Live birth/parity number and the risk of incident hypertension among parous women during over 13 years of follow‐up

Abstract The effect of live birth/parity number on incident hypertension was investigated among Iranian parous women aged 30–70 years. The study population included 2188 normotensive women who were enrolled in 1999–2001. They were followed for incident hypertension (based on JNC 7 report) by 3‐year intervals up to April 2018. Multivariable Cox proportional hazard models, adjusted for a wide set of potential hypertension risk factors, reproductive factors, and pregnancy complications, were applied to estimate hazard ratios (HRs) and 95% confidence intervals (CIs) of the number of parity/live birth(s) for incident hypertension. Additionally, as a sensitivity analysis, age‐scale Cox regression was also done. During a median follow‐up of 13.5 years, 935 incident hypertension have occurred. Compared to those with two live births, the participants who had 3 and ≥4 live births were at higher risk of hypertension development by the HRs of 1.25 [95% CI: 1.02–1.55] and 1.39 [1.12–1.72], respectively, in the full‐adjusted model. Moreover, each additional live birth increased the risk of hypertension by a HR of 1.06 [95%CI: 1.02–1.11]. Results of parity number were also similar. Considering age as time scale also did not change the results generally. The authors found a significant interaction between live birth/parity number and age groups; the adverse effect of higher live birth/parity numbers on hypertension development was mainly found among those aged < 50 years. To sum up, compared to the live birth/parity number of two, Iranian women with ≥3 live birth/parity had a higher risk of incident hypertension; the issue was more prominent among younger mothers.

In addition to known hypertension risk factors such as aging, dietary factors (high sodium intake, low potassium intake, and unhealthy diet), obesity, lack of physical activity, and other potential hypertension risk factors, 1 childbearing and some other reproductive factors were suggested to have an association with blood pressure (BP) level and hypertension among women. [4][5][6] Physiological changes during pregnancy, labor, and also postpartum periods influence the BP levels. 7 Moreover, physiologic cardiometabolic changes (weight gain, insulin resistance, increased plasma glucose, and dyslipidemia) and complications that women might experience during pregnancy exert potential short-and long-term effects on postpartum outcomes, especially cardiovascular ones. [8][9][10][11][12][13] Most of the previous findings of the effect of live birth and parity on BP level or hypertension were limited to cross-sectional studies with inconsistent results. Although some previous studies suggested that parity can have lowering effects on BP level, 14,15 many more have reported that as the live birth or parity number increased, the prevalence of hypertension became higher. [16][17][18] Moreover, findings from prospective studies on this issue, all conducted in Western countries, were also limited and inconclusive. 6,[19][20][21] The current study aims at determining whether the live birth/parity number is an independent risk factor for hypertension development among Iranian women aged 30-70 years, with 13.5 years of follow-up, using a population-based cohort, the Tehran Lipid and Glucose Study (TLGS).

Study design and study population
The TLGS is a prospective cohort study. This study was conducted on a representative sample of Tehranian citizens that resided in district 13 of Tehran. Determining the prevalence, incidence, and other epidemiologic aspects of non-communicable diseases (NCDs) and also prevention of NCDs by advancing healthier lifestyles were the original aims of the TLGS. The details of the design, measurement methods, and enrollment strategy of the TLGS have been described elsewhere. 22 To summarize, the first phase of the TLGS, which was on January 31,

Clinical and laboratory measurements
According to the TLGS protocol, 22

Definition of main exposure, confounding factors, and outcome
Main exposure: The sum of the number of live birth and the stillbirth (defined as the birth of an infant which died in the mother's uterus after 20 weeks of gestation) was considered as parity number.
Confounder factors: Diabetes mellitus (DM) was defined as a FPG level of ≥7.0 mmol/L or pharmacologically treated with glucoselowering drugs. Smoking status was categorized into two groups; current smokers versus former/never smokers. Prior diagnosed CVD in female first-degree blood relatives aged < 65 years or male firstdegree blood relatives aged < 55 years was defined as a positive family history of premature CVD. Based on the World Health Organization's definition, the absence of spontaneous menstrual bleeding for more than 12 months, without other pathologic or physiologic cause could be determined as menopause 23 ; for participants with missing information, the menopausal age was considered at ≥50 years. 24 Based on the standard definition of the onset of a BP level ≥140/90 mm Hg with proteinuria > 0.3 g/24 h after 20 weeks' gestation, 25 preeclampsia diagnosis was made as a part of routine maternal care in Iran. At the enrollment phase, using a validated self-reporting questionnaire, 26 female participants were asked about their history of preeclampsia. 27 Finally, gestational DM (GDM) was defined as the presence of macrosomia or a history of GDM that self-reported by participants. 28 Macrosomia was considered as a birth weight > 4 kg. 29 Outcome: Based on the seventh report of the joint national committee on prevention, detection, evaluation, and treatment of high BP (the JNC 7 report), hypertension, whether at baseline or as an outcome during follow-up, was defined as the presence of at least one of the following criteria: (a) having SBP≥140 mm Hg, (b) having DBP≥90 mm Hg, and (c) initiation of anti-hypertensive drugs usage. Moreover, SBP 120-139 mm Hg or DBP 80-89 mm Hg were defined as prehypertension. 30

Statistical analyses
Descriptive statistics were used to describe baseline characteristics as Cox proportional hazard models were used to evaluate the association of parity and live birth number with incident hypertension. Parity and live birth number were considered as both continuous and categorical variables in the Cox models, separately. We considered live birth/parity number of two as the reference group, given that this number was associated with the lowest risk for CVD event among Iranian women. 13 In the current study, the selection of confounders was derived from our recent study among the Iranian female population. 3 It should be noted that we used the baseline data of the covariates. Using the Schoenfeld residual test, the proportionality was assessed for the Cox models. All proportionality assumptions were appropriate in this study. STATA version 14 (StataCorp LP, College Station, TX, USA) statistical software was used for statistical analyses of the current study and a two-tailed p-value ≤.05 was considered significant.

