The analyses focused on 4 hypotheses. The first predicted that for the 373 SDPS male probands who had data from alcohol challenges and four retrospective SRE scores, all five LR measures would correlate cross-sectionally and/or prospectively with a person’s maximum drinks or number of alcohol problems. This hypothesis was supported by multiple aspects of the data beginning with the zero-order correlations in Table 1 where alcohol challenge results and every LR measure correlated significantly with at least one of the alcohol-related variables at Time 1 and/or Time 2. In Table 5, after considering covariates all five proband LR values correlated significantly with cross-sectional maximum drinks and all, but SRE-5, correlated significantly with this outcome prospectively. All LR scores other than SRE-5 and the alcohol challenge correlated cross-sectionally with the number of AUD criteria endorsed, but only the alcohol challenge and SRE-3 related significantly to this outcome prospectively. The results of the regression analyses in Table 6 continued to support the most robust and most consistent relationship of SRE-3 to the alcohol variables both cross-sectionally and prospectively, but depending on the
