DSM-IV or DSM-5 criteria, hence enhancing diagnostic precision. Moreover, we excluded individuals with alcohol use problems but not meeting AUD criteria and other substance use disorders from the control group, thereby increasing statistical power. All of these approaches maximize our ability to screen for PGS with the highest estimability. Then we tested the generalizability of the best PGS in the AOU and IB cohorts, which were not ascertained to study AUD. The large sample sizes and low prevalence of AUD in both the AOU and IB cohorts ensured that PGS distributions remained unbiased and may resemble those in general populations. The ORs for the AOU and IB cohorts were 1.96 and 2.07, respectively, which were comparable to those for a family history of AUD (OR, 1.91-2.38).14,47 We note that the estimability is smaller than the PGS of some diseases; however, for preventable disorders like AUD, the estimabilities of the PGS could be totally reduced or amplified by nongenetic factors (eg, culture or religion can prevent individuals with high PGS from drinking, whereas traumatic events can increase the probability of AUD for those with low PGS). Consequently, higher PGS estimability is not expected for preventable diseases. It is noteworthy that the
