The results in Table 3 address more about how SRE and alcohol challenge LR values overlap regarding future drinking parameters. When both T15-based SRE5 scores and T1 alcohol challenge results were entered into regressions predicting the outcomes across T20 through T30, the R2s were a modest .04 to .07, but all were significant. Alcohol-related outcomes are affected by many things, including changes in jobs and relationships as well as health concerns and medical treatment, and explaining ~5% of the variance across 15 years is noteworthy. Regarding regressions that included SRE5, alcohol challenge values significantly contributed to each outcome, while SRE5 contributed to future AUD status, with a trend (β = .14, p <.09) for alcohol problems. For regressions that included SRET, R2 values were higher (.07 to .15) and the SRE contributed to all four regressions with alcohol challenge values contributing significantly to regressions for maximum drinks, with trends for usual drinks (β = −.15, p=.054) and problems (β = −.14, p=.07). Thus, the two SRE and alcohol challenge-based LRs appear to complement each other in their ability to predict alcohol outcomes, with both measures often contributing significantly to regressions.
