Long-term prognosis of hypertension in pregnancy

Objectives: To assess the prevalence of subsequent hypertension in women wi th hypertensive pregnancies and evaluate it according to the subclassificat ions of hypertension in pregnancy. Methods: A survey was carried out in 476 women with hypertensive pregnancie s (cases) and 226 normotensive controls delivered between 1973 and 1991 in a tertiary-level teaching hospital. They were invited to participate by mai l and 273 cases (57%) and 86 controls (38%) completed the analysis. Outcome s assessed were prevalences of hypertension, diabetes, and hypercholesterol emia, together with cardiovascular morbidity. Results: Among responders, age and parity were similar in both groups altho ugh follow-up time was longer in controls. Subsequent hypertension was more frequent within cases. After excluding chronic and unclassifiable hyperten sion, the mean blood pressure was higher in all other forms of pregnancy hy pertension (103 +/- 13 mm HE versus 94 +/- 13 mm Hg, p < 0.001); long-term hypertension prevalence was 45% in cases and 14% in controls [odds ratio (O R) = 5.1; 95% confidence interval (95% CI) = 2.5-9.8; p < 0.001]. There wer e no differences with respect to the prevalences of subsequent diabetes or hypercholesterolemia. Remote hypertension was more common following gestati onal hypertension (54%) than in preeclampsia (38%), eclampsia (14%), or nor motensive cases (14%) (OR for gestational hypertension versus normotensives = 7.2; 95% CI = 3.4-14.8, p < 0.001, and OR for preeclampsia versus normot ensives = 3.7; 95% CI = 1.7-7.9, p < 0.001). Conclusions After an average of 13.6 years since the index pregnancy, women with hypertensive pregnancies have an increased risk of subsequent hyperte nsion. Gestational hypertension is the hypertensive disorder of pregnancy w ith the highest incidence of subsequent hypertension. Women with preeclamps ia have a greater tendency to develop hypertension than women with normoten sive pregnancies. By contrast, women with eclampsia do not.