Preeclampsia and subsequent cardiovascular disease: villain or innocent bystander?

Not long ago, women who experienced preeclampsia were reassured that there were no long-term health risks associated with their condition—with the notable exception of an increased risk of recurrent preeclampsia in future pregnancies. Although proteinuria and hypertension would sometimes take weeks or even months to resolve, spontaneous remission after delivery was the rule. Although preexisting hypertension, kidney disease, and diabetes mellitus were recognized risk factors for preeclampsia, the large majority of women with preeclampsia were healthy primiparous women without significant medical problems. Medical students were taught, and obstetricians counseled, that preeclampsia resolves after delivery and that there were no long-term maternal risks to worry about. A little over a decade ago, the tide began to shift, and it gradually became clear that this once comforting mantra is wrong. A series of retrospective cohort studies, many based on data from national or regional birth registries including Norway (1), Scotland (2,3), Washington State, United States (4), and Ontario, Canada (5), demonstrated with remarkable consistency an increased risk of subsequent cardiovascular disease 10–20 years after preeclampsia. Two largemeta-analyses have shown that women with prior preeclampsia have an increased risk for hypertension (relative risk [RR], 3.7), ischemic heart disease (RR, 2.16 to 2.33), cerebrovascular disease (RR, 1.81 to 2.03), cardiovascular mortality (RR, 2.29), and overall mortality (RR, 1.49) (6,7) compared with women without prior preeclampsia. Women with severe preeclampsia (4), or preeclampsia with preterm delivery (1) or low birthweight (2)—both markers of preeclampsia severity—are at highest risk. Preeclampsia also carries an almost 5-fold increase in ESRD later in life (8). However, all of these studies and the meta-analyses that summarize them share one important limitation: failure to completely measure and control for the presence of cardiovascular risk factors before pregnancy. Preeclampsia and cardiovascular disease share many risk factors, including chronic hypertension, diabetes, renal disease, and obesity (9,10). Studies vary considerably in their approach to these potential covariates. Several did not attempt to identify, exclude, or control for prepregnancy hypertension. Others did exclude (4,11) or control for (5) some of these comorbidities, but relied on billing diagnosis codes, which may be unreliable. With all retrospective studies, identification of risk factors—especially if subtle or subclinical (insulin resistance, prehypertension)—will inevitably be incomplete. This begs the question: Does preeclampsia lead to subsequent cardiovascular disease, perhaps due to persistent endothelial dysfunction triggered by preeclampsia? Or do preeclampsia and cardiovascular disease reflect different manifestations of a common high-risk vascular phenotype? In other words, is preeclampsia truly an independent risk factor for cardiovascular disease? Microalbuminuria is an early marker for cardiovascular disease (12). If preeclampsia is causally implicated in subsequent cardiovascular disease, one would expect an association between preeclampsia and subsequent albuminuria. Failure to detect such an association would support the alternative hypothesis that the association between preeclampsia and cardiovascular disease simply reflects shared risk factors. In this issue of CJASN, Sandvik and colleagues (13) report the largest single study to date of albuminuria after preeclampsia. Using national registry data, they identified all women living in the Bergen area of Norway who experienced preeclampsia in their first pregnancy. The diagnosis of preeclampsia was determined clinically by the attending midwife or doctor using standard criteria. Women with diabetes, renal disease, hypertension, and rheumatic disease were excluded, as were women with recurrent preeclampsia in subsequent pregnancies. Controls without preeclampsia were matched by age and year of first pregnancy. Women were contacted approximately 10 years after their first pregnancy; 49% of preeclampsia cases and 38% of controls agreed to participate, for a total of 89 preeclampsia cases and 69 controls. The primary outcome variable was urinary albumin excretion above the 75th percentile (urinary albumin/creatinine ratio (ACR) .0.70 mg/mmol or .6.2 mg/mg) in three morning urine samples. Analyses were adjusted for age, body mass index, marital status, annual household income, and highest educational level. There was no significant difference in median ACR between the women with prior preeclampsia (0.53 mg/mmol) and controls (0.50 mg/mmol). In adjusted analyses, the odds ratio (OR) for ACR above the 75th percentilewas 1.08 (P50.85). Only onewoman in the preeclampsia group (and none in the control Lehigh Valley Health Network, Allentown, Pennsylvania; and University of South Florida Morsani College of Medicine, Tampa, Florida