Gestational Diabetes Mellitus and Subsequent Risks of Diabetes and Cardiovascular Diseases: the Life Course Perspective and Implications of Racial Disparities

Abstract

Purpose of Review

Gestational diabetes mellitus (GDM) is one of the most common pregnancy complications worldwide and the prevalence is continuously rising globally. Importantly, GDM is not an isolated complication of pregnancy. Growing evidence suggests that individuals with GDM, compared to those without GDM, have an increased risk of subsequent type 2 diabetes (T2D) and cardiovascular diseases (CVD). Substantial racial and ethnic disparities exist in the risk of GDM. However, the role of race and ethnicity in the progression from GDM to T2D and CVD remains unclear. The purpose of the current review is to summarize recent research about GDM and its life-course impacts on cardiometabolic health, including 1) the peak time of developing T2D and CVD risks after GDM, 2) the racial and ethnic disparities in the risk cardiometabolic diseases after GDM, 3) the biological plausibility and underlying mechanisms, and 4) recommendations for screening and prevention of cardiometabolic diseases among individuals with GDM, collectively to provide an updated review to guide future research.

Recent Findings

Growing evidence has indicated that individuals with GDM had greater risks of T2D (7.4 to 9.6 times), hypertension (78% higher), and CDV events (74% higher) after GDM than their non-GDM counterparts. More recently, a few studies also suggested that GDM could slightly increase the risk of mortality. Available evidence suggests that key CVD risk factors such as blood pressure, plasma glucose, and lipids levels are all elevated as early as < 1 year postpartum in individuals with GDM. The risk of T2D and hypertension is likely to reach a peak between 3–6 years after the index pregnancy with GDM compared to normal glycemia pregnancy. Cumulative evidence also suggests that the risk of cardiometabolic diseases including T2D, hypertension, and CVD events after GDM varies by race and ethnicity. However, whether the risk is higher in certain racial and ethnic groups and whether the pattern may vary by the postpartum cardiometabolic outcome of interest remain unclear. The underlying mechanisms linking GDM and subsequent T2D and CVD are complex, often involving multiple pathways and their interactions, with the specific mechanisms varying by individuals of different racial and ethnic backgrounds. Diabetes and CVD risk screening among individuals with GDM should be initiated early during postpartum and continue, if possible, frequently. Unfortunately, adherence to postpartum glucose testing with either obstetrician or primary care providers remained poor among individuals with GDM.

Summary

A life-course perspective may provide critical information to address clinical and public health gaps in postpartum screening and interventions for preventing T2D and CVD risks in individuals with GDM. Future research investigating the racial- and ethnic-specific risk of progression from GDM to cardiometabolic diseases and the role of multi-domain factors including lifestyle, biological, and socio-contextual factors are warranted to inform tailored and culture-appropriate interventions for high-risk subpopulations. Further, examining the barriers to postpartum glucose testing among individuals with GDM is crucial for the effective prevention of cardiometabolic diseases and for enhancing life-long health.

Diagnosis and Epidemiology of GDM

Gestational diabetes mellitus (GDM) is defined as glucose intolerance with onset or first recognition during pregnancy. Although there is a general agreement that the screening and diagnosis of GDM should be conducted at gestational weeks 24–28, but no global consensus on the diagnostic criteria. The disagreements include whether to use the one-step approach ((i.e., 75-g 2-h glucose tolerance test (OGTT)) or the two-step approach ((i.e., a positive 50-g 1-h glucose challenge test (GCT) followed by 100-g 3-h OGTT test)) and what cut-off values to apply for the positive results. While the one-step approach is endorsed by the World Health Organization (WHO) and the International Association of Diabetes and Pregnancy Study Groups (IADPSG) [1], the two-step approach is recommended by the National Institutes of Health (NIH) and American College of Obstetricians and Gynecologists (AJOG) [2] and more commonly applied in the United States (U.S.)., and both approaches are endorsed by the American Diabetes Association [3]. Thus, it is important to consider the impact of different diagnostic criteria for GDM in different settings, countries, or regions when comparing studies on GDM.

GDM is characterized by pancreatic beta cell function being insufficient to meet the body’s insulin need, usually due to progressive insulin resistance developed during pregnancy [4]. The established risk factors for GDM are advanced maternal age, parity, pre-pregnancy overweight or obesity, previous delivery of macrocosmic newborns, history of GDM, family history of diabetes [5], lifestyle factors (i.e., unhealthy diet and low physical activity before and during pregnancy) [6, 7], and psychosocial factors [8–10].

Recent data have shown a substantial rise in the prevalence of GDM among individuals of various racial and ethnic backgrounds [11–15], and would likely further increase as the obesity rate is continually increasing globally. Globally, the prevalence of GDM ranged from 1.0% to 25.0%, with remarkable regional differences [16, 17]. In general, Southeast Asia, Middle East, and North Africa have high GDM prevalence, while the Americas, Africa and Western Pacific appeared in the middle, and Europe had the lowest GDM prevalence among all the WHO regions. Although such global differences could be caused by the different diagnostic criteria utilized by different countries and studies, the more important implication could be profound risk differences across racial and ethnic groups.

