Abstract Objectives The retrospective self‐report of the effects of alcohol (SRE) questionnaire generates three scores reflecting drinking quantities required for effects: SRE‐5 (first five times of drinking), SRE‐3 (recent 3 months of drinking), and SRE‐H (heaviest drinking). Data to date indicate that SRE‐5 scores, potentially reflecting genetic effects, have high reliabilities across 9 months. Data presented here evaluated the consistencies of all three SRE scores obtained five times over a decade. Method Participants were 331 young adults (45% male) from the Collaborative Study on the Genetics of Alcoholism (COGA) Prospective Study of youth who completed the SRE and were evaluated at least five times using the Semi‐Structured Assessment for the Genetics of Alcoholism interview. Analyses documented changes in SRE scores, changes in drinking quantities, and the relationship between these measures over time using repeated measures ANOVA and regression analyses. Results SRE‐5 scores did not change significantly across evaluations despite significant changes in drinking quantities. SRE‐3 and SRE‐H demonstrated significant changes over time that related to changes in drinking quantities during each period. Participants with low levels of response (low LRs) to alcohol, based on a median split of SRE‐5 at Time 1, were likely to remain low LR throughout the decade. Conclusions This is the first report on the longer‐term consistency of all three SRE scores during a decade of life in which a person's heaviest drinking and AUD development are likely to occur. Results indicate that retrospective SRE‐5 scores, potentially reflecting genetic predispositions to AUD, do not change over a decade despite changes in drinking patterns.
Schuckit et al. (2026) studied this question.