LAUSR

OBJECTIVE To better quantify the impact of specific on- and off-premise drinking contexts on population-level alcohol-related problems, we evaluated context-specific risks relative to frequency of use of each context. METHOD We surveyed 860 adult (21-100 years) past-year drinkers in the California East Bay, sampled in areas of high versus low median household income and off-premise alcohol outlet densities. We examined associations of context-specific drinking frequencies in seven on- and off-premise drinking locations with individual and area characteristics using negative binomial regression. Next, we used heteroscedastic ordered logistic regression to relate context-specific drinking frequencies and continued volumes to five drinking-related problems (Alcohol Use Disorders Identification Test scores, physiological problems, risky sex, social problems, and driving after drinking too much). To estimate population-level effects, we assessed drinking frequencies relative to mean past-year use of each drinking context. RESULTS Higher individual annual income (>$60,000) was associated with more frequent drinking in all on-premise drinking contexts (bars/clubs, restaurants, and stadiums). Heavier overall drinking was associated with drinking more frequently at bars. Drinking more frequently in respondents' own homes and heavier drinking at friends'/relatives' homes were associated with most drinking-related problems. The population-level effects of physiological problems and driving after drinking too much were highest for parties and friends'/relatives' homes, whereas that of risky sex outcomes was highest for bars. CONCLUSIONS Assessing context-specific risks related to heavy and/or frequent drinking, in combination with scaling these risks to determine population-level impacts, can help tailor interventions to reduce alcohol-related problems across different on- and off-premise contexts.