Context: Polycystic ovary syndrome (PCOS) is closely linked to hyperandrogenism (HA). In PCOS, HA has been associated with metabolic disturbances that increase the risk for cardiovascular disease (CVD).
Objective: To assess the association of high serum androgen levels, as a postmenopausal remnant of PCOS, with the prevalence of atherosclerosis and incidence of CVD in postmenopausal women.
Design: The Rotterdam Study, a prospective population-based cohort study. Median follow-up was 11.36 years.
Setting: General community.
Participants: A total of 2578 women aged >55 years. Exclusion criteria were missing informed consent or follow-up data, perimenopausal status, and menopause by surgical intervention or at an unnatural age (age <40 or >62).
Intervention: None.
Main outcomes and measures: Linear, logistic, and Cox regression models assessed the association of top quartiles (P75) of serum testosterone, free androgen index (FAI), dehydroepiandrosterone, and androstenedione and sex hormone-binding globulin with coronary artery calcium, carotid intima-media thickness (IMT), pulse wave velocity, peripheral artery disease, and incidence of coronary heart disease (CHD), stroke, and CVD.
Results: Mean age (standard deviation) was 70.19 (8.71) years, and average time since menopause was 19.85 (9.94) years. Highest quartile FAI was associated with higher pulse wave velocity (β [95% confidence interval (CI)], 0.009 [0.000 to 0.018]). Highest quartile dehydroepiandrosterone [β (95% CI), -0.008 (-0.015 to -0.001)] and androstenedione [β (95% CI), -0.010 (-0.017 to -0.003)] levels were associated with a lower IMT. We found no association between high androgen levels and incident stroke, CHD, or CVD.
Conclusion: Postmenopausal high androgen levels were not associated with an elevated risk for CVD. Cardiovascular health in women with PCOS might be better than was anticipated.