The Association of Adverse Pregnancy Outcomes and Cardiovascular Disease: Current Knowledge and Future Directions

Abstract

Purpose of review

Adverse pregnancy outcomes are associated with increased risk for future cardiovascular disease. The goal of this review is to share what is currently known about the increased risk and to identify areas for future research.

Recent findings

Severe studies have identified a strong association between adverse pregnancy outcomes and cardiovascular disease such as heart failure, valvular disease, ischemic heart disease, stroke, hypertension, and metabolic syndrome. The recognition of this increased risk is reflected in recent changes in prevention guidelines. The guidelines now recognize sex-specific risks such as preeclampsia and preterm delivery and recommend incorporating a pregnancy history to identify them earlier. However, no robust risk prediction tools incorporating these pregnancy risk factors have been developed and validated. While smaller clinical trials have been performed in reducing cardiovascular risk factors in the postpartum timeframe, there remains a paucity of large-scale randomized clinical trials that continue to show a risk reduction in these women.

Summary

While there is increasing recognition of the long-term cardiovascular risks associated with adverse pregnancy outcomes, there remains a need for interventional studies aimed at reducing this risk and for incorporation of pregnancy risk factors into traditional cardiovascular risk prediction tools.

Introduction

Adverse pregnancy outcomes (APOs) are common and occur in 17–20% of all pregnancies in the USA [1, 2, 3•]. These include hypertensive disorders of pregnancy (HDP), such as preeclampsia and gestational hypertension, gestational diabetes mellitus (GDM), preterm delivery, and small for gestational age (SGA) infants. While APOs can result in short-term complications during pregnancy, APOs are increasingly associated with deleterious long-term complications, especially cardiovascular disease (CVD), such as coronary artery disease (CAD), stroke, or heart failure [4–6]. This has led to the hypothesis that pregnancy may act as a “stress test” for future CVD as it may unmask underlying risks of poor or suboptimal cardiovascular health in women.

Certainly, APOs and CVD share similar risk factors, including obesity, hypertension, metabolic syndrome, race, and low socioeconomic status (Fig. 1). It is unclear though, whether the increased CVD risk is related to these shared risk factors or whether other mechanisms and attributes contribute to the elevated long-term CVD risk associated with APOs. In this review, we explore the epidemiology for common APOs, what is currently known regarding their association with increased risk of CVD, and what remains to be learned.

Fig. 1.

Shared risk factors between adverse pregnancy outcomes and cardiovascular disease.

Preeclampsia

Hypertensive disorders of pregnancy (HDP) are a common cause of morbidity during pregnancy and include chronic hypertension, gestational hypertension, preeclampsia, preeclampsia superimposed on chronic hypertension, HELLP syndrome, and eclampsia. Among the common variations of HDP, preeclampsia carries the greatest morbidity and mortality risk, affecting 5–10% of all pregnant women [7–9]. It is defined by new onset hypertension (systolic blood pressure > 140 mmHg or diastolic blood pressure > 90 mmHg) and end organ dysfunction, diagnosed after 20 weeks of gestation [10]. The pathogenesis underlying preeclampsia remains elusive, but abnormal placentation, an imbalance in angiogenic factors, and heightened inflammation are thought to be key mediators. In most women, the overt clinical manifestations of preeclampsia resolve in the immediate postpartum period. However, numerous studies show an increased long-term risk for CVD in women with a history of preeclampsia [11–14].

The observed association between CVD risk and preeclampsia may in part be due to shared risk factors between preeclampsia and CVD. Diabetes mellitus, hypertension, obesity, and smoking all increase risk for preeclampsia and are known risk factors for CVD [15–19]. Cohort studies have also demonstrated higher triglyceride levels, higher levels of atherogenic particles like ApoE, and higher ApoB/ApoA1 ratios, conferring increased risk, among women with preeclampsia [18, 20]. In the years after preeclampsia resolution, women are at higher risk for CVD. For instance, within the first year postpartum, women with HDP have a very high risk of developing chronic hypertension, with hazard ratios between 11.6 (95% CI 10.4–12.8) and 24.5 (95% CI 21.8–27.6) reported [21], and while this decreases over time, the risk remains twice that for women with normotensive pregnancies more than two decades later [21, 22].

