Abstract
INTRODUCTION: Daily data collected over 14 consecutive days were used to examine whether extreme drinking was more likely on days college students reported prepartying (i.e., drinking before going out) or playing drinking games in a multi-ethnic sample of college seniors (analysis subsample: N=399; 57% women; M age=21.48years, SD=.40). METHODS: Multilevel modeling with drinking occasions at Level 1 (1265 drinking days) nested within persons at Level 2 (399 drinkers) was used to predict four extreme drinking behavior outcomes at the daily level: consuming at least 8/10 (women/men) drinks, reaching an estimated blood alcohol concentration (eBAC) of .16% or greater, drinking enough to stumble, and drinking enough to pass out. RESULTS: Prepartying only (29% of drinking days) was more common than playing drinking games only (10%) or engaging in both behaviors on the same day (13%). Odds of extreme drinking were greater among students who frequently engaged in prepartying (ORs: 1.86-2.58) and drinking games (ORs: 1.95-4.16), except prepartying frequency did not predict drinking enough to pass out. On days students prepartied (ORs: 1.58-2.02) and on days they played drinking games (ORs: 1.68-1.78), odds of extreme drinking were elevated, except drinking games did not predict eBAC of .16% or greater. CONCLUSIONS: Extreme drinking is attributable to both person-level characteristics (e.g., preparty frequency) and specific drinking behaviors on a given day. Prepartying and drinking games confer elevated risk of extreme drinking and are important targets in alcohol interventions for college seniors.