Abstract
INTRODUCTION: Drinking at an early age (AO) and quickly progressing to drinking to intoxication (Delay=Age of Intoxication[AI]-AO) confer risk for alcohol use and alcohol-related problems. However, inconsistencies exist in the literature, which may reflect the use of different definitions of AO and AI. We evaluated whether 1) defining AO as age at first sip of alcohol (AO sip) versus age at which at least one standard drink was consumed (AO drink); and 2) defining AI as age at first "drunk" (AI drunk) versus age at first binge episode (≥5 standard drinks consumed; AI binge) resulted in different self-reported ages or differentially predicted drinking outcomes. METHODS: 248 high school students (53.6% male; 16.50[1.19] years; 71.4% White) completed anonymous surveys assessing alcohol use. RESULTS: Participants reported a younger AO (sip) than AO (drink) and a younger AI (drunk) than AI (binge), resulting in significantly different Delay values for the four AO-AI pairings. Univariate general linear models indicated that AO-Delay pairings accounted for more variance in maximum drinks and alcohol-related problems than did the individual AO and AI variables. Pairings comprising AO (drink) and Delay (drink-binge) and AO (sip) and Delay (sip-binge), respectively, uniquely accounted for variance in both maximum drinks and problems. CONCLUSIONS: Clearly defining AO and AI using objective definitions that reflect specific amounts of alcohol (e.g., first sip; first standard drink; first binge) appears to outperform subjective definitions of alcohol use (e.g., first drunk).