The racial and ethnic differences in the GDM risk are better evidenced in the studies that are conducted in the same countries and applied the same diagnosis criteria. In the U.S., the prevalence of GDM in 2016 was estimated to be 11.1% in Asian individuals, 6.6% in Hispanic individuals, 4.8% in non-Hispanic Black individuals, and 5.3% in non-Hispanic White individuals [18]. In a more recent study, the prevalence of GDM among the first live birth all increased from 2011 to 2019 but varied across racial and ethnic groups, with age-standardized rate of 10.3% in Asian individuals, 6.7% in Hispanic individuals, 5.6% in Black individuals, and 5.8% in White individuals in 2019 [19]. The above two studies both used national birth certificate data which are likely underestimating the GDM prevalence [20]. Other studies using the self-reported survey data or medical records reported higher rates. For example, using questionnaires from the Pregnancy Risk Assessment Monitoring System (PRAMS), the prevalence of GDM in 2010 was 16.3% in Asian individuals, 12.1% in Hispanic individuals, 10.5% in Black individuals, and 6.8% in White individuals [21]. Using the medical records from Kaiser Permanente of Northern California (KPNC), the GDM prevalence was lowest in White (4.2%) and Black individuals (4.4%), followed by other Hispanics (5.4%) and Mexicans (7.1%), and was highest among Asian such as Chinese (7.9%), Filipinas (9.6%), and Asian Indians (11.1%) [22]. Similar patterns of the racial difference were also been reported in Australia [23], UK [24, 25] and Sweden [26]. Thus, available evidence from difference countries and using various methods and data sources consistently indicates that that Asian, Hispanic, and other racial and ethnic groups have an evaluated risk of GDM compared to White individuals. However, the molecular mechanisms and upstream determinants for the racial disparities in GDM risk remain largely understudied.

Postpartum and Long-term Cardiometabolic Diseases After GDM

Pregnant individuals with GDM are at greater risk for other pregnancy complications such as hypertensive disorders of pregnancy (gestational hypertensive and preeclampsia) and cesarean delivery [27]. Although most of them regain normal glucose tolerance after delivery, individuals with GDM are still featured with a higher risk of metabolic disorders in post-partum and later life. In this review, we focused on the risk of overt type 2 diabetes, hypertension, CVD events, and CVD mortality after GDM and discussed them in separate sections.

Risk of Type 2 Diabetes After GDM

Individuals with GDM have an increased risk of developing impaired glucose tolerance (IGT) and T2D in their later life. This association has been repeatedly demonstrated in studies from different regions and countries, regardless of the study populations and the diagnosis criteria of GDM and T2D. According to the pooled estimates from 4 systematic reviews and meta-analyses, the risk of development of T2D subsequent to the index pregnancy ranged from 7.4 to 9.6 times greater in individuals with GDM than their non-GDM counterparts [28–31]. The risk of T2D after GDM is greater among individuals with older maternal age, family history of diabetes, and greater body mass index (BMI) during the follow-up, but varied widely by length of follow-up, race and ethnicity, as well as the geographic regions.

Risk of Hypertension After GDM

In the past two decades, several studies in North America, Europe, and Asia have investigated the association between GDM and hypertension later in life. A meta-analysis published in 2021 pooled the results from 48 studies and found that mothers with GDM had 2.47 mmHg (95% CI: 1.74–3.40) higher pooled systolic blood pressure (SBP) and 1.89 mmHg (95% CI: 1.32–2.46) higher diastolic blood pressure (DBP) than those without GDM [32]. It is worth noting that this meta-analysis only estimated the unadjusted associations without accounting for potential confounders. More recently, our team conducted another systematic review and meta-analysis of 15 cohort studies with 3,959,520 participants and 106,560 cases of hypertension. We found that individuals with GDM had a 78% (pooled RR = 1.78; 95% CI: 1.47–2.17) higher risk of hypertension later in life [33] (Liu X et al., 2024, Hypertension, under the revision review). Individuals with younger age at first birth, preeclampsia, or preterm deliveries have been reported as the risk factors of the progression from GDM to hypertension. However, the role of race and ethnicity is less clear.

