Investigating associations of age at natural menopause with neuroimaging indicators of brain health in later‐life, using a British population‐based birth cohort

Abstract

AbstractBackgroundEarlier menopause age associates with poorer cognitive performance post‐menopause, but links with dementia risk are mixed and the potential mechanisms underlying such associations are unclear. Few studies have examined the relationship between menopause age and later‐life brain health.MethodOur study included 126 women from Insight 46, the neuroscience sub‐study of the 1946 British birth cohort, which included combined amyloid PET and MRI scanning at mean age 70.7 years (standard deviation = 0.66). Age at natural menopause was self‐reported, during midlife, as years since birth at the time of period cessation; women who reported surgical menopause were excluded (n = 63). We tested for interactions of menopause age with APOE‐ε4 status and used multivariable regression analyses to test associations of menopause age with later‐life total brain volume (TBV) and hippocampal volume. A generalised linear model with log‐linked gamma distribution tested associations with white matter hyperintensity volume (WMHV). Models were cumulatively adjusted for total intracranial volume and age at scan (model 0/M0), hormone therapy (HT) use (M1:M0+HT) and for prospective early‐life and sociodemographic (M2:M1+childhood cognition, education, adult socioeconomic position), reproductive (M3:M2+menarche age, parity), and health‐related (M4:M3+smoking, BMI and blood pressure at age 36 years, APOE‐ε4 status) covariables.ResultThe association of menopause age with WMHV was modified by APOE‐ε4 status (p = 0.024); each 1‐year increase in menopause age associated with a 7.4% decrease in WMHV for ε4 carriers (n = 35) and a 2.4% WMHV increase for non‐carriers (n = 90). Menopause age did not associate with hippocampal volume, but there was a positive association with TBV (Table 1) which remained significant until we accounted for health‐related factors in model 4 (Figure 1). In post‐hoc analysis, adjustment for blood pressure at age 36 attenuated the association by 11.24%, explaining more of the attenuation than other model 4 covariables.ConclusionWe find evidence for links between age at natural menopause and later‐life neuroimaging measures. Later menopause associated with greater brain volume, partly explained by midlife blood pressure. Interactive effects of APOE‐ε4 status on associations with WMHV, an indicator of brain pathology, warrant further investigation.