Reproductive lifespan and cardiovascular health in postmenopausal women: a non-linear, behavior-driven association with cross-population validation

Reproductive lifespan (RLS), the interval from menarche to natural menopause, has been associated with cardiovascular disease (CVD) risk. However, most prior studies have assumed a linear relationship, and the potential non-linearity of this association, as well as the underlying behavioral and biological mechanisms, remain largely unexplored.

This cross-sectional study included 1,673 naturally postmenopausal women (weighted N = 6,885,059) from eight NHANES cycles (1999-2010 and 2017-March 2020). Cardiovascular health was assessed using the American Heart Association's Life's Simple 7 (LS7) score. Restricted cubic spline and segmented non-linear analyses examined non-linearity. Mediation analyses evaluated indirect effects of alkaline phosphatase (ALP) and the neutrophil-to-lymphocyte ratio (NLR). Sensitivity analyses decomposed LS7 into behavioral and biological sub-scores. An independent cohort of 303 Chinese postmenopausal women was used to assess cross-population generalizability.

RLS exhibited a significant inverted U-shaped association with LS7 (P for non-linearity = 0.025). In the ascending segment (RLS ≤ 37 years), each additional year of RLS was associated with a higher LS7 score (β = 0.038, P = 0.048); beyond 37 years, the association became non-significant. NLR significantly mediated this association (P < 0.001; 8.8% mediated), whereas ALP showed no mediation. The non-linear association was driven entirely by the behavioral sub-score, with no contribution from the biological sub-score. The non-linear association was independently validated in the Chinese cohort (P for non-linearity = 0.014), and the sLS7 score differed significantly across RLS tertiles (5.71 vs. 5.69 vs. 4.94, P = 0.001).

In naturally postmenopausal women, RLS showed an inverted U-shaped association with cardiovascular health, peaking at approximately 37 years. This association was partially mediated by NLR and driven entirely by behavioral components. These findings support incorporating reproductive history into cardiometabolic risk assessment and highlight behavioral interventions for women with shorter reproductive lifespans.