Risk of CVD Events After GDM

More recently, attention has been made to investigate whether having GDM during pregnancy is a risk factor for CVD events and mortality. One systematic review and meta-analysis with 7 cohort studies published in 2018 found individuals with GDM had a 74% higher risk of CVD events later in life (pooled RR = 1.74; 95% CI: 1.28–2.35) than those without GDM [34]. The risk was slightly greater for coronary artery disease (pooled RR 2.09; 95% CI: 1.56–2.80) than stroke (1.25; 1.07–1.48). Another meta-analysis with 9 studies (8 cohorts and one case–control study) published in 2019 also reported that individuals with GDM had a 98% higher risk of CVD events later in life (1.98; 1.57–2.50) [35]. Interestingly, all the studies included in these two meta-analyses were conducted in North America or Europe, except one in Israel. A more recent one published in 2022 that included 15 studies estimated a slightly weaker association between GDM and CVD events than the previous meta-analyses, with a 45% (95% CI: 1.36–1.53) higher risk of CVD and cerebrovascular diseases in individuals with GDM as compared to those without GDM [36]. Still the association of GDM with risk of heart diseases (1.72; 1.40–2.11) was greater than that of cerebrovascular disease (1.40; 1.29–1.51). This most recent meta-analysis included one study in Karen, Israel, Iran, respectively, in addition to all other studies in North America or Europe.

A history of GDM was also associated with increased risks for specific types of CVD. In the recent meta-analysis in 2022, individuals with GDM had increased risks of coronary artery diseases (RR = 1.40; 95% CI: 1.18–1.65), myocardial infarction (1.74; 1.37–2.20), heart failure (1.62; 1.29–2.05), angina pectoris (2.27; 1.79–2.87), cardiovascular procedures (1.87; 1.34–2.62), ischemic stroke (1.49; 1.29–1.71), hemorrhagic stroke (1.44; 1.16–1.78), and venous thromboembolism (1.28; 1.13–1.46), compared to individuals without GDM [36].

Risk of Mortality After GDM

A few studies also investigated the association of GDM with mortality. In the U.S. Nurses’ Health Study II (NHS II), individuals with a history of GDM had a 25% greater risk of death compared to their counterparts without GDM [37]. The increased risk of death was persistent regardless of the subsequent development of T2D and largely attributable to CVD (RR = 1.59; 95% CI: 1.03–2.47). It is noteworthy that the majority of the participants in NHS II were Non-Hispanic Whites (~ 90%). In another large U.S. study with both White and Black (45%) individuals using data from the Collaborative Perinatal Project, individuals with GDM or IGT during pregnancy had a slightly higher risk of all-cause mortality (RR = 1.14; 1.00–1.30). Interestingly, this study found an increased risk only for diabetes related death (RR = 3.12; 95% CI: 2.22–4.37) among individuals with GDM but not for CVD-related death (0.93; 0.74–1.20) [38]. We are not aware of studies in other regions and countries that investigated the risk of mortality associated with GDM.

The Cardiometabolic Risk Trajectory and Peak Time After GDM

The first question we want to address is when individuals with GDM have the highest risk of developing T2D. This question can be addressed by studies reported absolute risks (e.g., number of disease cases developed in 100 people) and/or relative risks (RR). The absolute risk of T2D after GDM can be estimated in studies that only followed individuals with GDM over time and used cumulative incidence or rate of progression to T2D. The reported crude cumulative incidence of T2D after GDM varied from 2.6% to 93.4%, over studies followed 6.6 weeks to 30 years postpartum [16, 28, 29, 31, 39]. An early systematic review in 2002 reported that cumulative incidence of T2D after GDM increased markedly in the first 10 years and then slowed down afterward [31]. However, in a more recent meta-analysis in 2020, the pooled cumulative incidence of T2D among individuals with GDM was estimated as 9.2% (95% CI: 7.2%-12.3%) in studies that followed 1 to 5 years and increased to 16.2% (95% CI: 15.8%-16.5%) for studies that followed for 10 years or longer [40]. In another recent meta-analysis including 129 studies published in 2021, the cumulative incidence of T2D after the GDM increased in a more linear way as the study duration increased, with 12% higher (95% CI: 8%-16%) risk for each additional year after the index pregnancy [29]. This is not surprising because for almost all chronic diseases, the cumulative incidence increases with time.