RESULTS
The study population consists of 2188 female participants with a mean age (SD) of 43 ( Table 3 and Table S2).
In a sensitivity analysis, considering age scale Cox regression, each additional live birth or parity was associated with a 4% higher risk of incident hypertension in model 4 (for parity, the risk was marginally significant). Moreover, after adjustment for a wide series of confounders in model 2, compared to the reference group (live birth/parity number = 2), having 3 and ≥4 live birth/parity(s) increased the risk Moreover, in some other cross-sectional studies, live birth/parity number did not significantly associate with BP levels and hypertension, or lose their significance after adjustment for confounders. [35][36][37][38] Based on a study on perimenopausal and postmenopausal women from Italy, Giubertoni and coworkers found that compared with nulliparous women, those with at least one child had higher prevalence of hypertension; however, the incidence of hypertension during follow-up was not related with parity. 21 Prospective data from US women who were parous at baseline showed that adjusted mean BP did not differ by the number of subsequent births; however, for those who were nulliparous at baseline, adjusted mean BP decreased by subsequent births. 6 Results from the HUNT study in Norway also demonstrated that successive pregnancies were associated with lasting and clinically relevant reductions in SBP and DBP; it took about a decade for parous women to reach the levels they had experienced before pregnancy. 20 In contrast to the studies mentioned above, among Danish premenopausal women, researchers found that as the number of live-birth pregnancies increased, the risk of incident hypertension increased between live-birth pregnancies during follow-up. 19 Similar to results from Denmark, we also found an unfavorable impact of higher numbers of live birth/parity among Iranian women; the issue was more prominent among younger ones. The non-significant association among older participants may be due to greater contribution of aging and other traditional risk factors on hypertension development in the elderly, especially after menopause. Another possible reason is the limited number of older participants in our study. In contrast to our results, a population-based study in Switzerland also demonstrated that parity had a decreasing and increasing effect on BP level in younger and older groups, respectively (p-value for interaction < .001); however, due to cross-sectional design of this study, their finding was not comparable with our prospective results. 31 The observed higher hypertension risk and prevalence by increasing in live birth/parity number was suggested to be mainly attributed to the pathway of metabolic changes (weight gain, dyslipidemia, insulin resistance, and increased plasma glucose) that occur during pregnancy physiologically. 39 By multiple pregnancies, the exposure time to these changes was increased. This accumulative effect of repeating parity can be associated with incident metabolic diseases such as metabolic syndrome, obesity, T2DM, and CVD in the future life of mothers, [9][10][11][12]40,41 and consequently, hypertension development, since these disorders are known hypertension risk factors. We suggest that although this pathway can have a role, but it could not justify the unfavorable impact of higher live birth/parity number completely, because in our data analysis, we adjusted for general and central obesity, DM, lipid profile, and also prehypertension as confounders and results remained significant. Moreover, even after considering the history of preeclampsia and GDM, which were extreme of these metabolic changes pathologically, results did not change. As another explanation, some human and animal studies suggested that multiple pregnancy can cause endothelial dysfunction and greater pressure response to vasoconstrictive agents. [42][43][44][45][46][47] In addition to the biological pathway, multiparity was found to develop CVD through an unhealthier lifestyle and socioeconomic factors. 8,48,49 It was reported in Iran that household income per capita is lower in higher family size. 50 Furthermore, at the recruitment time of this study, Iranian government had a policy of reducing population growth, so there was minimal economic support for Iranian parents. 51 Therefore, mothers with more children were at higher economical pressure and have lower leisure time. It can lead to unhealthier diet (less access to high fiber and fruit rich diet), physical inactivity, poor access to health care, and other socioeconomic problems. Therefore, health policymakers need to pay special attention to education on health, nutrition, and special economic support of mothers willing to have more than two children.
The key strengths of this study are its long duration of follow-up, standardized measurements for confounders and outcome rather than relying on self-reported data, and using a wide set of confounders, especially complications of pregnancy that poorly addressed in previous studies. However, several important limitations need to be considered. First, no access exists for valid data on job status, income, and diet habits of participants. These factors can be the socioeconomic and lifestyle factors that have a potential effect on hypertension development. Second, the number of female participants with no children at baseline was too low (n = 8) to compare the impact of nulliparity with ever parity. Third, we relied on office BP measurements using non-automated device rather than ambulatory or home BP measurements; hence, we cannot diagnose those with "white coat" hypertension or "masked" hypertension. However, the method applied in other population-based studies. 6,20 Finally, all of our participants belonged to an urban area only, and our findings may be unable to be generalizable to rural populations.
To sum up, during more than a decade of follow-up, among residents of the metropolitan city of Tehran, compared to live birth/parity number of two, those with ≥3 live birth/parity had higher risk of incident hypertension, independent of well-known hypertension risk factors and reproductive factors; the issue was more prominent among younger mothers. Further investigations are needed to evaluate this issue in other parts of the country and also discover the potential role of socioeconomic and lifestyle factors in the pathway between live birth/parity and hypertension development.

ACKNOWLEDGMENT
The authors would like to express their appreciation to the TLGS participants and staff for their kind cooperation.

CONFLICTS OF INTEREST
There are no conflicts of interest.

Study conception and design: Seyyed Saeed Moazzeni and Farzad
Hadaegh; Analysis and interpretation of data: Samaneh Asgari and