Abstract
The role of estradiol in depression and Alzheimer's disease - brain disorders that disproportionately affect females - is debated. Results from observational studies are inconsistent and limited by confounding and reverse causation. To overcome these limitations, we perform two-sample Mendelian randomization. We run genome-wide association studies on sex-specific brain age gap, a proxy of brain health, and female-specific estradiol levels using data from the UK Biobank. We test for causal links between genetically-predicted factors related to estradiol exposure (estradiol levels in pre- and postmenopausal samples, reproductive span, age at menarche, age at menopause, number of childbirths) and brain age gap, Alzheimer's disease and depression as outcomes. We replicate our analyses on estradiol levels in males. Here, we find no significant associations between estradiol exposure and brain health across samples and robust methods, indicating an absence of constant causal effects and suggesting that hormonal fluctuations may drive links between estradiol and brain health.