Comparing the absolute risk of T2D after the GDM across studies from different populations may not be appropriate because some populations have an inherently higher risk of T2D than others. On the other hand, the RR of T2D, comparing individuals with GDM to those without GDM can better reveal the risk and timing of postpartum T2D among individuals with GDM. In two meta-analyses which used the 5 years interval after the index pregnancy, the pooled RR of T2D comparing individuals with GDM to those without GDM varied from 4.69 (95% CI: 2.84–7.75) in an early meta-analysis [30] to 17.06 (8.95–32.55) in a more recent one [40] among studies that followed less than 5 years. However, the pooled RRs in studies followed more than 5 years were similar in these two meta-analyses [9.34 (3.42–25.54) [30] vs.10.42 (5.68–19.22) for 5–10 years follow-up and 8.09 (4.34–15.08) > 10 years follow-up [40]. In another two meta-analyses that used 3 years as the interval, the pooled RR in studies followed less than 3 years varied from 4.82 (2.19–10.61) [41] to 11.0 (3.4–35.1) [29]; increased to a peak of 16.16 (9.96–26.24) and 18.2 (14.4–23.1) in studies followed between 3 and 6 years; attenuated to 6.57 (3.56–12.11) and 5.4 (3.8–7.7) in studies that followed between 6–10 or 6–9 years, respectively; and remained at 6.02 (1.61–22.51) and 5.8 (2.6–12.8) in studies followed longer than 15 [41] or 12 [29] years, respectively [41]. Thus, results from several systematic reviews and meta-analyses tend to suggest that the magnitude of the positive association of GDM with subsequent T2D risk increased over time, especially after the index pregnancy through six years postpartum but then attenuated and remained significant beyond 10 years after the index pregnancy, compared to their non-GDM counterparts. However, the RR had greater variations in early years postpartum and these meta-analyses only have aggregated data.

Such large variations of the RRs can be attributed to differences in the study samples (i.e., population-based or hospital-based) and diagnosis criteria of GDM and T2D. Moreover, the frequency of diabetes screening also plays an important role, as evidenced by the few studies that performed regular diabetes screening after the index pregnancy. For example, the SWIFT study in the U.S. followed 1,010 individuals with GDM but not T2D from 6–9 weeks postpartum to 2 years postpartum, with the T2D screened by annual OGTT tests. The cumulative incidence was 6.8% in year 1 and 11.8% in year 2 [42]. Adopting a similar design of annual OGTT testing, the Tianjin GDM observational study in China reported that compared to individuals without GDM, the RR of T2D among those with GDM was 13.0 in the follow-up of 4.40 years [43], and 8.90 when the follow-up extended to 10 years [44]. Taken together, available evidence seems to suggest that the risk of T2D peaked between 3–6 years after the index pregnancy.

Research regarding the risk trajectory of other CVD risk factors or events is limited. As shown by a 2021 meta-analysis [32], the SBP, DBP, glucose, and triglycerides in those with GDM were all elevated as early as < 1 year postpartum and remained elevated at 1–5 years, 5–10 years and > 10 years post-partum compared to individuals without GDM. The greatest magnitude elevations of both SBP and DBP were shown in 5–10 years after the index pregnancy, while the triglyceride level seems to peak in 1–5 years postpartum. However, in the recent meta-analysis with aggregated data (Liu X et al., 2024, under revision review), we also found the association of GDM with hypertension risk was slightly stronger in studies with a follow-up of < 5 years vs. 5–15 or ≥ 15 years (RR: 2.15 vs. 1.68 or 1.81).

Individuals with GDM also had a greater risk of dyslipidemia, shortly after the delivery. A Canadian study found GDM was an independent predictor of total cholesterol, LDL-cholesterol, and triglyceride levels measured 3 months after delivery [45]. In a U.S. study conducted in Portland, individuals with GDM and normal glucose tolerance had higher levels of total cholesterol (5.06 vs. 4.56 mmol/L), LDL-cholesterol (3.17 v. 2.57 mmol/L), and triglyceride (1.02 vs. 0.86 mmol/L), and lower HDL-cholesterol levels (1.53 vs. 1.73 mmol/L; p = 0.001) than those without GDM from 2 to 24 months after the index pregnancy [46]. A more recent study among Chinese individuals found a total of 38.5% of the 589 individuals with GDM developed dyslipidemia in 6–12 weeks postpartum [47].

In summary, available evidence tends to suggest the key CVD risk factors such as blood pressure, plasma glucose, and lipids levels all elevated as early as < 1 year postpartum in individuals with GDM. The risk of T2D also increases shortly after delivery and likely reach peak between 3–6 years after the index pregnancy. The risk of hypertension is also greater within the 5 years after the index pregnancy. Given both T2D and hypertension are strong risk factors for CVD events, risk screening among individuals with GDM may need to be offered within first year of postpartum and continue up to 5 or 6 years. Future research is needed to identify the optimal frequency and the cost-effective approaches of risk screening.

Racial Disparities of Type 2 Diabetes and CVD Outcomes After GDM

One question being debated for many years is whether the subsequent CVD risk after GDM varied by race and ethnicity. Although Asian, Hispanic, and Black individuals all have a higher incidence and recurrence rate of GDM than Whites [48, 49], it is unclear whether their risks of subsequent CVD disease after GDM are also greater than Whites.