Similarly, there is about a two-fold higher risk of ischemic heart disease (IHD) and stroke, with HR 1.8 (95% CI 1.3–2.6) reported for IHD and 2.98 (1.11–7.96) for non-fatal stroke after 10–15 years of follow-up [11, 23•, 24]. Subclinical markers of coronary disease, such as coronary artery calcium score (CACS), are also higher in women with prior preeclampsia [25–27]. Aside from IHD, women with a history of early-onset preeclampsia also have greater later life CVD risk factors, including diabetes, insulin, triglycerides, total cholesterol, and metabolic syndrome [26, 28–31].

In the acute setting, preeclampsia is associated with peripartum cardiomyopathy, possibly due to increased anti-angiogenic factors [32, 33]. Specifically, levels of angiogenic markers, soluble fms-like tyrosine kinase-1 (sFLT1), and soluble endoglin (sENG) correlate with myocardial dysfunction during preeclampsia [34]. Aside from the acute setting, several longitudinal studies also demonstrated an association between incident heart failure and history of HDP, with some study reporting HR 1.7 (95% CI 1.04–2.06) [23•].

A study in women with preeclampsia 1–2 years postpartum showed left ventricular hypertrophy and dysfunction, while another study in older women with prior HDP showed abnormal diastolic parameters and larger left atrial size [35, 36]. Interestingly, a recent, large observational study also reported an increased risk for valvular disease, including aortic stenosis (HR 2.9; 95% CI 1.5–5.4) and mitral regurgitation (HR 5.0; 95% 1.5–17.1) in women with HDP [23•].

Of note, women with recurrent preeclampsia are at much higher risk for future subclinical and clinical CVD when compared to women with only one pregnancy affected by preeclampsia. These patients have increased carotid-intima thickness, abnormal diastolic echocardiographic parameters, and elevated left ventricular mass [37–41]. Women with recurrent preeclampsia have an increased risk of hypertension (RR 2.3; 95% CI 1.9–2.9) and IHD (RR 2.4; 95% CI 2.2–2.7) [42, 43]. Higher incident heart failure among women with recurrent preeclampsia compared to controls (9.83 vs. 1.67 per 10,000 person-years) has also been reported [42].

Gestational diabetes mellitus

GDM is diabetes mellitus that manifests during the second or third trimester of pregnancy and is distinct from type 1 or type 2 diabetes mellitus that occurs prior to pregnancy [44]. In the USA, GDM has become increasingly prevalent, paralleling the increased prevalence in obesity [45•, 46], and now affects approximately 6–9% of pregnant women in the USA [47].

Women with GDM are at increased risk of developing CVD risk factors after pregnancy compared to women without GDM. Specifically, those with GDM are more than 7 times more likely to develop type 2 diabetes mellitus (RR 7.43 [95% CI 4.79, 11.51]) [48] and have nearly a twofold higher risk of developing hypertension and hyperlipidemia (hypertension OR 1.88 [95% CI 1.34, 2.64]; dyslipidemia OR 1.76 [95% CI 1.28, 2.44]) [49–51]. The risk of type 2 diabetes mellitus and hypertension is even higher in those with maternal obesity and GDM vs. no obesity and no GDM adjusted HR of 2.4 [95% CI 1.6, 3.5] compared to women with either GDM or gestational hypertension alone (adjusted HR 1.4; [95% CI 1.2, 1.7]) [49, 52].

Additionally, GDM is a risk factor for subsequent CVD, as it is associated with an increased risk of major cardiovascular events (RR 1.98 [95% CI 1.57, 2.50]) independent of the development of type 2 diabetes mellitus [53•]. Not surprisingly, the risk of CVD among women with GDM is amplified in the presence of other APOs. For example, women with both GDM and gestational hypertension have higher risks of CVD (adjusted HR 2.4 [95% CI 1.6, 3.5]) compared to women with either GDM or gestational hypertension alone (adjusted HR 1.4 [95% CI 1.2, 1.7]).

Epigenetics, subclinical inflammation, and endothelial/vascular dysfunction have all been proposed as mechanisms to explain the association between GDM and early-onset CVD [54, 55]. Inflammation has been associated with early atherosclerosis among women with GDM, as elevated levels of inflammatory markers, including C-reactive protein and interlukin-6, and lower levels of anti-inflammatory biomarkers, including adiponectin, have been demonstrated in women with GDM compared to normoglycemic women [56]. Endothelial dysfunction may also play a role, and women with GDM have been found to have increased carotid artery intimal thickness [57].

