Article in English.
Abstract on Pubmed: The role of social modelling and structural factors of the work-place in predicting the probability of heavy drinking was investigated in a sample of 3267 Norwegian male and female waiters and cooks. In the logistic regression analysis, the probability of heavy drinking was increased by two social modelling factors and one structural factor. Having co-workers who, at least weekly, took an end-of-work drink at the work-place gave an odds ratio for heavy drinking of 2.8 (95% CI 1.9-4.1), and having co-workers who went out after work at least every week gave an odds ratio of 1.8 (95% CI 1.2-2.8). Working at a place with a liberal alcohol policy gave an odds ratio 1.5 (95% CI 1.1-2.2). Among the background factors, only household type significantly predicted heavy drinking. As compared with living with children, the odds for heavy drinking when living alone was 4.3 (95% CI 2.9-6.4). The results indicate that preventive measures in the restaurant business should not only concentrate on the individual, but also deal with factors related to the occupational activity that promote and sustain heavy drinking.