Racial Disparities of Type 2 Diabetes After GDM

In a recent large UK study that followed individuals with GDM for 26 years, South Asian (RR = 2.09; 95% CI: 1.52–2.85) and Black individuals (1.65; 1.05–2.62) with GDM were more likely to develop T2D compared to White individuals with GDM [50]. In an Australia hospital-based cohort study among all individuals with GDM, Asians had 2.1 times (95% CI: 1.7–2.7) higher risk of developing T2D compared to Whites [51]. Another smaller Australian study also reported the crude incidence of T2D or IGT was higher in South Asian (69%), South East and East Asian (41%) than in White individuals (42%) [52]. In a most recent U.S. population-based study in New York with 22,338 individuals with GDM, the 8-year incidence rate of T2D was 18.5% in Black, 16.8% in South and Southeast Asian, 14.6% in Hispanic, 5.5% in East and Central Asian, and 5.4% in White individuals. Compared with White individuals, the RR was 4.0 (95% CI 2.4–3.9) in Blacks, 2.9 (2.4–3.3) in South and Southeast Asian, 3.3 (2.7–4.2) in Hispanics, and 1.0 (0.9–1.4) in East and Central Asian [53]. A recent meta-analysis in 2020 reported the cumulative incidence of T2D was 16.5% in studies with mixed race/ethnicity, 15.6% in studies with non-Whites, and 9.9% in studies with Whites [40]. Another meta-analysis published in 2021 with 129 studies also found the risk of T2D was 57% lower in studies with White individuals than those with non-White individuals [29]. Taken together, the absolute risk of developing subsequent T2D after GDM appeared to be lower in Whites than in other racial and ethnic groups. However, because these studies only compared the incidence of T2D across racial and ethnic groups among individuals with GDM, the results may only reflect the overall lower incidence of T2D among White individuals but not the relative risk of T2D associated with GDM among White individuals compared to other racial and ethnic groups.

Results from several meta-analyses provide suggestive evidence of a higher risk of progression from GDM to T2D among White individuals compared to other racial/ethnic groups, yet only with aggregated data on specific racial/ethnic groups (i.e., Asian, Black, Hispanic, mixed). In an early meta-analysis of 20 cohort studies published between 1960 and 2009 with at least 6 weeks of follow-up, the pooled RR of T2D after GDM was 12.76 (95% CI: 2.31–70.63) across 7 studies with Whites, followed by 7.44 (95% CI: 2.43–22.83) across 6 studies with non-Whites, and 6.36 (95% CI: 2.06–19.61) across 7 studies with mixed race/ethnicity [30]. In a more recent meta-analysis published in 2021 including studies until October 2019, 13 studies with the majority of participants of non-White European origin showed a smaller magnitude of the pooled RR of T2D after GDM (5.1; 2.6–9.9) than 32 studies with majority of White European origin (11.2; 9.0–13.9) [29]. Similar findings were reported in another recent meta-analysis published in 2021 including 39 cohort studies published between 2009 and 2019, the pooled RR of T2D after GDM was 7.18 (5.08–10.15) in 7 studies with Asians, 5.93 (0.75–46.92) in 3 studies with Hispanics, 7.85 (6.66–9.25) in 16 studies with mixed race/ethnicity, and 14.91 (10.84–20.51) in 12 studies with Whites [28]. Given the limited studies among Asians, none of these meta-analyses reported risk estimates across Asian subpopulations (e.g., Chinese, Korean, Japanese, Indian, Pilipino, Vietnamese, etc.).

Only a few studies that included multi-racial and ethnic groups allowed for a direct comparison in the relative risk of T2D after GDM with individual level data between individuals with and those without GDM; however, the results are contradictory. One early population-based cohort conducted in Ontario, Canada found White individuals with GDM had a greater risk of T2D (RR = 13.6; 95% CI: 13.2–14.0) than either Chinese (9.2; 8.1–10.3) or South Asian (9.6; 8.8–10.5) individuals compared with their counterparts without GDM [54]. A large U.S. study using Kaiser Permanente Southern California healthcare data (n = 77,666) with 15 years of follow-up showed that Black individuals with GDM had a highest risk of T2D (RR = 9.2; 95% CI: 7.2–11.7) comparing to those without GDM, followed by Hispanic (7.4; 6.6–8.2), White (7.0; 5.8–8.4), and Asian individuals (6.2; 5.1–7.6) [55]. Further adjustment for pre-pregnancy BMI attenuated the association but did not explain the observed racial differences. More recently, a U.S. population-based study linked hospital discharge and vital statistics data to the HbA1C registry data among 336,276 pregnant individuals in New York City [56]. With the 9 years of follow-up, the RR of T2D, comparing individuals with GDM to those without GDM, was 7.7 (95% CI: 6.1–9.6) in South/South East Asians, 10.0 (95% CI: 9.2–11.5) in Blacks, 12.2 (9.2–11.5) in Hispanics, 7.7 (6.1–9.6) in South/South East Asians, and 12.5 (10.0–15.6) in Whites, with significant difference between South/South East Asians and Whites (P < 0.001). It is worth noting that this study only used HbA1C levels to define diabetes, which is known to have a poor sensitivity and a high false negative rate. In contrast, in the Louisiana State University Hospital-Based Longitudinal Study among U.S. individuals (n = 19,998) who received healthcare from public hospitals and were followed up to 20 years, we found that the risk of T2D in individuals with GDM was highest among Asian (RR = 18.1; 0.27–205.00), followed by Hispanic (11.3; 4.37–29.4), Black (7.43; 6.34–8.72), and White individuals (4.63; 3.61–5.95) compared with their counterparts without GDM [57]. It is noteworthy that the sample size of Asians (n = 134) in our Louisiana study is small and the estimate is underpowered as indicated by the wide 95% CI. A recent study comparing two cohorts (the Coronary Artery Risk Development in Young Adults study, consisting of 1,066 Blacks and 792 Whites participants) in the U.S. and the Tianjin GDM observational study (1968 Chinese in China) reported that GDM was associated with a higher risk of T2D within 5 years postpartum in Chinese individuals (RR = 71.50; 95% CI: 9.76–525.00) than in Black individuals (9.29; 2.23–38.7) [44]. However, the risk within 10 years was highest among White individuals (52.60; 5.73–462.00), followed by Blacks (9.98; 4.45–22.4) and Chinese individuals (8.96, 4.39–18.3) [44].

