Pregnancy Complications as Indicators of Cardiovascular Disease Risk in Women: How Do We Tackle Cardiovascular Disease Prevention in Women Who Have Failed the Cardiac Stress Test of Pregnancy?

In the 1980s and 1990s, cardiovascular disease (CVD) epidemiology took a paradigm‐shifting turn after David Barker proposed that cardiometabolic diseases in adulthood might have causal origins in fetal life. ¹ Although Barker's interpretations of his study results have since been contested, and the developmental origins of disease hypothesis has evolved through several iterations, the concept of life‐course epidemiology, which is the idea that events early in life can be associated with increased risks of disease decades later, has revolutionized chronic disease epidemiology. ²

Against this backdrop, a handful of prescient researchers in the early 2000s began to consider that the dysregulated cardiometabolic profiles exhibited by women with pregnancy complications such as preeclampsia and gestational diabetes might have long‐term consequences for maternal health. Two decades later, it is clear that women with a history of preeclampsia, preterm delivery, pregnancy loss, or gestational diabetes have higher rates of CVD than women who did not experience these complications. ³ , ⁴ , ⁵ , ⁶

In hindsight, these findings are hardly surprising. The stresses on a woman's body during pregnancy have been compared with those experienced while running a marathon; the body's metabolic activity during pregnancy is on par with that displayed in athletes who participate in endurance activities such as ultramarathons and the Tour de France. ⁷ The 2011 American Heart Association statement on reproductive history and later CVD referred to pregnancies as periods of potential vulnerability, ⁸ hinting that some aspect of pregnancy gone wrong might be responsible for the increased risk of CVD associated with a history of pregnancy complications. Although the impact on future health of poor placentation, placental dysfunction, and maternal responses to placental pathology cannot be dismissed, the idea of pregnancy as a cardiac stress test that unmasks an existing predisposition to CVD before traditional risk factors become clinically apparent, better accounts for the wealth of evidence linking pregnancy complications with later CVD risk. ⁹ , ¹⁰ Given the finely tuned cardiometabolic balance that must be maintained for a successful pregnancy, any factor that impedes optimal cardiometabolic functioning is likely to increase the risk of a pregnancy complication or adverse pregnancy outcome, as the mother's systems struggle to adapt to the demands of pregnancy and fail. Furthermore, any factor that prevents a woman's body from fully adapting to the demands of pregnancy (eg, behaviorally determined traits such as maternal obesity or advanced age at first birth, or genetically influenced traits such as diminished vascular/endothelial responsiveness or increased cholesterol levels) will probably have consequences for long‐term health as well.

Against this background, Cécile et al in this issue of the Journal of the American Heart Association (JAHA) examined whether hyperemesis gravidarum might also be associated with the risk of various cardiovascular outcomes after accounting for exposure to preeclampsia, either in the same pregnancy or in separate pregnancies. ¹¹ As was previously the case for preeclampsia, hyperemesis, which is characterized by debilitating nausea and vomiting resulting in dehydration, electrolyte imbalance, weight loss, and even metabolic disturbances, has traditionally been considered a condition whose consequences are limited to pregnancy. However, this belief is eroding, particularly because women with hyperemesis have increased risks of placental function disorders (preeclampsia, placental abruption, stillbirth, fetal growth restriction). ¹² Cécile and colleagues found that hyperemesis was independently associated with later CVD, regardless of exposure to preeclampsia. ¹¹ Interestingly, the strongest associations were observed for nonischemic CVD outcomes, whereas there was little evidence of associations with ischemic outcomes such as myocardial infarction, ischemic heart disease, and ischemic stroke. The study and its results raise several points that are highly relevant at a time when researchers and clinicians from a broad swathe of specialties are trying to translate the links between pregnancy complications and later CVD into recommendations for postpartum monitoring and interventions.

Many studies of pregnancy complications and later CVD risk have only examined associations with ischemic outcomes, likely due to the role thrombosis and ischemia play in many instances of preeclampsia, placental abruption, and stillbirth. However, strict adherence to a narrow mechanistic assumption may have obscured a bigger picture. Cécile and colleagues found that hyperemesis was associated with increased risks of many nonischemic outcomes, including heart failure, cardiomyopathy, conduction disorders, valve disease, angina, inflammatory heart disease, and pulmonary embolism, but not with the ischemic outcomes (myocardial infarction and stroke). ¹¹ Similarly, although preeclampsia is strongly associated with later myocardial infarction and ischemic stroke, it has also recently been associated with increased risks of cardiomyopathy, heart failure, atrial fibrillation, and peripheral artery disease. ³ , ¹³ , ¹⁴ Most pregnancy complications are etiologically heterogeneous (eg, Cécile and colleagues discuss several pathways that may be associated with hyperemesis ¹¹ ), and there are often a variety of mechanisms associated with increased risk of disease occurrence, severity, and outcome. The broad range of CVD outcomes that has been found to be associated with hyperemesis and preeclampsia should remind us that there are likely multiple different pathways leading from adverse pregnancy events to the observed overall increase in the risk of CVD among affected women, and induce us to consider a broad palette of potential outcomes when evaluating the impact of pregnancy complications on postpartum maternal health.

