[o2–08–02]: sex‐specific analysis of the adsp case‐control whole‐exome sequencing dataset

Women comprise nearly two-thirds of all Alzheimer disease (AD) cases, suggesting gender-specific risk and protective factors. For example, a number of studies have established that apolipoprotein E (APOE) genotype modifies risk of AD differently in men and women, with the ApoE4 allele conferring greater risk in women and the neuroprotective effects of the ApoE2 allele being more prominent in men. Several other candidate gene studies implicating sex-specific effects on AD from genes such as BDNF, LDLR and ABCA1, also suggests biological differences between the sexes are important in development and progression of AD. Thus, we analyzed the Alzheimer's Disease Sequencing Project (ADSP) case-control whole-exome sequencing (WES) dataset using a sex-stratified approach in order to identify sex-specific genetic associations with AD. Sex-stratified analysis was performed on non-Hispanic white (NHW) subjects in the ADSP case-control WES study (N=10,436; 56.4% females cases, 59.1% female controls) to investigate genetic risk differences between males and females. Both single-variant (logistic regression) and gene-based tests (SKAT-O) were performed separately in males and females. All analyses adjusted for age-of-onset /age-at-exam, principal components and sequencing center. Results for males and females were evaluated separately for known and novel associations. Heterogeneity tests were applied to test for different effects on male and female risk. Gene-based testing discovered three significant sex-specific associations for AD in the genes ZNF471 (male P=0.986; female P=1.26x10-6), AP4S1 (male P=2.31x10-7; female P=0.183) and TOX (male P=1.30x10-6; female P=0.136). Additionally, several variants show marked differences in strengths of association between males and females, including a synonymous variant in the known GWAS risk gene PICALM that icreases risk for males (P=8.57x10-6) but shows no association in females (P=0.414) (chi-square heterogeneity P=0.004). Replication of our top results is ongoing. We identified several sex-specific associations with AD. AP4S1, for which recessive mutations cause the neurological disease hereditary spastic paraplegia, is part of the endocytic pathway, a known AD pathway. TOX is a growth-hormone-responsive transcription factor involved in immune-response, and ZNF471 is hypermethylated in nonagenarians. Understanding the nature of these significant associations could help explain differential risk and progression for AD between the sexes.