Racial Disparities of CVD Risk Factors or Events After GDM

Studies that investigated the racial difference in hypertension and other cardiovascular risk factors, or CVD events after GDM are limited and the results are mixed. In a large UK study among 9,118 individuals with GDM, South Asian (RR = 0.85; 95% CI: 0.40–1.77) and Black individuals (1.35; 0.60–3.02) did not have a higher risk of developing hypertension than White individuals [50]. However, this study did not compare the hypertension risk in individuals with GDM to those without GDM, thus only suggesting the absolute risk of hypertension after the GDM. In a U.S. study using the Massachusetts General Hospital medical record data among 4,110 individuals who delivered between 1998 and 2007, GDM was associated with an elevated risk of hypertension (RR = 1.75; 95% CI: 1.28–2.37) among all participants; However, the risks were greater among Hispanic (3.25; 1.85–5.72) and non-Hispanic White individuals (1.68; 1.10–2.57) as compared to Black (0.91; 0.37–2.45) and Asian individuals (1.04; 0.05–20.71) [58]. Notably, this study was possibly underpowered to examine the Asian and Black individuals (i.e., the sample size < 100). In our recent meta-analysis with aggregated data, we found the association between GDM and subsequent hypertension was slightly stronger in studies conducted in Asian countries vs. North American or European countries (2.36 vs. 1.76 or 1.82) (Liu, 2024, Hypertension).

In the U.S. GENNID study (332 with GDM and 662 without GDM), history of GDM was associated with higher risk of CVD (RR = 1.59; 95% CI: 1.02–2.49) among all participants [59]. When stratifying by race/ethnicity, the risk was highest among Hispanics (2.91; 1.06–8.02), followed by Whites (1.62; 0.84–3.12) and Blacks (1.27; 0.62–2.61). A more recent study among 917 individuals who participated the U.S. National Health and Nutrition Examine Survey (NHANES) 2007–2016 found that Black individuals with GDM had a higher risk of any chronic cardiometabolic conditions (any of CVD, T2D or hypertension) compared to White individuals (RR = 2.40, 95% CI: 1.28–4.50) [60]. However, both GDM and CVD were self-reported in these two studies and GENNID was among individuals with a family history of diabetes. Finally, the U.S. CPP study with 46,042 individuals reported that the association of GDM with all-cause mortality was similar between Black individuals (RR = 1.16; 95% CI: 0.96–1.39) and White individuals (RR = 1.16; 95% CI: 0.97–1.39) [38].

In summary, available evidence suggests that the risk of cardiometabolic diseases including type 2 diabetes, hypertension, and CVD events after GDM varies by race and ethnicity. However, whether the risk is higher in certain racial and ethnic groups and whether the pattern may vary by the postpartum cardiometabolic outcome of interest remain unclear. Evidence from the meta-analyses should be interpreted with caution due to the potential ecological biases and different confounders adjusted by individual studies. Only a few studies included multi-racial/ethnic populations that can estimate the RR of T2D, hypertension, or CVD events after GDM which allowed the direct comparison across racial and ethnic groups. However, the limited sample size of the racial and ethnic minority groups is a concern. Most studies do not have adequate statistical power to make the meaningful comparisons. In addition, whether the racial and ethnic disparities in the progression from GDM to cardiometabolic diseases are due to genetics, environment, lifestyle or other differences across racial and ethnic groups warrant further investigation. Understanding these disparities is essential for developing targeted interventions and providing personalized health care for at-risk populations.