Preterm delivery

Preterm delivery, defined as birth between 20 and 37 weeks of gestation causes significant neonatal mortality worldwide and is a common reason for antenatal hospitalization [58]. In addition, preterm birth is associated with future maternal CVD risk, as shown in a recent large meta-analysis of more than 5.8 million women and 338,000 preterm births which highlights increased risk of future composite CVD (RR 1.43; 95% CI 1.18–1.72), cardiovascular mortality (RR 1.81; 95% CI 1.55–2.10), CAD (RR 1.49, 95% CI, 1.38, 1.60), and stroke (RR 1.65, 95% CI, 1.51, 1.79) [59•].

The most important potential confounder is the presence of hypertensive disorders of pregnancy. Two studies have compared spontaneous preterm birth with medically indicated preterm birth. A Norwegian study found that all gestational lengths shorter than 39 weeks were associated with future elevated CVD mortality [60]. Interestingly, spontaneous preterm delivery was associated with increased risk for CVD with HR up to 2.4 (95% CI 1.7–3.4), and medically indicated preterm delivery was associated with an increased CVD risk with HR up to 6.2 (95% CI 4.2–9.3) [60]. While the elevated future CVD mortality in the medically indicated group can be partially explained by the high prevalence of hypertensive disorders of pregnancy in this group, it is notable that the future CVD risk remains elevated in the spontaneous preterm labor group as well. Similarly, a study of Scottish women reported an association between preterm delivery and future risk of IHD. Women who have had medically indicated preterm delivery are at higher risk for future IHD events (HR 1.81; 95% CI 1.61–2.04) and IHD death (HR 2.49; 95% CI 1.89–3.30). However, women with a history of spontaneous preterm delivery were also at increased risk for IHD events (HR 1.46; 95% CI 1.33–1.61) and IHD death (2.14; 95% CI 1.70–2.70) [61]. These studies suggest that PTD remains a risk factor for future CVD even when accounting for the prevalence of hypertensive disorders of pregnancy.

Social disparities and APOs

Importantly, APOs disproportionately affect racial and ethnic minorities, particularly in the USA. These disparities are present not only in the prevalence of specific APOs but also in the cardiovascular sequelae of APOs and maternal mortality [62–64]. For example, African American and Native American women experience the highest rates of preterm birth, while Asian, Hispanic, and Native American women have the highest rates of GDM [65–67]. For HDP, there are conflicting data regarding racial disparities in disease prevalence [68]; however, African American women are more likely to suffer from severe maternal morbidity related to preeclampsia, including stroke, peripartum cardiomyopathy and cardiac arrest [69].

In addition, low socioeconomic status (SES) is also a significant risk factor for APOs, and this association is independent of race or ethnicity [64, 70, 71, 72•]. However, it is a weaker predictor of APOs than racial/ethnic group identification, and improving SES for racial/ethnic minorities does not eliminate disparities in APO rates between racial/ethnic minorities and white women [71, 72•]. Etiologies for racial disparities in APOs are multifactorial, with prior studies implicating differences in inflammatory markers, genetics and epigenetics, endothelial function, and psychosocial factors—including racial discrimination at individual and systems levels and other underlying risk factors for APOs [62, 66, 73]. Additional research is needed to further clarify etiologies for these disparities to pinpoint potential targets for solutions (Table 1).

Table 1.

