A signal detection approach to the combined effects of work stressors on alcohol consumption

Objective: To test the relationship between work stress and alcohol consumption and the efficacy of signal detection analysis in a work stress-alcohol consumption study. Method.' The study was part of a Work Site Health Project, which included a stress reduction program, at a manufacturing company in Osaka, Japan. The study design was a cross-sectional survey. The participants were Japanese male white-collar workers (N = 457; 30 to 50 years old) in the head administrative office of a large manufacturing company. Measurements consisted of demographic variables of the workers (age, marital status, manner of living and Type A personality), work stress measures, working hours and alcohol consumption. Results: Of eight independent variables, six significant predictors of heavy alcohol consumption were identified in the first signal detection analysis. Of these six variables, job demand and skill discretion were work-stress-related measures. In the second multiple regression analysis, skill discretion and job demand were significant predictors of alcohol consumption among the workers in two of the seven subgroups categorized by the first signal detection analysis. Conclusions: It was revealed that work stress did relate to alcohol consumption. However, the effect of the work stress upon alcohol consumption was limited to specific types of workers with multiple specific attributes. It was also inferred that the signal detection approach is an effective approach to the work stress-alcohol consumption relationship.