Exploring the pleiotropic genetic architecture of substance use disorders: Findings from phenome-wide association studies

Complex and common, substance use disorders (SUDs) are heterogeneous in presentation and difficult to treat. Moreover, SUDs are frequently comorbid with other psychiatric disorders and medical issues. While it is known that SUDs are moderately heritable, the underlying genetic etiology of SUDs, and their comorbidities, has remained elusive. Here, we synthesize findings from two types of datasets, namely electronic health record (EHR)-based datasets and deeply phenotyped datasets, to examine genetic liabilities for substance use and associations with other phenotypes. Each dataset type brings unique strengths and limitations to our understanding of the genetics of SUDs. We review phenome-wide association studies (PheWAS) of genetic liability for four common substances of abuse (alcohol, opioid, tobacco, and cannabis) as well as a transdiagnostic addiction factor. These analyses consistently identified associations with other substances, psychiatric disorders, demographic characteristics, and comorbid medical conditions. We conclude the review by proposing the ideal dataset and future directions for the field.