Is the Association Between Parity and Coronary Heart Disease Due to Biological Effects of Pregnancy or Adverse Lifestyle Risk Factors Associated With Child-Rearing?
Findings From the British Women’s Heart and Health Study and the British Regional Heart Study
Background— Parity is associated with coronary heart disease (CHD) risk. In the present study, we assessed the associations between number of children and CHD in both women and men.
Methods and Results— A total of 4286 women and 4252 men aged 60 to 79 years from 24 British towns were studied. Number of children was positively associated with body mass index and waist-hip ratio in both sexes. In women but not in men, number of children was inversely associated with high-density lipoprotein cholesterol and was positively associated with triglycerides and diabetes. For both sexes, similar “J” shaped associations between number of children and CHD were observed, with the prevalence lowest among those with 2 children and increasing linearly with each additional child beyond 2. For those with at least 2 children, each additional child increased the age-adjusted odds of CHD by 30% (odds ratio, 1.30; 95% confidence interval, 1.17 to 1.44) for women and by 12% for men (odds ratio, 1.12; 95% confidence interval, 1.02 to 1.22). Adjustment for obesity and metabolic risk factors attenuated the associations between greater number of children and CHD in both sexes, although in women some association remained.
Conclusions— Lifestyle risk factors associated with child-rearing lead to obesity and result in increased CHD in both sexes; biological responses of pregnancy may have additional adverse effects in women.
Received September 9, 2002; revision received November 21, 2002; accepted November 25, 2002.
The association between parity and coronary heart disease (CHD) risk in women has been assessed in a number of prospective studies.1–4 The majority,2–4 although not all,5 found a positive association. Because pregnancy is a state of relative insulin resistance, most investigators have stressed biological pathways for this association.1–4 Studies that have shown an association between pregnancy and abnormal metabolic risk factors have largely been conducted on pregnant women or women of reproductive age,5–7 and it is unclear whether any effect of repeated pregnancies on these risk factors persists into later life. Further, the possibility that the association is the result of residual confounding due to socioeconomic position8 or lifestyle risk factors associated with child-rearing9 have not been adequately examined.
A study of the association between number of children and incidence of various cancers demonstrated that comparisons between women and men were a useful method for distinguishing whether the mechanism for the association involved biological processes related to pregnancy or lifestyle factors that were influenced by or influenced family size.10 Only one small study previously looked at this possibility with respect to CHD, and it found that having ≥4 children, compared with none, was associated with increased CHD risk in both sexes.4
In the present study, we assessed the association between number of children and CHD risk factors and prevalent CHD in a large study of older women and men.
Data from the British Regional Heart Study and the British Women’s Heart and Health Study were used. The British Regional Heart study is a prospective study of cardiovascular disease involving 7735 British men followed since 1978 to 1980.11 Between February 1998 and February 2000, a clinical re-examination of all surviving members of the British Regional Heart Study, now aged 60 to 79 years, was undertaken (n=4252, with a 76% attendance of cohort survivors). The British Women’s Heart and Health Study12 was established in 1999 and is a cohort study of women designed to parallel the British Regional Heart Study. Sampling was stratified by town and by 5-year age group to ensure that the distribution was proportionately matched with that of the British Regional Heart Study. A total of 4286 women (60% of the 7173 invited) participated, and baseline data were collected between May 1999 and March 2001.12 Ethics committee approvals were obtained for both studies, and participants gave informed consent for review of their medical records. In the present study, cross-sectional analyses were made using data from the 1998 to 2000 re-examination of the men and the 1999 to 2001 baseline data from the women. Protocols for obtaining data were similar in both studies.
Outcome and Explanatory Variables
Full details of all measurements have been previously reported.12–14 For men, number of children was determined from the baseline (1978 to 1980) questionnaire because this question was not repeated at the re-examination. This assumes that these men, who were aged 40 to 59 years at baseline, did not have any further children. At the 1998 to 2000 follow-up, men were asked to report the birth weights of their offspring, and 97% gave the same number of birth weights as they had number of children at baseline. For simplicity throughout the text, number of children is used to describe the number of live births to women and number of children of the men. CHD was defined as self-report of a doctor diagnosis of myocardial infarction, angina, or heart failure and/or a record of a diagnosis of any one of these conditions in the medical records.15
The association between number of children and prevalent CHD was first assessed graphically. To assess a possible nonlinear relationship (ie, J-shaped curve), which has been previously reported,2,16 age-adjusted quadratic models were fitted for women and men separately; interaction tests of both the linear and quadratic components identified whether these patterns were similar between women and men. Multiple logistic regression was used to assess the association between number of children and prevalent CHD. Because homeostasis model assessment (HOMA; a measure of insulin resistance) scores are not calculated for subjects with diabetes, this measurement could not be entered into a regression model without exclusion of all individuals with diabetes. To enter full data on insulin-resistance diabetes status in the final models, a categorical variable combining both HOMA scores and diabetic status was created, with the first 5 categories defined by the quintiles of the HOMA scores and a sixth category containing individuals with diabetes. Triglyceride levels and HOMA scores were log-normally distributed; geometric means are presented in tables, and the natural log of these was used in the regression models.
