Decomposing associations between acculturation and drinking in Mexican Americans.

BACKGROUND: Acculturation to life in the United States is a known predictor of Hispanic drinking behavior. We compare the ability of 2 theoretical models of this effect-sociocultural theory and general stress theory-to account for associations between acculturation and drinking in a sample of Mexican Americans. Limitations of previous evaluations of these theoretical models are addressed using a broader range of hypothesized cognitive mediators and a more direct measure of acculturative stress. In addition, we explore nonlinearities as possible underpinnings of attenuated acculturation effects among men.
METHODS: Respondents (N = 2,595, current drinker N = 1,351) were interviewed as part of 2 recent multistage probability samples in a study of drinking behavior among Mexican Americans in the United States. The ability of norms, drinking motives, alcohol expectancies, and acculturation stress to account for relations between acculturation and drinking outcomes (volume and heavy drinking days) were assessed with a hierarchical linear regression strategy. Nonlinear trends were assessed by modeling quadratic effects of acculturation and acculturation stress on cognitive mediators and drinking outcomes.
RESULTS: Consistent with previous findings, acculturation effects on drinking outcomes were stronger for women than men. Among women, only drinking motives explained acculturation associations with volume or heavy drinking days. Among men, acculturation was linked to increases in norms, and norms were positive predictors of drinking outcomes. However, adjusted effects of acculturation were nonexistent or trending in a negative direction, which counteracted this indirect normative influence. Acculturation stress did not explain the positive associations between acculturation and drinking.
CONCLUSIONS: Stress and alcohol outcome expectancies play little role in the positive linear association between acculturation and drinking outcomes, but drinking motives appear to at least partially account for this effect. Consistent with recent reports, these results challenge stress models of linear acculturation effects on drinking outcomes and provide (partial) support for sociocultural models. Inconsistent mediation patterns-rather than nonlinearities-represented a more plausible statistical description of why acculturation-drinking associations are weakened among men.