Future directions and key research questions

Areas of investigation	Future directions and key research questions
Preeclampsia	Are there specific proteomic, metabolomic, and epigenomic profiles that are correlated with increased risk for CVD and heart failure? What is the relationship between pro/anti-angiogenic factors and endothelial function? Does controlling postpartum and long-term blood pressure reduce CVD risk in women with a history of preeclampsia?
Gestational diabetes	Are there biomarkers that can predict future CVD in women with a history of GDM? What is the relationship between markers of inflammation and endothelial dysfunction and risk of future CVD in women with GDM? Do these markers help predict groups at greatest risk for developing CVD later in life? What is the role of anti-hyperglycemic medications in the prevention of CVD in women with GDM?
Preterm delivery	What are the mechanisms underlying the association between preterm delivery and CVD? Does prophylactic use of aspirin prevent spontaneous preterm delivery?
Pathophysiology	Are APOs causal of CVD pathogenesis or merely unmask a woman’s pre-existing CVD risk? What are the biologic mechanisms leading to future maternal CVD and heart failure after APOs? Are there pathophysiologic mechanisms underlying socioeconomic and racial disparities in APOs?
Risk prediction and interventions	What are the appropriate methods and frequency of screening for CVD and CVD risk factors after delivery? Who should be screened? Does reducing modifiable risk factors decrease CVD risk in women with history of APOs? Why do APOs fail to enhance predictive discrimination for CVD in risk prediction models despite their well-established associations with future CVD events? What is the best method to identify women with APOs at high risk for future CVD?

APO adverse pregnancy outcome, CVD cardiovascular disease

Role of aspirin in the prevention of APOs

Mechanisms and biologic plausibility of aspirin

Because of the many adverse effects, there is great interest in reducing the incidence and severity of APOs. In addition to addressing social determinants of health, aspirin therapy has been increasingly utilized during pregnancy, particularly in preventing morbidity from preeclampsia [74]. Aspirin is a non-steroidal anti-inflammatory drug that inhibits cyclooxygenase isoenzymes 1 (COX-1) and 2 (COX-2) [75]. The latter converts arachidonic acid to prostaglandins, which have vasodilatory effects and inhibit platelet aggregation. In platelets, however, prostaglandin H2 is converted into thromboxane A₂, which has the opposite effect of vasoconstriction and platelet activation [74, 75].

Aspirin preferentially inhibits COX enzymes in platelets rather than the endothelium, leading to inhibited thromboxane production without affecting prostacyclin production. One mechanism of preeclampsia is an imbalance in thromboxane A₂ and prostacyclin production, resulting in uterine spiral arterial remodeling, vasoconstriction, and uteroplacental insufficiency [75, 76]. Ratios of thromboxane A₂ to prostacyclin have been shown to be decreased after treatment with low-dose aspirin, which is proposed to be a mechanism by which aspirin may be beneficial in preeclampsia [76]. In addition to pathways of vasoconstriction, aspirin also decreases profibrotic factor activin A levels, which is associated with cardiac fibrosis and hypertrophy, as well as improve global longitudinal strain, suggesting possible mechanisms of prevention of cardiac dysfunction in women with preeclampsia [77].

Preeclampsia

Large randomized controlled trials of aspirin therapy report mixed results in women at elevated risk for preeclampsia or pregnancy-induced hypertension [83–86]. Two older trials failed to demonstrate differences in preeclampsia, perinatal death, preterm delivery, or intrauterine growth restriction (IUGR) with aspirin use [85, 86]. In contrast, the Aspirin for Evidence-Based Preeclampsia Prevention trial demonstrated that among 1776 women at high risk for preterm preeclampsia, aspirin reduced the incidence of preterm preeclampsia (OR 0.38 [95% CI 0.20, 0.74]) [84] The US Preventive Service Task Force (USPSTF) performed a systematic review investigating aspirin therapy in pregnancy in 2014 and reported a 24% RR reduction in preeclampsia (95% CI 0.62, 0.95), 14% reduction in preterm delivery (95% CI 0.76, 0.98), and a 20% reduction in IUGR (95% CI 0.65, 0.99) using pooled estimates [87•]. Aspirin was also noted to be relatively safe, with no change in placental abruption, perinatal mortality, or postpartum hemorrhage rates. Therefore, the USPSTF concluded that there is substantial net benefit of low-dose aspirin use and recommended starting low-dose aspirin therapy (81 mg/day) after 12 weeks of gestation for women at higher risk for preeclampsia [78], which is also endorsed by ACOG and the Society for Maternal and Fetal Medicine task forces [74] (Table 2).

Table 2.

