Compared to PRS, the association between the first-degree family history and AUD was of a greater magnitude (although 95% CI overlapped). This is expected, because the first-degree family history reflects both genetic and shared environmental effects, while PRS estimates only a portion of the genetic component of risk (as captured by common variants). However, the ORs in full sample and FH+ subsample were 1.96 and 1.86, respectively, which are comparable to the estimated ORs of the first-degree family history based on U.S. national surveys (1.91–2.38) (Dawson et al., 1992, Karriker-Jaffe et al., 2021). Therefore, like family history, PRS could be used in evaluating AUD risk. In addition, by adjusting for PRS effects, important non-genetic factors related to AUD can be identified, and interactions between these factors and PRS can also be investigated. These non-genetic factors, combined with PRS, can further improve our ability to evaluate AUD risk. In our dichotomized PRS analysis, we used 10% as the cut-off to define high PRS individuals. For comparison purpose, we also used 5% as the cut-off and results were similar with much wider 95%CI due to the smaller numbers of individuals having high PRS.
