Figure 4. Pooled cumulative incidence and multivariable-adjusted associations of menopause and leukocyte telomere length (LTL) categories with incident coronary artery disease (CAD) events.

The upper panels show the pooled cumulative incidence of CAD events among (A) all postmenopausal women, (B) women without a history of gynecologic surgery, and (C) women with a history of gynecologic surgery during follow-up truncated at 12.7 years. The lower panel (D) shows the adjusted HRs for incident CAD events during follow-up meta-analyzed across cohorts. P-values correspond to Cox regression models adjusted for age, age², the first ten principal components of genetic ancestry, race/ethnicity, current/former smoking status, body mass index, diabetes status, current hormone therapy use, antihypertensive medication use, cholesterol-lowering medication use, systolic blood pressure, total cholesterol, and high-density lipoprotein cholesterol. Analyses in the WHI were further adjusted for inverse probability of sampling weights, while those in the UK Biobank were further adjusted for Townsend deprivation index. Menopause status and LTL were included as combined nonordered, categorical variables, with women without premature menopause and above-median LTL constituting the reference category. No corrections for multiple comparisons were made.