Associations of endogenous and exogenous hormonal exposures and cardiovascular disease in women – A FinnGen study

Background: Earlier age at menopause and shorter reproductive span (time from menarche to menopause) have been linked to an increased risk of cardiovascular diseases (CVD), presumably because of limited lifetime exposure to endogenous estrogen. Our intention is to determine whether genetic liability to earlier vs later menarche and menopause are associated with risk of common cardiovascular diseases in women. We also aim to investigate effects of exogenous estrogen exposure in the form of systemic hormonal contraceptive and menopausal hormone therapy use. Methods and results: We determined GWAS summary statistics for age at menopause, age at menarche , and length of the reproductive period with data derived from UK Biobank (UKBB) European-ancestry participants. We calculated polygenic scores for women from FinnGen study (N=184 132) that indirectly capture genetic variation in endogenous estrogen exposure: age at menopause, age at menarche, and length of reproductive span. We also investigated exogenous hormone exposure associated with use of menopausal hormone therapy or systemic hormonal contraceptives. We used Cox proportional hazards model to investigate associations between PGSes and exogenous estrogen exposure and risk of hypertension, stroke, and coronary heart disease (CHD) events. Median follow-up time was 25.8 years. During the follow-up 56 143 women experienced hypertension, 18 200 women experienced strokes, and 13 879 women experienced major CHD events. Genetic liability to later menopause and longer reproductive span were weakly associated with a higher risk of stroke: in models adjusted with smoking and BMI, the hazard ratio (HR) per one standard deviation increase in PGS was 1.03 [1.01-1.05] for both. Menopausal hormone therapy use was associated with lower risk of stroke (HR 0.85 [0.82-0.89]) and CHD (HR 0.80 [0.76-0.84]). Systemic hormonal contraceptive use was associated with lower risk of hypertension (HR 0.96 [0.93-0.99]), stroke (HR 0.84 [0.80-0.99]) and CHD events (HR 0.83 [0.78-0.89]). Conclusions: Although observational evidence consistently associates a longer reproductive span with lower cardiometabolic risk, the polygenic component of metabolic timing (PGSes for reproductive span and age at menopause) showed the opposite direction of association. This discrepancy likely reflects the fact that these scores capture genetic pathways only partially overlapping with phenotypic lifetime estrogen exposure. Importantly, the observed cardioprotective associations of menopausal hormone therapy and systemic hormonal contraception underscore that genetic predisposition and exogenous hormonal exposure represent distinct biological dimensions relevant to cardiovascular risk in women.