Interestingly, Cécile et al found that hyperemesis and preeclampsia each had an independent impact on the risk of CVD, ¹¹ despite previously documented associations between the 2 conditions. ¹² There was little sign of interaction between hyperemesis and preeclampsia. Instead, the risk associated with having both conditions was roughly what one would expect by multiplying the individual risks together, again suggesting that several underlying mechanisms might each make a separate contribution to later CVD risk.

As Cécile and colleagues note in their Discussion, ¹¹ if pregnancy complications unmask a predisposition to CVD, we may have an opportunity for early intervention to reduce a woman's long‐term risk of these outcomes. A recent study suggested that >50% of the association between preeclampsia and later CVD, and >80% of the association for gestational hypertension, might be mediated by 4 major traditional risk factors for CVD: chronic hypertension, hypercholesterolemia, type 2 diabetes, and obesity. ¹⁵ This suggests that postpartum interventions to optimize blood pressure, cholesterol levels, blood sugar, and weight; to improve diet; and to encourage exercise and smoking cessation, might reduce the burden of CVD in at least some women with a history of pregnancy complications. However, few studies of behavioral and pharmacological interventions to reduce CVD risk in women have focused specifically on the impact of improving CVD risk profiles in this group. ¹⁶ A recent notable study found that aggressive treatment of hypertension immediately after delivery in women who had hypertensive disorders of pregnancy resulted in blood pressure reductions sustained up to 4 years later. ¹⁷ Whether improvements in risk factor profiles can be sustained over decades, and whether these improvements are sufficient to reduce long‐term CVD risk, remain to be determined. Furthermore, given mounting evidence that pregnancy complications are associated with both ischemic and nonischemic forms of CVD, will targeting traditional risk factors for ischemic heart disease be sufficient? We do not yet know whether adding information on pregnancy complications to prediction algorithms improves the prediction of nonischemic cardiac end points beyond what is possible based on conventional measures of CVD risk, nor do we know whether improving traditional risk factor profiles in these women reduces the risk of nonischemic CVD outcomes.

Assuming we can intervene effectively to reduce the risk of CVD in women who have failed the cardiac stress test of pregnancy, current evidence indicates that such interventions must be delivered fairly soon after an affected pregnancy. Figure 2 in the article by Cécile and colleagues suggests that approximately 15% of women who had preeclampsia, hyperemesis, or both conditions will develop some form of CVD within 10 years of delivery. ¹¹ Our own work on preeclampsia paints a similarly bleak picture: depending on their age at the time of an affected pregnancy, 14% to 32% of women who had a hypertensive disorder of pregnancy were diagnosed with treatment‐requiring hypertension within 10 years of delivery, ¹⁸ and up to 4% of women with preeclampsia suffered either a stroke or a myocardial infarction within 20 years of delivery. ¹⁹ To take advantage of the warning provided by pregnancy complications such as hyperemesis, preeclampsia, and placental disorders, we need systems in place to intervene as early as possible, preferably in the immediate postpartum period when the focus on pregnancy‐associated pathology and maternal health and well‐being has not yet waned.

In the era of post–COVID‐19 telehealth solutions, and with the increasing usefulness of applications using artificial intelligence algorithms, options for optimizing traditional CVD risk factor profiles through postpartum screening, monitoring, coaching, and behavioral and pharmacological interventions have mushroomed. In addition to the more traditional postpartum transition clinics being piloted in several settings, such as for women who had hypertensive disorders of pregnancy, ²⁰ other possible interventions could involve the use of remotely monitored mobile applications that flag women for education on risk factor optimization or referral for clinical screening/monitoring based on frequently updated information entered by the women themselves, remote monitoring of blood pressure and blood sugar, or mobile coaching for behavior modification. ¹⁷ However, the idea of scaling up such interventions to cover entire populations, with all their attendant personnel and systems needs, is daunting. In practice, without strong advocacy and a clear demonstration of benefits, how and to what extent postpartum interventions are implemented is likely to depend on the availability of resources at both national and local levels, the perceived benefit to the health care system, and health care system structure (universal versus fee‐for‐service). Furthermore, who should be responsible for the postpartum follow‐up of women who have had pregnancy complications remains unclear. Many medical societies, covering specialties including obstetrics, general practice, and cardiology, are currently discussing how to handle the so‐called fourth trimester and postpartum monitoring of at‐risk groups, as well as patient transitions between specialties. ¹⁷ Perhaps it would not be too astonishing if we were to witness the birth of a new medical subspecialty combining aspects of obstetrics, internal medicine, family medicine, and preventive medicine and tasked with postpartum follow‐up of women identified by their pregnancy experiences as being at risk of later chronic disease. Perhaps, like life‐course epidemiology, this too is an idea whose time has come.

Disclosures

The International Society for the Study of Hypertension in Pregnancy sponsored (conference registration, airline tickets, and accommodation) Dr Boyd's attendance at their annual meeting in August 2022 in Montpellier, France, at which she delivered an invited talk on the long‐term consequences of preeclampsia.