeclampsia (preeclampsia with or without severe features, or superimposed preeclampsia or eclampsia regardless of the timing of onset) and antepartum gestational hypertension. We considered HDP overall and by timing of delivery (preterm vs. term). PTB was defined as delivery before 37 completed weeks of gestation and was further stratified by whether it was spontaneous or indicated. 22 Participants with stillbirths were rare (N = 16) and were not excluded from the reference or APO groups.
The primary outcome of MetS was identified at the follow-up visit 2–7 years after delivery using the American Heart Association standard definition, which requires meeting three or more of the following criteria: (1) waist circumference >35 inches (88 cm) for non-Asians and >31.5 inches (80 cm) for Asians; (2) triglycerides >150 mg/dL or medication treatment for high triglycerides; (3) high-density lipoprotein (HDL) <50 mg/dL or medication treatment for low HDL; (4) a serum glucose ≥100 mg/dL or a diagnosis of diabetes mellitus; and (5) systolic blood pressure ≥130 mmHg or diastolic blood pressure ≥85 mmHg, or medication treatment for hypertension. 24
Early pregnancy MetS was defined using blood pressure, waist circumference, and glucose and lipids collected at the early pregnancy visit based on the same criteria, although this definition is not validated for measures collected during pregnancy as it is common for lipids, including triglycerides and HDL, to increase during the first trimester. 25
Waist circumference over the iliac crest and height were measured using a nonstretch study measuring tape. Weight was measured using balance beam or digital scales. Blood specimens were taken using venipuncture, stored at the study's central biorepository, and assayed in batch at the core laboratory (Lundquist Institute) for triglycerides, HDL, and glucose. Most (86.9%) specimens were fasting. Blood pressure was measured according to a standardized research protocol using (during pregnancy) aneroid sphygmomanometers and (during the nuMoM2b-HHS visit) the OMRON HEM-907XL with manual sphygmomanometry as a secondary measure in the event of device malfunction. Self-report during participant interviews at the nuMoM2b-HHS visit was used to identify the use of medication for high triglycerides, low HDL, or hypertension, as well as the diagnosis of diabetes mellitus.
Demographic characteristics were collected during the first pregnancy visit at 6 weeks and 0 days to 13 weeks and 6 days of gestation. These included age, marital status (single/never married; married; or separated, divorced, or widowed), completed education (less than high school, high school or general educational development certificate, some college with no degree, associate/technical degree, bachelor's degree, and degree beyond bachelor's), and health insurance (commercial/military, government, and self-pay/other).
Race and ethnicity were self-identified from standard options (race: White; Black, African American, or African descent; American Indian or Alaskan Native; Asian Indian; Other Asian [to be specified]; Native Hawaiian or Other Pacific Islander [to be specified]; and Other [to be specified]) (ethnicity: Hispanic or Latino origin or descent [yes/no]) and categorized as non-Hispanic White, non-Hispanic Black, Hispanic, Asian, and other (American Indian or Alaska Native, Native Hawaiian or other Pacific Islander, multiracial, or “other” self-reported race). Body mass index (BMI) was calculated in kg/m2, with a BMI 25–29.9 kg/m2 classified as overweight and 30 kg/m2 or higher as obese. Tobacco use during the prior month (categorized as “yes/no”) was self-reported.
Statistical analysis
The demographic and cardiovascular characteristics of the cohort were described using means and standard deviations (SD) for continuous characteristics and frequency and percent for categorical characteristics, overall and by primary exposure (GMD, HDP, and PTB). We used robust Poisson regression to estimate the crude and adjusted relative risk (aRR) and associated 95% confidence intervals (CIs) of prevalent MetS at the nuMoM2b-HHS visit, comparing the risk among participants with specific APOs or GDM to a reference group of participants with none of these outcomes (models A [unadjusted] and B [adjusted]) or to a reference group of participants without the specific outcome of interest (models C [unadjusted] and D [adjusted]). 26 The adjusted analysis included age, BMI, self-reported race/ethnicity, insurance type, and smoking status at the first nuMoM2b study visit. Participants with missing outcome data or missing covariate data were excluded from analyses involving those data.
