The number of episodes of major psychiatric and substance use disorders as an index of genetic risk and genetic heterogeneity

Kenneth S Kendler,Henrik Ohlsson,Jan Sundquist,Kristina Sundquist
DOI: https://doi.org/10.1038/s41380-024-02727-x
2024-09-02
Abstract:We investigate whether number of episodes (NoEs) meaningfully reflect genetic risk and genetic heterogeneity for five primary disorders-Drug Use Disorder (DUD), Alcohol Use Disorder (AUD), Major Depression (MD), Bipolar Disorder (BD), and Schizophrenia (SZ) ascertained from Swedish population registries. We utilize Genetic Risk Ratios (GRR)-defined as the ratio of the genetic risk for secondary disorders to the genetic risk for the primary disorder-derived from Family Genetic Risk Scores (FGRS). For all five primary disorders, genetic risk rose robustly with increasing NoEs. For both AUD and DUD, the GRR for all six secondary disorders-selected to have a likely genetic relationship with the particular primary disorder-declined with increasing NoEs so that cases of AUD and DUD with high versus low NoEs had both a higher genetic risk and a purer genetic signal. With MD, genetic risk maximized at an intermediate NoEs. While the GRRs for AUD and DUD in MD cases dropped sharply with increasing NoEs, GRR for BD increased. For BD, genetic risk rose sharply with increasing NoEs while for all secondary disorders the GRRs showed a mixture of modest increases and decreases. Like AUD and DUD, but even more markedly, selecting BD cases with high rates of recurrence would produce a sample with a high overall genetic risk and a relatively homogeneous genetic signal. For SZ, genetic risk rose moderately with increases in NoEs. GRRs for other non-affective psychoses (ONAP) and autism spectrum disorder (ASD) fell quite slowly with increasing NoEs, and more rapidly for other secondary disorders. Cases of SZ with high recurrence rates had a high genetic risk and a relatively pure signal, albeit with contributions from ONAP and ASD. In summary, NOEs are a robust index of genetic risk and genetic heterogeneity across our primary disorders with important inter-disorder differences.
What problem does this paper attempt to address?