AbstractBackgroundHippocampal atrophy is an important pathophysiological event in Alzheimer’s disease (AD) etiopathology. Still the mechanistic link between these volumetric abnormalities and AD remains poorly understood. In this study, we estimated polygenic overlap with mixed effect directions between AD and bilateral hippocampal volumes, and identified commonly shared genetic variants between two traits, involved in several synaptic and immunomodulatory genes.MethodFor this study we used GWAS summary statistics on AD (1,11,326 cases and 6,77,663 controls) (Bellenguezet et al., 2022) and hippocampal volumes (33,224 populations controls) (Smith et al., 2021). First, we applied Bivariate causal mixture model (MiXeR version1.3, Frei et al., 2021) to quantify the polygenic overlap between AD and bilateral hippocampal volumes. For each AD‐volume pair, MiXeR estimated the shared polygenic architecture between two traits, irrespective of the direction of individual variants effects.To explore the MiXeR estimated shared genetic basis at individual variant level, we used the conjunctional false discovery rate (conjFDR) analysis (Andreassen et al., 2013) on the association of AD with bilateral hippocampal volumes, and vice versa. The conjFDR analysis is built on an empirical Bayesian statistical framework. Four regions (MHC, 8p23.1, MAPT, APOE) with complex LD patterns were excluded from the analyses because complex correlations in regions with intricate LD might bias the estimation.ResultMiXeR univariate analysis estimated the number of AD risk variants to be around 343 genetic variants, and the number of variants for bilateral hippocampal volumes between 901 and 942 variants. MiXeR bivariate analysis estimated that the number of shared variants between AD and bilateral hippocampal volumes was around 75 and 96 variants (see Figure 1). At conjFDR < 0.05, we identified 10 common loci shared by AD and left hippocampal volume,6 lead SNPs had consistent effect directions, while 4 loci have opposite effect directions according to their Z value. 8 common loci shared by AD and right hippocampal volume, 5 lead SNPs have consistent effect directions, 3 have opposite effect directions. (see Figure 2 and Table 1)ConclusionThis study provides evidence of a shared genetic architecture between AD and bilateral hippocampal volumes. The identification of common loci helps to understand the role of hippocampus in AD pathology.