Biological Plausibility and Underlying Mechanisms

Although the increased CVD risk associated with GDM can be partially attributable to the development of T2D, studies suggest that other pathways independent of T2D exist. In two meta-analyses, GDM was associated with 9% (RR = 1.09, 95% CI: 1.06–1.13) [36] or 56% greater risk of subsequent CVD events (1.56; 1.04–3.39) even among individuals who did not develop T2D [35]. In a recent large-scale study using the UK biobank data with a mediation analysis, overt diabetes explained 23%, hypertension explained 11%, and dyslipidemia explained 10% of the association between GDM and overall cardiovascular outcomes [61]. Another population-based cohort in Denmark also estimated that diabetes explained about 23.2% of the association between GDM and later CVD risk [62]. Thus, GDM may be associated with underlying vascular changes or other mechanisms that adversely affect cardiovascular health independent of diabetes. These pathways could include high blood pressure, systemic inflammation, dyslipidemia, endothelial dysfunction, and other vascular damages (Fig. 1).

Fig. 1.

Proposed pathways through which the risk of subsequent CVD increased following GDM

GDM and T2D

GDM and T2D share some common etiology and pathophysiologic, including both insulin resistance and impaired β-cell function. Although the etiology and pathogenesis of GDM is not fully understood, most individuals with GDM appear to have β-cell dysfunction that occurs on a background of chronic insulin resistance [63, 64]. Pregnant individuals have increased insulin secretion (200–250%) and decreased insulin sensitivity (40–50%) compared to their non-gravid state [65]. In some individuals, β-cell function cannot compensate for the increased insulin requirements over pregnancy, and as a result they develop GDM or hyperglycemia [66, 67]. Fasting plasma glucose [68] and insulin levels [69, 70] during pregnancy and postpartum are strong predictors for subsequent T2D risk among individuals with GDM, suggesting that the progression from GDM to T2D is closely related to the progressive β-cell function failure to compensate for ongoing insulin resistance. It has also been hypothesized that additional insulin resistance postpartum should worsen the β-cell function among individuals with GDM [71]. Thus, amelioration of insulin resistance could stabilize or improve the β-cell function and reduce the T2D risk among individuals with GDM. In other words, reducing the insulin demanding by improving insulin sensitivity could greatly lower the risk of developing T2D among individuals with priori GDM. Indeed, results from the TRIPOD trial have shown that individuals with GDM who received troglitazone had a greater reduction (~ 56%) in T2D risk compared to their counterparts who received the placebo during a median follow-up of 30 months, mostly attributed to the increase of insulin sensitivity during the first 3 months [72]. Further, 86 participants who did not develop T2D at the end of the trial were enrolled in an open-label observational PIPOD study and received the pioglitazone treatment for up to 3 years [73]. The risk of diabetes was lower in individuals with the greater reduction in insulin output. The evidence across the TRIPOD and PIPOD studies revealed that thiazolidinedione drugs may enhance insulin sensitivity, reduce insulin secretory demands, and preserve pancreatic β-cell function, which in turn are all associated with a lower risk of T2D. These findings also support the concept that insulin resistance causes, or at least contributes to, the β-cell failure in individuals with GDM and provide the rationale for focusing on β-cell rest for preventing T2D risk.

It is worth mentioning that the TRIPOD and PIPOD studies only included Hispanic individuals with prior GDM. The similar patterns were observed in studies with a very small sample size (n < 20) in Blacks [74] and Whites [75]. Data on other racial and ethnic groups including Asians are sparse. While impaired β-cell function and insulin resistance before pregnancy are both important in the pathogenesis of progression of GDM to T2D, the relative contributions of these two factors may vary across different racial and ethnic groups. The precise underlying mechanisms for the progression to insulin resistance postpartum in individuals with GDM remain to be elucidated. Several pathways have been postulated, including increased adiposity [76], decreased levels of adiponectin, increased inflammation [77, 78], oxidative stress [79], and mitochondrial dysfunction [80], and alterations in the insulinsignaling pathway [81, 82]. Mechanistic studies are warranted to understand the racial differences in more detail, along with influences of genetics, environment, lifestyle, and other factors that may contribute to these differences.

GDM, Hypertension and Other Vascular Dysfunction

It is plausible that the observed association between GDM and hypertension reflects pre-existing insulin resistance, as insulin resistance is linked with both GDM and hypertension [83, 84]. GDM may also be associated with hypertension directly through vascular damage [85] or indirectly through pregnancy-induced hypertension [86, 87]. Moreover, GDM apparently increases progression to subclinical atherosclerosis, even in the absence of overt metabolic abnormalities. For example, GDM has been associated a higher risk of coronary artery calcification [88] and carotid artery intima media thickness ( a subclinical measure of early atherosclerosis) [89], independent of hyperglycemia and adiposity.

