Longitudinal analysis of cardiovascular risk parameters in women with a history of hypertensive pregnancy disorders: the Doetinchem Cohort Study
Women with hypertensive pregnancy disorders (HPD) are at increased risk of developing hypertension and cardiovascular disease later in life; however, it is not known how cardiovascular risk develops throughout life. We evaluated the longitudinal trends in cardiovascular risk factors in women after hypertensive pregnancy disorders compared with women with normotensive pregnancies.
Design and population
All women of the Doetinchem Cohort Study (1987–91), a population-based cohort study, were included.
Women were examined (questionnaires and physical examination) four times at 5-year intervals. History of HPD was assessed from questionnaires. We compared 5-year changes in risk factors between women with and without HPD, by analysing longitudinal trends using generalised estimating equation analysis to estimate the effects of HPD and mean age, adjusting for treatment, body mass index (BMI), smoking and socio-economic status.
Main outcome measures
Change over time in traditional cardiovascular risk factors, including systolic blood pressure (SBP) and diastolic blood pressure (DBP), BMI, total and high-density lipoprotein (HDL) cholesterol for women with and without a history of HPD.
A total of 2703 women with normotensive pregnancies (mean age 40.5 years, SD 10.4) and 689 women with a history of HPD (mean age 38.4 years, SD 9.5) were included. Compared with normotensive women, in women with a history of HPD, SBP was 2.8 mmHg higher (95% CI 1.7–3.9), DBP was 2.3 mmHg higher (95% CI 1.6–3.0) and BMI was 0.7 kg/m2 higher (95% CI 0.4–1.1). Total cholesterol (−0.05; 95% CI −0.1 to 0.0) and HDL cholesterol (0.02; 95% CI −0.0 to 0.1) were similar in both groups. No difference in annual change in blood pressure or in the other risk factors was observed between women with and without a history of HPD.
Women with a history of HPD have higher levels of SBP, DBP and BMI compared with normotensive women, but the increase with ageing is similar in both groups.
In the last decade differences in cardiovascular disease (CVD) between men and women have been extensively studied. Well-known cardiovascular risk factors such as hypertension, hypercholesterolaemia and diabetes mellitus are important for both genders. In addition, several female-specific risk factors have been identified, mostly connected with reproductive history.[1-3] A history of complications in pregnancy, such as gestational diabetes, hypertensive pregnancy disorders (gestational hypertension, pre-eclampsia and HELLP [haemolysis, elevated liver enzyme and low platelet count] syndrome) and placental dysfunction have been associated with increased CVD risk. The severity of pregnancy disorders is related to future cardiovascular risk.[4-8] Women with a history of pre-eclampsia, defined as hypertension and proteinuria in the second half of pregnancy, have at least a two times higher risk of cardiovascular disease at older age.[9-11] Gestational hypertension is more benign in terms of pregnancy outcome and long-term maternal health; however, risk of chronic hypertension and coronary artery calcification is also increased in these women.[12-14] Several hypotheses about the mechanism between hypertensive pregnancy disorders (HPD) and CVD have been postulated, and endothelial dysfunction seems to play a crucial role in the development of future CVD.[15-19]
Until now there has been a gap in information on the development of cardiovascular risk between the ages of 40 and 60 years in women after HPD. A recent study showed an increased 10-year cardiovascular risk in women with a history of HPD 18 years postpartum compared with women with normotensive pregnancies or previous gestational diabetes. However, longitudinal follow-up data are not available. Current guidelines identify women with a history of gestational diabetes, gestational hypertension and pre-eclampsia at risk for future cardiovascular disease, although there are no specific recommendations on timing of monitoring or prevention in these high-risk women; neither is there evidence on the cost-effectiveness of preventive measures in this target group.[21, 22]
The objective of this study is to gain more insight into the presence and development of cardiovascular risk factors over time in women with a history of HPD. In this study we investigate in a population-based cohort, the Doetinchem Cohort Study, the longitudinal trends in the cardiovascular risk factors blood pressure, body mass index (BMI), total cholesterol and high-density lipoprotein (HDL) cholesterol in a subgroup of women with and without a history of HPD.
