Back

Is anxiety a pathway to Alcohol Use Disorders? A phenome-wide association study of the GABRA2 coding variant rs279858

Merikangas, A. K.; Kember, R. L.; Plawecki, M. H.; Kamarajan, C.; Chan, G.; Bauer, L.; Meyers, J. L.; Nurnberger, J. I.; Kramer, J.; Porjesz, B.; Edenberg, H. J.; Almasy, L.

2022-11-22 genetic and genomic medicine
10.1101/2022.11.21.22282301 medRxiv
Show abstract

Alcohol use disorders (AUDs) and related electrophysiological endophenotypes have been associated with the GABRA2 gene. However, the causal variants in GABRA2 and their mechanisms of influence on AUD and its correlates have not been established. Here we investigate the phenotypic spectrum of a GABRA2 coding variant (rs279858) through a phenome-wide association study (PheWAS) in two open-source datasets. We applied the PheWAS approach to identify a broad range of phenotypes associated with rs279858 in the MRC IEU OpenGWAS PheWAS and the Open Targets Genetics Portal. These databases extend the array of phenotypes beyond those available in electronic health records (EHR) to include numerous non-medical phenotypes and traits. We then followed up the results from those exploratory associations by examining the genetic correlations between our "top hits" and alcohol- and smoking-related phenotypes. In both data sources, rs279858 (C effect allele) was associated with anxiety-related phenotypes, including reduced risk-taking behavior and an increase in nervous feelings, as well as reduced number of lifetime sexual partners. Follow-up analyses revealed that these phenotypes were genetically correlated with each other and with alcohol- and smoking-related phenotypes. This work illustrates the utility of the PheWAS approach, particularly for phenotypes that extend beyond those that are typically captured in EHR data. In fact, the associations described here are all behavioral rather than clinical phenotypes. We postulate that these traits may be related to anxiety or behavioral inhibition that has been identified as a risk factor for AUD, and may represent pathophysiological intermediaries between GABRA2 and AUD.

Matching journals

The top 2 journals account for 50% of the predicted probability mass.