Reproductive factors and brain volume in community-dwelling individuals: the Ohasama Study

Reproductive history can be conceptualized as a life-course exposure integrating biological and social factors. Although reproductive factors have been associated with stroke risk and cognitive function, their relationships with structural brain changes remain unclear. Evidence from general population-based studies linking reproductive factors to brain volume, an intermediate marker of cognitive decline, is limited. This study examined whether age at menarche, age at menopause, estrogen exposure duration, and parity are associated with brain volume in postmenopausal women.

This cross-sectional study included 361 postmenopausal women aged ≥55 years from the Ohasama Study in Japan. Reproductive factors were obtained from a 1998 questionnaire were categorized into quartiles, and MRI data from 2009 to 2019 were analyzed with voxel-based morphometry. Associations were assessed using analysis of covariance adjusted for intracranial volume, age, birth cohort, and BMI. Effect sizes were reported as percent change and standardized effect sizes, with the smallest effect size of interest (SESOI) based on one-year age-related brain volume change.

After covariate adjustment, most associations between reproductive factors and brain volume were attenuated. Earlier age at menarche showed a weak trend toward larger temporal lobe volume (β = 0.59 cm³, 95% CI -0.03 to 1.21; p = 0.06), corresponding to a 0.63% change and 0.06 SD, which was comparable to the SESOI (±0.54% or 0.06 SD). Earlier age at menopause was associated with smaller occipital lobe volume (β = -0.52 cm³, 95% CI -1.01 to -0.03; p = 0.04), corresponding to a -1.08% change and 0.10 SD, exceeding the primary SESOI threshold (±0.64% or 0.06 SD), although confidence intervals included effects smaller than the SESOI. No significant associations were found for estrogen exposure or parity.

Reproductive factors were not associated with moderate or large differences in brain volume in postmenopausal women. The slight, region-specific differences observed were comparable in magnitude to short-term age-related changes. These findings suggest that the clinical significance at the individual level is likely limited. However, subtle associations at the population level cannot be entirely excluded.