The Doetinchem Cohort Study is a prospective population-based study, set up to study the impact of (changes in) lifestyle factors and biological risk factors on health. Between 1987 and 2007, four measurement rounds were completed. At baseline (R1: 1987–91), 12 405 inhabitants aged 20–59 years of Doetinchem, a town in a rural area of the Netherlands, were examined as part of the Monitoring Project on Cardiovascular Disease Risk Factors (MP-CDRF). From the first round, a random sample of 7769 participants was re-invited for a second examination (R2: 1993–97), and again 5 and 10 years later for a third (R3: 1998–2002) and fourth (R4: 2003–07) examination. A detailed description of the sampling and data collection procedures has been published previously. The study protocol was approved by the external Medical Ethics Committee of the Netherlands Organization of Applied Scientific Research Institute and all participants provided written informed consent. To evaluate the effects of HPD we only selected the women in the Doetinchem Cohort study (n = 3392); throughout the study there were 2703 women without a history of HPD and 689 women with a history of HPD (i.e. a positive answer on the question ‘did you ever have high blood pressure in pregnancy’).
Biological risk factors
The biological risk factors of interest for the present study that were collected in each round included (1) systolic blood pressure (SBP) and diastolic blood pressure (DBP), (2) total and HDL cholesterol and (3) height and weight. Body weight and height were measured in light indoor clothing with emptied pockets and without shoes. Height was measured with a wall-mounted stadiometer to the nearest 0.5 cm. Body weight was measured with a balance-beam scale to the nearest 0.5 kg. BMI was calculated as weight (kg) divided by height (m) squared. Blood pressure was measured twice, with the participant in a sitting position. The mean value of the two measurements was used in the analysis. In R1–3 blood pressure was measured with a random zero sphygmomanometer (Hawksley & Sons, Lancing, UK) and the first and the fifth phases Korotkoff sounds were the criteria for SBP and DBP, respectively. In examination R4, blood pressure was measured using an automated Speidel–Keller meter (Welch Allyn, Skaneateles Falls, NY, USA). In all rounds, a 30-ml nonfasting serum blood sample was drawn. Total and HDL cholesterol were determined at the Lipid Reference Laboratory of the Erasmus Medical Centre in Rotterdam, using an automated enzymatic procedure.
Cardiovascular risk factors
Hypertension was defined as an SBP ≥140 mmHg and/or a DBP of ≥90 mmHg and/or use of antihypertensive medication. Presence of diabetes mellitus was defined as self-reported treatment for diabetes mellitus; hypercholesterolaemia was defined as nonfasting total cholesterol ≥6.5 mmol/l and/or use of statins. Smoking was defined as current smoking, ex-smoking or no smoking.
In each round, participants completed a questionnaire on demographic characteristics, lifestyle characteristics and presence of chronic diseases or risk factors thereof. Use of antihypertensive medication and cholesterol-lowering therapy was classified into yes/no. Educational level was assessed as the highest level of completed education during follow-up and classified into three categories: low (intermediate secondary education or less), medium (intermediate vocational or higher secondary education) and high (higher vocational education or university). A history of HPD was defined by self-reported history of HPD. This was based on a positive answer of the question ‘Did you ever have high blood pressure in pregnancy’ at one of the four rounds. A history of gestational diabetes was also based on self-reported history.
Baseline characteristics were described as mean with standard deviation for continuous data, and as percentages for categorical data. Differences between women with and without HPD for continuous data were analysed by Student's t test for independent groups and for categorical data with chi-square test or Fisher exact test when appropriate.
To evaluate the effect of HPD and age on blood pressure, cholesterol measurements and BMI over the follow-up time we used generalised estimating equation analyses (GEE) with the genmod procedure in SAS software (SAS Institute, Cary, NC, USA). The GEE procedure extends the generalised linear model to allow for analysis of repeated measurements at different time-points. The intercept of this model is an estimate of the difference in risk factors between women with and without HPD. For this analysis we modelled HPD time-dependently, which means that women were defined as having had HPD from the moment they first reported HPD. The regression coefficient is an estimate of the annual change in risk factor. Whether presence of HPD yes or no modifies the age-related increase in risk factors was evaluated by including the interaction term ‘HPD × age’ into the model. Adjustments were made for use of antihypertensive medication, statins, BMI, smoking and socio-economic status. All P-values were two-sided, statistical significance was assumed when two-tailed probability value was <0.05.
