Do sex and gender modify the effect of apoe-4 on neurocognitive impairment? a multi-ancestry analysis in the uk biobank

Genetic sex and psychosocial factors relating to sex and gender influence a person's risk of developing neurocognitive impairment (NCI). Yet their role in mechanisms underlying APOE-ɛ4 pathophysiology of NCI remains unclear. We explore whether sex and gender independently modify the association between APOE-ɛ4 and NCI. We conducted effect modification analyses in N=364,793 participants from the UK Biobank pan-ancestry dataset (return #2442) without prevalent cardiovascular disease or NCI. APOE-ɛ4 carrier status was inferred based on diplotypes derived from phased genotypes. NCI cases were identified in hospitalization, mortality, and self-report questionnaire datasets. Gender was measured using a previously constructed literature-based femininity score (FS) that leverages six psychosocial factors. We estimated adjusted generalized linear models (GLM) for NCI, as random effects for ancestry groups were not informative. To evaluate APOE-ɛ4 effect modification by gender, we stratified models by sex and introduced interaction terms for APOE-ɛ4 and FS. We estimated conditional effects corrected for multiplicity to illustrate modification across FS. To evaluate APOE-ɛ4 effect modification by sex, we introduced interaction terms for APOE-ɛ4 and sex and conditioned on FS. Sensitivity analyses were conducted with NCI cases identified from primary care data (N=169,125). We identified N=6,123 participants with incident NCI in the sex-combined sample (1.7%), of which 2,990 were female. APOE-ɛ4 was associated with increased risk of NCI in females (p Both sex and gender influence the effect of APOE-ɛ4 on NCI in diverse ancestries. More analyses are needed to clarify mechanisms by which gender influences risk conferred by APOE-ɛ4 in both sexes.