In summary, the underlying mechanisms linking GDM and subsequent T2D and CVD diseases are complex, and often involve multiple pathways and intercorrelations and interactions. Individual risk profile before pregnancy, its interplay with placental development and functioning, as well as the level and persistence of insulin resistance, chronic subclinical inflammation, dyslipidemia, endothelia dysfunctions, and other vascular damages may all play important roles. The precise molecular level mechanisms remain to be elucidated by mechanistic studies. It is important to note that the specific mechanisms may vary by individuals of different racial and ethnic backgrounds. Additionally, lifestyle factors such as diet, physical activity, and weight management play crucial roles in the long-term health outcomes of individuals with GDM.

Recommendations for Postpartum CVD Risk Reduction Among Individuals with GDM

Growing evidence has suggested that individuals with GDM are at higher risk of developing T2D, hypertension and other CVD events later in life. These individuals can show clinically manifested CVD abnormalities as early as one year postpartum. Therefore, postpartum screening for CVD risk factors and lifestyle modifications are crucial for reducing the future CVD risk in this population.

Given the well-established higher risk of T2D after GDM, guidelines from the ADA [90] and American College of Obstetricians and Gynecologists (ACOG) [91] both recommend that all individuals with recent GDM undergo a 75-g OGTT at 4 to 12 weeks postpartum. Individuals with a history of GDM should be screened for diabetes every 1 to 3 years with testing frequency similar to those with prediabetes. Those who are identified to have prediabetes should be offered intensive lifestyle intervention and/or pharmacological intervention with metformin [90]. A recent meta-analysis has shown that lifestyle intervention among individuals with GDM initiated within 3 years after delivery was highly effective and led to a 43% reduction in the long-term risk of T2D [92]. In addition, lactation, weight loss, exercise, and healthy diet have been identified as protective factors for reducing the future T2D risk after a pregnancy complicated by GDM. Thus, health education or behavioral therapies should be incorporated to both prenatal and postpartum care for individuals with GDM.

The 2020 updated American College of Cardiology guidelines for primary prevention of cardiovascular disease acknowledge GDM as an unique CVD risk factor related to female sex and recommend that individuals with GDM undergo CVD risk screening within 3 months postpartum [93]. However, the exact frequency of screening for CVD risk factors after 3 months postpartum and the type of diagnostic testing (e.g., laboratory and/or imaging markers) remain to be defined.

Unfortunately, adherence to the postpartum visits and glucose tests among individuals with GDM by either obstetrician or primary care providers remained poor, with only 16% to 23% of individuals with GDM undergoing postpartum diabetes testing within one year [94, 95]. Barriers to the adherence to diabetes screening and preventive care after GDM include adverse social determinants of health, practical obstacles to patient engagement in intensive lifestyle modification (e.g., limited childcare and social support, emotional stress, financial barriers), inadequate knowledge about the lifetime risk of T2D, limited or unequal access to postpartum care, and an untimely transition to primary care [95]. Thus, individualized care plans should be developed based on specific health status, risk factors, and preferences of each birthing person. A collaborative approach involving healthcare providers, dietitians, social workers, and other support services is essential for successful postpartum CVD risk reduction among individuals with a history of GDM.

Conclusions

GDM in large part represents the detection of preexisting defects in β-cell compensation for chronic insulin resistance. Therefore, individuals with GDM have almost 10 times higher risk of developing T2D and 1.5 to 3 times higher risk of hypertension and other CVD events compared to those without GDM. The elevated CVD risk among individuals with GDM is supported by the evidence of suggesting GDM and CVD share common pathophysiology and risk factors. Thus, individuals with GDM should be considered as a high-risk group for future cardiometabolic diseases. The risk screening should be initiated early postpartum and continue, if possible, every year. In addition, substantial racial and ethnic disparities exist in the risk of GDM, however, the role of race and ethnicity in the progression from GDM to subsequent T2D and CVD remains unclear. Future research with racial and ethnic diverse populations with adequate sample sizes are warranted to investigate the racial- and ethnic-specific risk of progression from GDM to subsequent cardioembolic diseases. Further, investigation on the role of multi-domain factors (i.e., lifestyle, biological, and sociocontextual factors) in the risk progression is needed to better inform preventive strategies targeting upstream factors. Nevertheless, individuals with GDM should be considered as a high-risk population for future cardiometabolic diseases and should be followed up with regular glucose testing and CVD risk factors screening starting from early postpartum. Risk prevention strategies for this population can include efforts targeting at a reduction in the risk of T2D, in addition to other efforts including managing optimal body weight, blood pressure, blood lipids, and other CVD risk factors. Ideally, such efforts will be considered and integrated into the current clinical care in both prenatal and postpartum care. Recognizing the lifetime risks and the risk trajectory associated with GDM requires greater awareness by clinicians, patients, and researchers. Future investigations will be needed to address how to perform risk classification or stratification for postpartum cardiometabolic diseases among individuals with GDM by leveraging multi-domain factors including lifestyle, biological, and sociocontextual factors.

Data Availability

No datasets were generated or analysed during the current study.