Table 1 shows characteristics of the Doetinchem women at all four follow-up rounds. Table 2 shows the regression analysis of the different continuous cardiovascular risk factors (SBP and DBP), BMI and cholesterol (total and HDL) for presence of HPD, ageing and interaction between HPD and age.
Table 1. Baseline characteristics of the women in the Doetinchem cohort
|Age (years)||40.5 (10.4)||38.4 (9.5)a||44.4 (9.5)||46.7 (10.4)a||49.3 (9.4)||50.8 (10.2)a||55.4 (10.1)||54.1 (9.2)|
|Parity (n)||2.3 (1.0)||2.3 (1.0)||2.3 (1.0)||2.4 (0.9)||2.3 (1.1)||2.4 (0.9)||–||–|
|SBP (mmHg)||117 (15.0)||118 (13.5)||122 (17.5)||123 (15.9)||125 (18.4)||128 (17.9)a||132 (20)||134 (18.8)a|
|DBP (mmHg)||75 (10.5)||76 (9.3)||78 (10.7)||79 (10.6)a||79 (10.6)||81 (10.9)a||83 (10.3)||85 (10.2)a|
|Antihypertensive medication (%)||4.9||1.6a||8.3||0.2a||12.6||5.4a||18.2||14.6|
|Total cholesterol (mmol/l)||5.4 (1.1)||5.3 (1.0)a||5.5 (1.1)||5.4 (1.0)a||5.7 (1.1)||5.7 (1.0)||5.6 (1.0)||5.7 (1.0)|
|HDL cholesterol (mmol/l)||1.4 (0.3)||1.4 (0.3)||1.5 (0.4)||1.5 (0.4)||1.5 (0.4)||1.5 (0.4)||1.6 (0.4)||1.6 (0.4)|
|Ratio Total/HDL||4.2 (1.3)||4.1 (1.4)a||3.8 (1.2)||3.7 (1.1)a||4.0 (1.3)||4.0 (1.3)||3.8 (1.2)||3.8 (1.2)|
|Statin use (%)||0.2||0||1.2||0.6||3.4||2.1||9.1||6.1|
|BMI (kg/m2)||24.2 (3.8)||24.4 (3.8)||25.2 (4.2)||25.6 (4.3)||25.8 (4.3)||26.3 (4.3)a||26.1 (4.6)||26.9 (4.7)a|
|Diabetes mellitus (%)||0.6||0.9||1.6||1.6||2.2||0.9||4.0||3.5|
|Gestational diabetes (%)||–||–||2.5||4.2a||3.8||7.0a||3.4||6.3a|
Table 2. Difference and annual change of cardiovascular risk factors (RF) for women with history of HPD and women without HPD (GEE analysis)
|SBP (mmHg)||3.10 (1.95−4.25)||0.84 (0.80−0.88)||3.39 (2.26−4.52)||0.80 (0.76−0.85)||2.78 (1.69−3.87)||0.72 (0.68−0.77)|
|DBP (mmHg)||2.58 (1.87−3.29)||0.29 (0.26−0.31)||2.76 (2.06−3.45)||0.26 (0.24−0.29)||2.30 (1.62−2.98)||0.22 (0.19−0.24)|
|BMI (kg/m2)||0.64 (0.29−0.99)||0.10 (0.09−0.11)||0.69 (0.34−1.03)||0.09 (0.07−0.10)||0.72 (0.38−1.06)||0.07 (0.06−0.09)|
|Total cholesterol (mmol/l)||−0.03 (−0.10 to 0.05)||0.034 (0.031−0.037)||−0.04 (−0.12 to 0.04)||0.040 (0.037−0.043)||−0.05 (−0.13 to 0.03)||0.039 (0.036−0.042)|
|HDL cholesterol (mmol/l)||0.01 (−0.02 to 0.04)||0.000 (−0.001 to 0.002)||0.01 (−0.03 to 0.04)||0.001 (0.000−0.002)||0.02 (−0.01 to 0.05)||0.003 (0.002−0.004)|
Blood pressure levels were significantly higher in the women with a history of HPD compared with those without HPD. SBP was 3.10 mmHg (95% CI 1.95–4.25) higher and DBP was 2.58 mmHg (95% CI 1.87–3.29) higher in women with HPD compared with women without a history of HPD. Adjustment for use of antihypertensive medication, statins, BMI, smoking and socio-economic status in the model slightly decreased the blood pressure differences between both groups, but differences remained statistically significant, SBP 2.78 mmHg (95% CI 1.69–3.87) higher and DBP 2.30 mmHg (95% CI 1.62–2.98) higher. No interaction effect for the presence of HPD and age-related increase in blood pressure was observed: the increase in blood pressure was 0.84 mmHg (95% CI 0.80–0.88) per year for SBP and 0.29 mmHg (95% CI 0.26–0.31) per year for DBP in both groups.
