Abstract
Altogether 99,029 men aged 40-59, employed on the Italian railroad system, were classified in three levels of physical activity at work and three levels of job responsibility and then followed up for five years in terms of mortality and cause of death. Physical activity and job responsibility have been investigated in different ways as possible risk factors of lethal events. When considering together the findings of univariate and bivariate analyses it appears that low physical activity and high job responsibility are attributes favouring myocardial infarction, and that high physical activity and low job responsibility are attributes favouring lethal chronic bronchitis and violent death. Overall, mortality from all cases is not significantly different in different classes of physical activity and job responsibility, and within total mortality these two characteristics play a role in the distribution of the various causes of death as a consequence of possible competing risks. A number of other non-measured factors, however, may have had a confounding effect.