Abstract P103: Associations Between Brain Volumes and Cerebral Blood Flow and Sex Hormones in the Look AHEAD Brain MRI Cohort

Background: Females have greater relative brain volume (BV) and cerebral blood flow (CBF) compared to males. BV decreases after menopause suggesting a possible role of sex hormones. We studied the association of BV, abnormal white matter hyperintensity volumes (WMHV) and cerebral blood flow (CBF) with sex hormones in adults with type 2 diabetes (T2DM), which is associated with risk of brain atrophy. Methods: The sample was 215 participants with overweight or obesity and T2DM from the Look AHEAD Brain MRI cohort (mean age 68 years [SD: 6.3], 27% male, 73% female [all postmenopausal]; without exogenous hormones), who had brain MRIs to evaluate their total BV, WMHV and CBF. The ratio of brain measurements to intracranial volume was analyzed to account for body size. Estradiol (E2) and testosterone levels (T) were estimated with electrochemoluminescence assays. In females, who have low E2 and T levels, we compared brain measures in those with detectable (vs. undetectable) hormone levels (E2<20 pg/mL, 79%; T<0.02 pg/mL, 37%). In males, we used Spearman correlation to assess the association between brain measures and sex hormone levels. The associations between BV, WMHV and E2 and T were adjusted for age and BMI using linear regression. Results: Females with detectable (vs. undetectable) T levels had higher BV (p=0.04) (Table), which was attenuated after adjustment for age and BMI. WMHV and CBF were not statistically associated with sex hormone levels in females. In males, no brain measures were significantly associated with sex hormones levels. Conclusions: In postmenopausal females with T2DM, detectable levels of T were associated greater BV, but not associated any CBF or WMHV. In males, none of the brain measures were associated with sex hormones. Our findings are limited by a small, convenient sample size and low sensitivity of hormone assays with a high proportion of undetectable levels. Our findings suggest that larger samples with high sensitivity hormone assays are needed to assess clinically important differences.