Body mass index was also significantly different between both groups, 0.64 kg/m2 higher (95% CI 0.29–0.99) in women with a history of HPD; the yearly BMI increase was not different according to HPD (0.10 m2/kg [95% CI 0.09–0.11] per year). Adjustment for potential confounders did not change the BMI differences between the groups nor the annual BMI increases in women.
At rounds 1 and 2 total cholesterol was higher in women without a history of HPD, later this difference disappeared. Levels of HDL cholesterol were similar at each round. After full adjustment, total and HDL cholesterol levels did not differ between women with and without history of HPD (difference between groups – 0.05 mmol/l; 95% CI –0.13 to 0.03 for total cholesterol, and 0.02 mmol/L; 95% CI – 0.01 to 0.05, for HDL cholesterol). A history of HPD also did not influence yearly changes of total and HDL cholesterol.
Prevalences of other traditional cardiovascular risk factors are described in Table 1. Smoking prevalence was lower in women with HPD at rounds 1 to 3. Prevalence of hypertension was not different between both groups, although use of antihypertensive medication was more prevalent in women without a history of HPD at rounds 1 to 3. Prevalence of hypercholesterolaemia, diabetes mellitus and family history of myocardial infarction were similar in both groups of women. Women with a history of HPD more often had previous gestational diabetes compared with women without hypertensive pregnancy complications (6% for women with HPD and 3% in women without HPD at round 4).
In this longitudinal population-based cohort study we observed that in women with a history of HPD mean SBP was 3 mmHg higher, mean DBP was 2 mmHg higher and mean BMI was 0.7 kg/m2 higher, compared with women without a history of HPD. Increases in blood pressure and BMI with age did not differ between the two groups. Mean levels of total and HDL cholesterol did not differ between women with and without a history of HPD.
Strengths and weaknesses
Our study has several limitations that have to be taken into account. An important limitation is the assessment of the self-reported history of hypertensive pregnancy disorders, based on the questionnaire. This precluded us from differentiating between mild and severe conditions. Furthermore, mild conditions may have been under-reported, resulting in misclassification. This misclassification of mild cases of HPD as normotensive women will however only have led to an underestimation of the true magnitude of the associations, and could be a good explanation for the modest changes in blood pressure and BMI that we found in our study compared with other studies in the literature.[8, 19] Besides this, we lack information on the timing of the affected pregnancy and the severity of the reported HPD, which may vary from limited gestational hypertension to full-blown severe pre-eclampsia or HELLP syndrome. Previous studies have shown that this has important consequences for future cardiovascular risk. In the women with a history of HPD in our analysis the complete spectrum of HPDs will be present. Data in women with severe pre-eclampsia only will result in larger effects, as we previously demonstrated.
Important strengths of our study are that we had data from almost 3400 young women from the general population with an extensive follow-up and high response rates during follow-up, between 75 and 80%. Additionally, risk factors were measured four times over a period of 20 years, with direct contact with the participants at each round. We were able, for the first time, to study the role of HPD in CVD risk factor patterns over time. As intermediate cardiovascular follow-up data in affected women are scarce, our study adds valuable information to the field of cardiovascular risk in women with history of hypertensive diseases of pregnancy.