We conducted an exploratory sensitivity analysis of incident MetS, excluding those who met the criteria for MetS based on the early pregnancy measures collected at the first pregnancy visit. Additional sensitivity analyses were performed to assess the effects of self-reported additional pregnancies, the time elapsed between the index pregnancy delivery and follow-up visit as model covariates, and used interaction terms to explore potential effect modification by self-reported race/ethnicity and socioeconomic status as measured by insurance type.
For APO-specific analyses, those with missing data for the APO of interest were excluded from that APO-specific analysis.
Analyses were performed using SAS 9.4 (SAS Institute, Inc., Cary, NC) using the GENMOD procedure.
Results
Early pregnancy characteristics and pregnancy outcomes
Tables 1 and 2 show the demographic characteristics and measures obtained at the first pregnancy visit overall and by pregnancy outcome. Details of these measures by subgroup of HDP and PTB are shown in Supplementary Appendix Tables SA and SB. The participants had a mean age of 27.0 years (SD 5.6). The cohort was racially diverse, and 61.5% were married, most had completed high school, a majority were covered by private health insurance, and 5.6% reported current tobacco use. The mean BMI was 26.6 kg/m2 (SD 6.5), with 24.1% classified as overweight and 23.2% as obese. A small proportion reported being prescribed medications to treat prepregnancy diabetes (2.3%), chronic hypertension (1.9%), and high triglycerides (0.1%). At the first study visit, 9.8% (425 of 4,328) of participants in the cohort met the standard criteria for MetS based on these early pregnancy measures.
Participant Demographic Characteristics At First Pregnancy Visit, According to Index Pregnancy Adverse Pregnancy Outcome or Gestational Diabetes
Characteristics at enrollment.
APO defined as HDP or PTB. Participants with stillbirths were rare and were not excluded from the reference or APO groups.
HDP defined as all cases of preeclampsia (mild, severe, or superimposed preeclampsia or eclampsia regardless of the timing of onset) and antepartum gestational hypertension.
APO, adverse pregnancy outcome; GDM, gestational diabetes; GED, general educational development certificate; HDP, hypertensive disorders of pregnancy; n, number in category; N, sample size; PTB, preterm birth; SD, standard deviation.
Anthropometric and Biological Measures of Metabolic Risk At First Pregnancy Visit, According to Index Pregnancy Adverse Pregnancy Outcome or Gestational Diabetes
Characteristics during early pregnancy.
APO defined as HDP or PTB. Participants with stillbirths were rare and were not excluded from the reference or APO groups.
HDP defined as all cases of preeclampsia (mild, severe, or superimposed preeclampsia or eclampsia regardless of the timing of onset) and antepartum gestational hypertension.
Elevated waist circumference defined as >31.5 inches (80 cm) for Asian women and >35 inches (88 cm) for non-Asian women.
BMI, body mass index; DBP, diastolic blood pressure; HDL, high density lipoprotein; MetS, metabolic syndrome; SBP, systolic blood pressure.
Overall, 22.8% (1,022 of 4,484) of participants met the criteria for GDM, HDP, or PTB during pregnancy. Specifically, 4.3% met the criteria for GDM, 13.6% HDP, and 8.6% PTB. Co-occurrence of outcomes was common. Among those with GDM, 19.4% (37 of 191) also had HDP and among those with HDP, 18.6% (114 of 612) also delivered preterm.
MetS at the follow-up study visit
On average, the follow-up study visit occurred 3.2 years (SD 0.9) after the nuMoM2b delivery; the visit timing did not differ by pregnancy outcome. Table 3 shows the anthropometric and biological measures overall and by GDM, HDP, and PTB. Participants had gained a mean of 7.4 lbs (SD 22.9) since their early-pregnancy weight, and 24.1% and 30.1% were now classified as overweight and obese, respectively. In addition, more people reported being prescribed medications to treat hypertension than at baseline. MetS criteria for hypertension were present in 13.2%, abnormal glucose or diabetes in 16.4%, and abnormal waist circumference in 62.4%. More than one-third (34.5%) met MetS criteria for low HDL, and 12.1% for high triglycerides.