Of the 4286 women who responded, 3828 (90%) provided details of numbers of live births. Complete data on number of children for all 4252 participants in the follow-up of the men were available. The distributions of numbers of children were similar for the women and men (mean of 2.3±1.4 children for women versus 2.2±1.3 for men), and the modal number of children for both women and men was 2, with 39% of women and 40% of men having 2 children.
Association Between Number of Children and Metabolic Risk Factors
The relationship between CHD risk factors and numbers of children for women and men are shown in Tables 1 and 2⇓. Among both women and men, there was a positive linear relationship between number of children and body mass index and waist-hip ratio. However, whereas the strength of relationship for women and men was similar for waist-hip ratio (P=0.36 for interaction), the relationship with body mass index observed in women was stronger than that observed in men (P=0.002 for interaction). In men, the association between number of children and waist-hip ratio was attenuated to the null after adjustment for childhood social class, lifestyle risk factors (smoking, alcohol, physical activity, and marital status), and indicators of adult socioeconomic position (social class, car ownership, and housing tenure), whereas the association with body mass index remained unaltered. In women, both of these associations remained unaltered after full adjustment. In women only, number of children was also associated with lower high-density lipoprotein cholesterol, higher triglyceride levels, and increased diabetes prevalence. These associations were not affected by adjustment for potential confounding factors. No linear relationships were found between number of children and insulin resistance, systolic blood pressure, or diastolic blood pressure in either sex, though women with 5 or more children had high HOMA scores.
Association Between Number of Children and Socioeconomic Position
Among women, there were strong linear associations between number of children and all indicators of adult socioeconomic position (social class, car ownership, and housing tenure) and also with childhood social class (Table 1). Among men, there were no linear associations between number of children and either adult or childhood social class or car ownership, although men with ≥5 children were more likely to be from manual social classes in both childhood and adulthood and less likely to have access to a car (Table 2). There was a positive linear association between number of children and living in local authority housing among men.
Association Between Number of Children and Prevalent CHD
The age-adjusted and fully adjusted prevalence of CHD by number of children for women and men are shown in the Figure. In both women and men, there were “J” shaped associations between number of children and age-adjusted prevalent CHD, with the prevalence being lowest among those with 2 children (probability values for quadratic components: P=0.037 for women and P=0.004 for men). Tests for interaction showed that the shape and magnitude of these associations did not differ between women and men (interaction probability values for both the linear and quadratic terms were >0.2). For those with at least 2 children, each additional child increased the age-adjusted odds of CHD by 30% (odds ratio, 1.30; 95% confidence interval, 1.17 to 1.44) for women and by 12% for men (odds ratio, 1.12; 95% confidence interval, 1.02 to 1.22).
The associations between number of children and prevalent CHD for those with at least 2 children, together with the effect on these associations of adjustment for potential confounding and explanatory factors, are shown in Table 3. Separate adjustments for socioeconomic position, lifestyle risk factors, and metabolic risk factors are presented so that the role of each in the causal pathway between number of children and CHD can be assessed. Adjustment for obesity and metabolic risk factors attenuated the associations between greater number of children and CHD in both sexes, although in women, some association remained. Adjustment for socioeconomic position and lifestyle risk factors (smoking, alcohol, and physical activity) also attenuated the association between greater number of children and CHD in women, but adjustment for these factors had little effect in men.
Increasing number of children was associated with increasing obesity in both sexes and, in women, was also associated with adverse lipid profiles and diabetes. There was a similar J-shaped association between number of children and CHD in both sexes. Adjustment for obesity and metabolic risk factors attenuated the associations between greater number of children and CHD in both sexes, although in women, some association remains.
This study is cross-sectional and may be affected by reverse causality and survivor bias. Reverse causality (CHD leading to increased numbers of children) seems unlikely. Survivor bias (those with a greater number of children dying prematurely of CHD) would lead to underestimation of the strength of association between number of children and CHD. Because men are a cohort of survivors from the 1970s, the association in men may be underestimated, and the true number of children—CHD association in men may be closer to that of women. Among women, the similarity between the findings in our study and those of prospective studies suggests that this is not an important source of bias in women.1–4 The ascertainment of number of children differed between women and men in this study. The similarity in distributions in numbers of children between women and men in this study and, in addition, the similarity between these distributions and fertility rates in England and Wales during the 1940s to 1970s (when these children will have been born)17 suggest that these differences will not have importantly biased our results.
