Feed

Bipolar disorder (BD) is a clinically heterogenous mental disorder with a diversity in clinical trajectories and treatment response. Twin studies have shown that BD is highly heritable, with estimates ranging from 60% to 90%. However, our understanding of this genetic component has yet to be translated into clinical interventions or prediction tools. In this review, we summarize findings from large-scale international collaborations, with the latest and largest study finding 298 genome-wide significant loci and 36 credible mapped genes associated with BD, alongside the first genes with an increased burden of rare genetic variants. We next highlight key biological insights from these findings, including widespread genetic overlap but distinct patterns of genetic correlation with other mental disorders and related traits; enrichment of gene expression within both brain-specific tissues and cell types and non-brain tissues, including pancreatic and large intestinal tissues; and the identification of novel drug targets and repurposing candidates, including calcium channel blockers. While these insights may take several years before they impact clinical practice, the potential for genetic advances to impact patient care through more accurate nosology and the development of clinically relevant prediction and stratification tools is more imminently achievable. We go on to summarize the most prominent methods for genetic prediction, including artificial intelligence-based methods, before discussing promising use-cases and key challenges to implementation. Looking to the future, cross-disciplinary collaboration, access to real-world data for robust validation, and pragmatic solutions that facilitate implementation across diverse healthcare systems will be crucial to achieving material advances in clinical care and improving outcomes for people with BD.