The predictive effect of family genetic risk scores as an indirect measure of causal effects of one disorder on another

Kenneth S. Kendler,Henrik Ohlsson,Jan Sundquist,Kristina Sundquist

DOI: https://doi.org/10.1017/s0033291723003847

IF: 10.592

2024-02-06

Psychological Medicine

Abstract:Background One potential cause of comorbidity is the direct causal effect of one disorder – A – on risk for subsequent onset of disorder B. Could genetic risk scores be utilized to test for such an effect? If disorder A causally impacts on risk for disorder B, then genetic risk for disorder A should be lower in cases of disorder A with v. without a prior onset of B. Methods In all individuals (n = 905 736) born in Sweden from 1980 to 1990, from six psychiatric and drug use disorders (major depression, anxiety disorders, alcohol use disorder, drug use disorder, bipolar disorder, and schizophrenia), we formed 14 pairs of disorders A and B. In these pairs, we compared, using Cox proportional hazards models, the predictive effect of the familial-genetic risk score (FGRS) for disorder B in those who had v. had not had a prior onset of disorder A. Results In all pairs, the impact of the FGRS for disorder B was significantly stronger in cases without v. with a prior history of disorder A. These effects were similar across sex, stable across levels of FGRS and not likely due to clinician bias. In many of our disorder pairs, previous clinical studies suggest a mechanism for a causal effect of disorder A on B. Conclusions Our findings provide indirect evidence that the occurrence of one psychiatric or substance use disorder often has a causal effect on risk for subsequent disorders. This mechanism may substantially contribute to the widespread comorbidity among psychiatric conditions.

psychiatry,psychology, clinical