Previous studies have emphasized 2 possible biological mechanisms for the association between parity and CHD in women. In the first, it is proposed that each pregnancy permanently “resets” ovarian function, leading to a reduced lifetime exposure to estrogen.18 The established inverse association between parity and breast cancer, a known estrogen-dependent disease, provides support for this hypothesis.19 However, the role of endogenous estrogen in the cause of CHD in women is unclear.9,20,21
A second biological mechanism suggests that because normal pregnancy is a state of relative insulin resistance, repeated pregnancies may result in permanent detrimental effects on lipid and glucose metabolism.5–7 Most studies that have shown an association between pregnancy and abnormal metabolic risk factors have been conducted on pregnant women or women of reproductive age,5–7 and it is unclear whether any effect persists into later life. One recent study found that parity was associated with obesity, low high-density lipoprotein cholesterol, and insulin resistance in women aged ≥55 years.22 Although we found an association between number of children and some metabolic risk factors in women, this did not fully explain the association between number of children and prevalent CHD; however, it confirms the previous findings of an association between number of children and carotid plaques in older women, which also remained after adjustment for lipid levels, insulin resistance, and obesity.22 In neither that study nor our study did we adjust for coagulation factors, which have also been found to be adversely affected by pregnancy.6 If these adverse effects of pregnancy persist into later life in mothers in the same way that lipid abnormalities seem to, then the remaining association with increased number of children after adjustment for components of the insulin resistance syndrome and potential confounding factors may be due to adverse coagulation in women with more children.
Our findings in men suggest that lifestyle factors associated with having large families may lead to obesity and increased CHD risk in both sexes. The increased prevalence of CHD in women and men with no children or only one child may reflect ill-health among these individuals. It is interesting to note that the risk factors for CHD that we have found to be associated with increased numbers of children in women are also those associated with pregnancy complications such as preeclampsia, preterm delivery, and small-for-date babies.5 Women with these complications of pregnancy may have limited the number of further pregnancies. In addition, polycystic ovarian syndrome provides a link between infertility and increased risk factors for CHD, which might explain the increased risk of CHD among women with one or no children,5 although the largest study of the association between polycystic ovarian syndrome and actual CHD found an increased risk for stroke only and not for CHD.23 Clearly, pregnancy complications and polycystic ovarian syndrome could not explain the increased risk of CHD among men with none or only one child.
The British Regional Heart Study and the British Women’s Heart and Health Study are funded by the Department of Health. The British Regional Heart Study is a British Heart Foundation Research Group. Dr Lawlor is funded by a Medical Research Council/Department of Health training fellowship. The views expressed in this publication are those of the authors and not necessarily those of the funding bodies. We thank all of the general practitioners and their staff who have supported data collection and the women and men who have participated in the studies.
Ness RB, Schotland HM, Flegal KM, et al. Reproductive history and coronary heart disease risk in women. Epidemiol Rev. 1994; 16: 298–314.
Green A, Beral V, Moser K. Mortality in women in relation to their childbearing history. BMJ. 1988; 297: 391–395.
Sattar N, Greer IA. Pregnancy complications and maternal cardiovascular risk: opportunities for intervention and screening? BMJ. 2002; 325: 157–160.
Greer IA. Thrombosis in pregnancy: maternal and fetal issues. Lancet. 1999; 10: 1258–1265.
Barrett-Connor E. Sex differences in coronary heart disease: why are women so superior? The 1995 Ancel Keys Lecture. Circulation. 1997; 95: 252–264.
Kravdal O. Is the relationship between childbearing and cancer incidence due to biology or lifestyle? Examples of the importance of using data on men. Int J Epidemiol. 1995; 24: 477–484.
Shaper AG, Pocock SJ, Walker M, et al. British Regional Heart Study: cardiovascular risk factors in middle-aged men in 24 towns. BMJ. 1981; 283: 179–186.
Lawlor DA, Bedford C, Taylor M, et al. Geographic variation in cardiovascular disease, risk factors and their control in older women: British Women’s Heart and Health Study. J Epidemiol Commun Health. 2003; 57: 134–140.
Wannamethee SG, Lowe GDO, Whincup PH, et al. Physical activity and hemostatic and inflammatory variables in elderly men. Circulation. 2002; 105: 1785–1790.
Walker MK, Whincup PH, Shaper AG, et al. Validation of patient recall of doctor-diagnosed heart attack and stroke: a postal questionnaire and record review comparison. Am J Epidemiol. 1998; 148: 355–361.
Colditz GA, Willett WC, Stampfer MJ, et al. A prospective study of age at menarche, parity, age at first birth, and coronary heart disease in women. Am J Epidemiol. 1987; 126: 861–870.
OPCS. Birth Statistics: Historical Series of Statistics From Registrations of Births in England and Wales, 1837–1983. Series FM1. No. 13. London: Her Majesty’s Stationary Office; 1987.
McPherson K, Steel CM, Dixon JM. ABC of breast diseases: breast cancer: epidemiology, risk factors, and genetics. BMJ. 2000; 321: 624–628.
Lawlor DA, Ebrahim S, Davey Smith G. Sex matters: secular and geographical trends in sex differences in coronary heart disease mortality. BMJ. 2001; 323: 541–545.
Lawlor DA, Ebrahim S, Davey Smith G. The role of endogenous oestrogen in the aetiology of coronary heart disease: analysis of age related trends in coronary heart disease and breast cancer in England and Wales and Japan. BMJ. 2002; 325: 311–312.
Humphries KH, Westendorp IC, Bots ML, et al. Parity and carotid artery atherosclerosis in elderly women: the Rotterdam Study. Stroke. 2001; 32: 2259–2264.