Several studies identified the increased risk of CVD for women with a history of HPDs, especially pre-eclampsia.[4-8] Prevalence of chronic hypertension, an important risk factor for CVD, is also high in this subset of women. As we have underlined before, little is known about the course of development of cardiovascular risk factors after having experienced gestational hypertension. A recent study identified an increased 10-year cardiovascular risk (based on Framingham risk score) in women with a history of pregnancy disorders (18 years postpartum). To our knowledge our study is the first analysis of longitudinal data on cardiovascular risk factors in women with previous HPD. We observed that blood pressures and BMI levels in these women were significantly higher at all time periods during follow-up, which is consistent with the literature.[17-19] However, the impact on clinical significance of the observed differences is limited and further research should confirm our data. The established differences between both groups do not further increase during follow-up. This information might shed some light on the different hypotheses about the relationship between hypertensive pregnancy disorders and future cardiovascular disease. From round 1 onwards, the exposed women are different, giving support to the hypothesis that the endovascular system of affected women has changed before pregnancy. In this study we were not able to evaluate the effect of prepregnancy factors, because we do not know the exact timing of the HPD in these women.
For other traditional cardiovascular risk factors, like lipids, we did not find a difference between women with a history of HPD compared with women with normotensive pregnancies. This is unexpected, as other studies have found a more unfavourable lipid profile in women with a history of HPD compared with reference women.[17, 25] The possibly more limited severity of the previous HPD in our cohort could be the reason for these findings. Women with a history of HPD were less often smokers, whereas previous smoking was similar within both groups. These factors therefore do not explain the known increased cardiovascular risk for women with a history of HPD. Blood pressure and BMI seem to be the most important factors for future cardiovascular risk in this subset of women. Unfortunately, our study was not powered to evaluate differences in cardiovascular events over time for women with and without history of HPD.
The latest American Heart Association guidelines for CVD prevention in women mention pregnancy disorders as risk factors for future CVD and advise taking a detailed obstetric history in women and monitoring their risk factors carefully. This is not yet common daily practice however, so more efforts are needed to create awareness among clinicians and their patients. According to our present data, measurements of blood pressure and lifestyle advice are to be recommended for women with a history of HPD between 40 and 55 years of age. The clinical impact in the long-term is unclear.
To further elucidate the relation between hypertensive pregnancy disorders and CVD risk we suggest that future research should focus on lifetime development of risk factors with additional prepregnancy measurements, careful registration of pregnancy disorders and long-term follow-up postpartum. In parallel, the effects of possible preventive measures should be evaluated.
In this longitudinal population-based cohort study we observed that women with a history of HPD have significantly higher blood pressure and BMI compared with women without a history of HPD. During follow-up both blood pressure and BMI increase similarly with age in both groups. Based on these data we could not advise long-term monitoring of cardiovascular risk factors postpartum for women with a history of HPD. Future research should focus on careful registration of pregnancy disorders, preferably with information on prepregnancy risk factor levels, and evaluation of lifetime development of cardiovascular risk factors.
Disclosure of interest
Contribution to authorship
All authors contributed to the study; the analysis and interpretation of the data was by JD and WV; drafting and revising of the article and approved the final version of the manuscript were by all authors.
Details of ethics approval
The study protocol was approved by the external Medical Ethics Committee of the Netherlands Organization of Applied Scientific Research Institute, and all participants provided written informed consent.
The Doetinchem Cohort Study was financially supported by the Ministry of Health, Welfare and Sport of the Netherlands and the National Institute for Public Health and the Environment.
The authors thank the epidemiologists and fieldworkers of the Municipal Health Service in Doetinchem for their contribution to the data collection for this study. Project director is WMM Verschuren; logistic management was provided by P Vissink and data management by A Blokstra. We thank Prof. H Boshuizen for statistical advice. The authors thank VRC Derks for help in proofreading and submitting the article.