Skip to content

Mullins et al. 2021 — Genome-wide association study of more than 40,000 bipolar disorder cases provides new insights into the underlying biology

One-paragraph summary

The Psychiatric Genomics Consortium analyzed 41,917 bipolar disorder cases and 371,549 controls of European ancestry. Sixty-four genomic loci reached genome-wide significance. Risk signal was enriched in synaptic-signaling and brain-expressed genes, especially neurons of prefrontal cortex and hippocampus; expression integration prioritized 15 genes, including potentially druggable targets HTR6, MCHR1, DCLK3 and FURIN. Bipolar I and bipolar II showed high but imperfect genetic correlation. The paper materially increased robust locus discovery but did not establish a diagnostic variant, clinical polygenic-score threshold or treatment selector (Mullins 2021, PMID 34002096).

Key findings

  • Sample: 41,917 cases and 371,549 controls, all described in the abstract as European ancestry.
  • Genome-wide significant loci: 64.
  • Enrichment: synaptic-signaling pathways and brain-expressed genes.
  • Cell/anatomical enrichment: neurons with specific expression in prefrontal cortex and hippocampus.
  • Drug-target enrichment included antipsychotics, calcium-channel blockers, antiepileptics and anesthetics.
  • Expression quantitative-trait integration robustly linked 15 genes to bipolar disorder.
  • Bipolar I and II shared substantial but incomplete genetic liability.

Limitations

  • Ancestry concentration limits immediate portability to global populations.
  • Case-control association does not identify the causal variant or direction of biological effect at every locus.
  • Enrichment among drug targets is not evidence that repurposing a particular drug will be effective.
  • The abstract does not report patient-level discrimination, calibration or treatment-response performance.
  • Subtype heterogeneity and ascertainment differences remain embedded in the pooled phenotype.

Why it matters

This study moved bipolar genetics from a small set of uncertain loci to a substantially denser, reproducible polygenic map. It supports neuronal and synaptic biology, supplies targets for functional work, and quantifies partial subtype distinction. Its clinical importance is mainly negative as well as positive: even at this scale, inherited risk is distributed and overlapping, so present genetics cannot replace longitudinal diagnosis.

Cited by wiki pages

  • Genetics and neurobiology