Reproductive patterns and heart failure risk: insights from five aggregated cohorts
E L Shalowitz, K Sompel, K Leyba, E Kudron, A Herrera Heredia, A Hendricks, D P KaoAbstract
Background
Women experience departures from normal cyclic reproductive hormone exposures during pregnancy, childbirth, and breastfeeding. The importance of these disruptions in relation to heart failure (HF) risk is not well understood.
Purpose
To determine if number of pregnancies, births, age at first, and breastfeeding is associated with HF risk.
Methods
Data from the Atherosclerosis Risk in Communities (ARIC), Cardiovascular Health Study (CHS), Framingham Heart Study (FHS), Jackson Heart Study (JHS), and Multiethnic Study of Atherosclerosis (MESA) comprising 25,253 women were obtained from the NHLBI’s BioLINCC resource and harmonized. Primary outcomes were incidence as adjudicated by each study and age at onset of HF. We evaluated associations between primary outcomes and pregnancy, birth, and breastfeeding history.
Results
Comparisons between women with vs. without incident HF from all studies are summarized in Figure 1. Association between variables of interest and age of HF onset are shown in Figure 2. As observed previously, later menopause and greater total years of menses were associated with less incident HF and later age of onset. Women without HF had fewer pregnancies than women with HF (Fig 1). In individual studies, concordant findings were observed in ARIC (3.4±2.3 vs. 3.9±2.6), JHS (3.8±2.5 vs. 5.1±3.5), and MESA (3.8±2.5 vs. 5.1±3.5, p<0.001 for all). Trends were concordant in FHS Gen III (2.2±1.5 vs. 2.8±1.7, p=0.14), CHS (2.6±1.9 vs. 2.7±1.9, p=0.18), and FHS Original (2.1±1.7 vs. 2.2±1.7, p=0.23). Women without HF had fewer live births compared to women with HF (Fig 1). In individual studies, concordant findings were observed in ARIC (3.2±1.9 vs. 3.6±2.2) and JHS (3.4±2.2 vs. 4.7±2.8, p<0.001 for both) and trends were concordant in CHS (2.6±1.5 vs. 2.7±1.7, p=0.38), FHS Offspring (2.5±1.4 vs. 2.7±1.5, p=0.14), and FHS Gen III (1.7±1.2 vs. 2.4±1.3, p=0.03). Earlier age at first birth was associated with increased risk of HF (Fig 1). Findings were observed in ARIC (22.4±4.2 vs. 21.7±4.2, p<0.001). Trends were concordant in all other available studies: FHS Offspring (24.6±4.8 vs. 23.9±7.4, p=0.27) and FHS Gen III (26.3±6.9 vs. 25.7±8.4, p=0.83). In studies that reported breastfeeding prevalence (ARIC and FHS Gen III), there was no difference between women with and without HF, however history of any breastfeeding was associated with earlier HF onset (Fig 2, β= -3.78±0.47 years, p<0.001).
Conclusions
Fewer pregnancies and births, and later age at first birth were associated with lower incidence and later onset of HF across several decades of observation. The mechanism is unclear, but may relate to interruptions in normal hormonal cycling, which was observed again to be associated with less HF. Breastfeeding was associated with earlier HF onset, though data were limited. Future research should examine how pregnancy, childbirth, and breastfeeding impact HF risk and can help guide prediction and prevention strategies.For image description, please refer to the figure legend and surrounding text.For image description, please refer to the figure legend and surrounding text.