Summary of indications for use of aspirin in pregnancy, shared pathophysiologic mechanisms, and emerging research

	Possible mechanisms	Society recommendations	Selected emerging research topics
Preeclampsia	Inhibits thromboxane production and vasoconstriction → reduces uteroplacental ischemia	ACOG: Low-dose aspirin for women at high risk for PE from 12 to 28 weeks until delivery [74] USPSTF: Low-dose aspirin after 12 weeks for women at high risk for PE [78]	Studies of biomarkers to understand the pathophysiologic benefits of platelet inhibition in PE and prevention of cardiac dysfunction [43, 77]
Preterm delivery		Not recommended in the absence of PE risk factors [74]	Newer trials and meta-analyses suggesting potential benefit [79, 80]
IUGR		Not recommended in the absence of PE risk factors [75]	Clinical models to predict preterm delivery to identify women who might benefit from aspirin [81, 82]

ACOG American College of Obstetricians and Gynecologists, PE preeclampsia, USPSTF United States Preventive Services Task Force, IUGR intrauterine growth restriction

Preterm delivery

In addition to preeclampsia, aspirin has been studied in the context of other APOs. Preterm delivery has been an outcome in many preeclampsia trials, with a few studies have reported the effect of aspirin on preterm delivery in women without risk factors for preeclampsia. In particular, the ASPIRIN trial, a recent large multi-national RCT, randomized 11,976 nulliparous women to low-dose aspirin between 6 weeks 0 days and 13 weeks 6 days of gestation and found an 11% relative reduction (95% CI 0.81, 0.98) in the risk of overall preterm delivery and a 25% relative reduction (95% CI 0.61, 0.93) in the risk of early preterm delivery < 34 weeks in women [79]. This study did not differentiate between spontaneous and iatrogenic preterm delivery. However, there were no differences in the rate of HDP or IUGR, the two main indications for iatrogenic preterm delivery, suggesting that the lower incidence of preterm delivery in the aspirin group may be due to lower rates of spontaneous preterm delivery. Additionally, there is evidence from other meta-analyses and preeclampsia trials that suggest a lower risk of spontaneous preterm delivery in women treated with aspirin [80, 88].

Intrauterine growth restriction

Finally, low-dose aspirin may also be effective at preventing IUGR, since both IUGR and preeclampsia share common pathophysiology underlying abnormal placentation [74]. However, clinical studies have reported conflicting results in women with small for gestational age fetuses [89, 90]. In contrast, patients with subsequent preeclampsia have consistently found a reduction in IUGR with aspirin therapy, especially when initiated prior to 17 weeks’ gestation [87, 91]. Ultimately, additional studies are needed to examine the mechanisms and effectiveness of aspirin in IUGR and to discover risk stratification tools, including biomarkers, that may identify women who are at higher risk of IUGR who may benefit from aspirin use (Table 1).

Interventions to reduce subsequent long-term CVD risk in women who have had APOs

Few clinical trials have been performed to investigate specific interventions to reduce long-term CVD risk in women with APOs. In one randomized controlled trial, postpartum women with a history of preeclampsia within the prior 5 years underwent an online intervention targeting modifiable risk factors, specifically diet and physical activity [92]. The intervention improved physical activity but did not decrease weight or blood pressure [92]. Since then, other clinical trials have started, including the RedCarRisk trial, which seeks to assess the efficacy of an intensive regular conditional workout on arterial stiffness in women after preeclampsia. (Table 1).

Because of the strong association between APOs with future CVD [93], the importance of early identification for women at risk of future CVD has been acknowledged by professional societies, including the AHA, ACOG, and the Society of Obstetricians and Gynecologists of Canada [93, 94•, 95]. Most recently, a joint presidential advisory from the AHA and ACOG urged that the postpartum time frame be considered a critical time for health promotion, cardiovascular health assessment, risk stratification, and focused counseling highlighting the need for screening and long-term follow-up [94•, 96]. More specifically, the ACOG Task Force on Hypertension in Pregnancy recommends periodic assessment for CVD risk factors (e.g., hypertension, hyperlipidemia, obesity, and diabetes) in women with a history of preeclampsia, especially for women with preterm or recurrent preeclampsia [10, 97•]. In addition, the recent 2019 American College of Cardiology (ACC)/American Heart Association (AHA) guideline on the Primary Prevention of Cardiovascular Disease has recommendations for women with HDP. The guidelines consider APOs to be CVD risk enhancers, which warrant further assessment of risk by obtaining a CACS or by considering intensified preventive interventions like statin therapy [98]. However, there are no specific recommendations on the frequency of screening. There are also no recommendations favoring chronic aspirin therapy.