Measures At the Follow-Up Study Visit, According to Index Pregnancy Adverse Pregnancy Outcome or Gestational Diabetes
APO defined as HDP or PTB. Participants with stillbirths were rare and were not excluded from the reference or APO groups.
HDP defined as all cases of preeclampsia (mild, severe, or superimposed preeclampsia or eclampsia regardless of the timing of onset) and antepartum gestational hypertension.
Elevated waist circumference defined as >31.5 inches (80 cm) for Asian women and >35 inches (88 cm) for non-Asian women.
Of those who attended the follow-up study visit, 4,402 had sufficient data to permit ascertainment of MetS. There was no association between pregnancy outcome and missing MetS status.
As shown in Table 3, 16.8% (738 of 4,402) of the participants included in the analytic sample met the criteria for MetS at the follow-up study visit. The presence of MetS was significantly more common among those whose pregnancy had been complicated by GDM, HDP, or PTB when compared to those without any of these complications. Specifically, 13.1% of those without any of these complications met criteria for MetS. In contrast, MetS was identified among 39.0% of those with a history of GDM, 29.2% of those with HDP, and 29.7% among those with PTB. The highest prevalence (48.5%) was observed among participants who had both PTB and HDP (data not shown).
The strength of association varied according to the pregnancy outcome (Table 4) when compared to those without GDM, HDP, or PTB. Relative risks adjusted for age, BMI, self-reported race/ethnicity, insurance type, and smoking status (Table 4, model B) were elevated for those with a history of GDM (aRR = 1.75; 95% CI 1.42–2.16), HDP (aRR = 1.49; 95% CI 1.27–1.75), and PTB (aRR = 1.78; 95% CI 1.49–2.12). The highest relative risk (aRR = 1.95; 95% CI 1.50–2.54) was observed among participants who had both PTB and HDP. When participants with specific pregnancy outcomes were compared to those without the specific outcome (Table 4, model D), estimates of association were mildly attenuated, but direction of association and statistical significance were unchanged.
Association of Adverse Pregnancy Outcomes and Gestational Diabetes with Prevalent Metabolic Syndrome 2–7 Years After Pregnancy
Model A: risk ratio for the comparison of specific APO or GDM versus no APO/GDM (absence of HDP, PTB, or GDM), unadjusted.
Model B: risk ratio for the comparison of specific APO or GDM versus no APO/GDM (absence of HDP, PTB, or GDM), adjusted for baseline age, BMI, race, insurance, and smoking status.
Model C: risk ratio for the comparison of specific APO or GDM versus absence of specific APO/GDM (absence of HDP, PTB, or GDM), unadjusted.
Model D: risk ratio for the comparison of specific APO or GDM versus absence of specific APO/GDM (absence of HDP, PTB, or GDM), adjusted for baseline age, BMI, race, insurance, and smoking status.
HDP defined as all cases of preeclampsia (mild, severe, or superimposed preeclampsia or eclampsia regardless of the timing of onset) and antepartum gestational hypertension.
APO defined as HDP or PTB. Participants with stillbirths were rare and were not excluded from the reference or APO groups.
CI, confidence interval; RR, relative risk.
Sensitivity analysis (Supplementary Appendix Table SC) showed these associations were largely unchanged when those with evidence of MetS at the time of the early pregnancy visit were excluded for the analysis. Additional sensitivity analysis (results not shown) indicated that these associations were not meaningfully different when controlling for the number of subsequent pregnancies or time from the index pregnancy delivery to the follow-up visit. There was no evidence of interaction of self-reported race/ethnicity or insurance status with the outcome.
Comment
Consistent with prior research, 8,27 we found that those whose pregnancies were complicated by GDM, HDP, and PTB were at an increased risk of MetS during the early postpartum years. We extend prior work in two ways. First, we demonstrate that the risk of MetS remains significantly elevated even after adjustment for potentially confounding covariates extant during early pregnancy, suggesting the association cannot be solely attributed to shared risk factors between APO and MetS, such as higher BMI. Second, we demonstrate that the magnitude of association with MetS depends upon the specific type of pregnancy complication and timing of birth.
