Multi-level genomic analysis identifies pleiotropic genetic liability for insomnia, anxiety and depression

Insomnia (INS) is a major factor determining the risk, severity, treatment resistance and relapse of depression (DEP) and anxiety disorders (ANX), suggesting a shared neurobiological vulnerability. Identifying common biological mechanisms that are therapeutically actionable is a foundational step towards improving current clinical care. Since INS, DEP, and ANX share a substantial proportion of their genetic risk, pinpointing the specific genetic factors that influence these three conditions simultaneously offers a route to discover pharmacological targets with transdiagnostic potential. We developed a novel framework of multi-level trivariate genetic analysis to identify the shared genetic component of INS, DEP and ANX using genome-wide association studies (GWAS). We trace shared components through global and local genetic correlations, overlapping genome-wide significant loci, shared effector genes, and variants with shared effects. We show that 55% of the genetic signal is shared across all three conditions. We identify 195 genomic loci with shared signal for at least two conditions, many of which are likely arising from the same causal variant (at least 50%) or effector gene (60-80%). Pairwise mediation analyses suggest that these shared likely causal variants are more consistent with models of vertical pleiotropy where DEP and INS are risk factors towards ANX, rather than with models of horizontal pleiotropy. We find convergence of shared effector genes on biological processes such as inhibitory synaptic transmission, neuronal organization, and axonal development. We find that the majority of shared effector genes that are also druggable are involved in synaptic transmission and located in synaptic membranes, identifying CACNA2D3, DRD2, GRIA1 and GRM5 as key targets for drug development to treat comorbid INS, DEP and ANX. These results reveal an interconnected but mechanistically diverse basis for shared genetic risk across INS, DEP and ANX, and offer potential candidates for future pathway-tailored transdiagnostic therapeutic targets.