Current knowledge gaps and future directions

Improvement in risk-prediction models

Current prevention guidelines for CVD have stressed the importance of pregnancy history in evaluating CVD risk, but few studies have supported the incorporation of pregnancy risk factors into risk prediction scores [99]. In particular, three studies have examined the incremental value of adding HDP, preterm delivery, IUGR, and low birth weight to well-established risk factors and have found no risk reclassification [35, 99–101]. In one study utilizing a Norwegian population-based registry, HDP and preterm delivery status did led to reclassification of CVD risk, but it was of minimal clinical significance [98]. In another study assessing a US cohort of women at low risk for CVD, incorporating HDP did not improve predictive discrimination despite being independently associated with increased 10-year CVD risk [99].

These studies used population-based registries and cohorts to create sample populations of parous women ≥ 40 years of age with no prior history of CVD, and thus a large proportion of women in these groups were beyond their reproductive years. Because the proportion of comorbidities, such as diabetes, obesity, dyslipidemia, and hypertension, increases with age, sampling of older women might have underestimated the risk attributed to APOs. In addition, APOs are not independent of each other and may occur simultaneously making it difficult to assess their individual risk. Hence, any future risk prediction modeling exercises should include risk factors that are independently highly predictive of CVD and should include populations that reflect the target population intended for screening (i.e., women of reproductive age) [97] (Table 1).

Cardiovascular dysfunction and atherosclerosis after APOs

For women who have had APOs, some studies reveal structural and functional cardiovascular changes as well as persistent anti-angiogenic, coagulopathic, and inflammatory states [35, 101].

Notably, one study in women with preeclampsia 5 years earlier are reported to have lower coronary flow reserve, higher carotid intima-media thickness, and higher high-sensitivity C-reactive protein values compared to normal controls [102]. Further, the Cardiovascular Risk Profile: Imaging and Gender-Specific Disorders (CREW-IMAGO) study has noted that 31% of women with previous preeclampsia aged 45–55 years had a CACS greater than zero, with 47% having coronary atherosclerotic plaques and 4.3% with significant coronary artery stenosis [26].

Of note, most of the studies assessing subclinical atherosclerosis have been performed in women of approximately 60 years, thus, by and large, 30–40 years after pregnancy. In addition, most of these findings are derived from single center studies. Thus, there should be future multicenter, phenomic and metabolomic studies that explore signals of early atherosclerosis, endothelial dysfunction, and early diastolic dysfunction in younger patients. These studies should also focus on the mechanistic link between APOs and cardiovascular dysfunction. In addition, the biological properties of aspirin, including immunomodulatory, anti-inflammatory, and antioxidant effects should be explored in future studies to further understand the mechanisms underlying potential therapeutic benefits in women who have had APOs (Table 1).

Behavioral interventions in women with APOs

Whether modifiable behavioral factors (e.g., exercise, weight loss) moderate CVD risk in women following an APO is an important question with clinical and public health implications. Women who have had a HDP and with higher BMIs have a higher risk of incident hypertension [103]. However, there are no large multicenter studies that have explored the effect of behavioral modifications on subclinical cardiovascular dysfunction. A potential study can assess the role of AHA’s Life’s Simple 7 recommendations on optimal control of CV risk factors, which target smoking, nutrition, exercise, weight loss, blood pressure, cholesterol, and diabetes, in the postpartum period.

Conclusions

Adverse pregnancy outcomes, including preeclampsia and HDP, GDM, and preterm delivery, are highly associated with future maternal CVD; however, it is unclear whether this is a causal relationship or whether APOs unmask shared risk factors with CVD. Unlocking the pathophysiologic link between an APO and the development of subclinical and overt CVD will help identify novel therapeutic targets and lifestyle recommendations for women with APOs. The role of aspirin in prevention of APOs and subsequent cardiovascular dysfunction will be important to explore as a possible therapeutic avenue. Timely recognition and aggressive risk factor modification and health promotion are vital, particularly among women with lower socioeconomic status. Future research should explore the efficacy of modifiable behavioral interventions in reducing CVD risk in women with APOs.