Specifically, those with HDP, who delivered preterm, were found to be at the greatest risk of having MetS 2–7 years after delivery. Taken together, these results reinforce that pregnancy complications are indicators of future cardiovascular risk and the importance of the early postpartum years as critical periods when high-risk individuals can be identified and early interventions could be initiated in an effort to influence long-term health.
Because this analysis was based on data gathered prospectively, we were able to incorporate early pregnancy data and explore both prevalent and incident MetS. In the absence of prepregnancy data, it has been challenging to be certain that the association of pregnancy complications with later cardiovascular risk is not merely a function of confounding by other factors. Prior research into the association between pregnancy outcome and trajectories of cardiovascular risk has yielded variable findings. The HUNT study found people with preeclampsia had elevated cardiovascular risk factors before pregnancy when compared to those without preeclampsia, and the pregnancy did little to alter the cardiovascular trajectory. 19 In contrast, analysis of data from the CARDIA study 28 suggested that pregnancy altered the cardiovascular trajectories for those who delivered preterm. While this study cannot confirm a causal link, these prospective data provide evidence that APOs may influence long-term cardiovascular risk.
There are important limitations that should be considered when interpreting the findings from this analysis. Time varied considerably across the participants from delivery to the follow-up assessment. However, this is unlikely to explain the association between pregnancy complications and MetS, as the timing of follow-up did not differ by pregnancy outcome. In addition, not all specimens from the follow-up study visit were fasting, which may have led to misclassification; there is no evidence, however, that any misclassification was non-random.
In addition, applying the definition of MetS to measures collected during early pregnancy is not validated and likely leads to misclassification; therefore, the prevalence estimates based on that measure are not reliable. There were some differences in the characteristics of individuals in the nuMoM2b cohort and those who attended the follow-up HHS visit (data not shown but available on request); however, these differences were not pervasive and are unlikely to substantively bias the findings. 22 Finally, an observational study design cannot prove a causal connection between the pregnancy outcomes and the later development of MetS.
These findings reinforce the practice of incorporating obstetric information into cardiovascular risk assessment during the early postpartum years. Future research should identify potentially modifiable factors (e.g., breastfeeding, healthy weight and nutrition, sleep, physical activity) that could help at-risk individuals avert subsequent adverse cardiovascular health. 29
Conclusions
People whose pregnancies are complicated by GDM, HDP, or PTB are at a higher risk of MetS within the first 2–7 years after delivery, independent of early pregnancy risk factors such as BMI. Furthermore, MetS risk varies by the type of pregnancy outcome.
Footnotes
Author Disclosure Statement
Dr. Bairey Merz serves as a consultant/advisor for iRhythm. Dr. Saade is a consultant for GestVision. The authors report no other relevant financial or other conflicts of interest related to this project.
Funding Information
This study is supported by grant funding from the Eunice Kennedy Shriver National Institute of Child Health and Human Development: U10 HD063036; U10 HD063072; U10 HD063047; U10 HD063037; U10 HD063041; U10 HD063020; U10 HD063046; U10 HD063048; and U10 HD063053.
The study is also supported by cooperative agreement funding from the National Heart, Lung, and Blood Institute and the Eunice Kennedy Shriver National Institute of Child Health and Human Development: U10-HL119991; U10-HL119989; U10-HL120034; U10-HL119990; U10-HL120006; U10-HL119992; U10-HL120019; U10-HL119993; U10-HL120018; and U01HL145358.
Support was also provided by the National Institutes of Health: Office of Research on Women's Health through U10-HL-119991; Office of Behavioral and Social Sciences Research through U10-HL119991 and U10-HL119992; and the National Center for Advancing Translational Sciences—UL-1-TR000124, UL-1-TR000153, UL-1-TR000439, and UL-1-TR001108; and the Barbra Streisand Women's Cardiovascular Research and Education Program, and the Erika J. Glazer Women's Heart Research Initiative, Cedars-Sinai Medical Center, Los Angeles.
The content of this article is solely the responsibility of the authors and does not necessarily represent the official views of the National Heart, Lung, and Blood Institute, the National Institutes of Health, or the US Department of Health and Human Services.
Supplementary Material
Supplementary Appendix Table SAA
Supplementary Appendix Table SAB
Supplementary Appendix Table SAC
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
