BACKGROUND:

Much of the medical and epidemiological research on the causes of cardiovascular disease has concentrated either on individual life style behaviors such as smoking and exercise or on biomedical risk factors such as hypertension, and diabetes. This research has provided the basis for improvements in personal health behaviors and in the treatment of hypertension which has reduced the incidence of cardiovascular disease in this country.

However, even the best statistical models, incorporating all such risk factors, leave a significant proportion of cardiovascular disease mortality and morbidity unexplained. Major differences in cardiovascular disease rates have also been observed between different social classes and different societies. These findings have led epidemiologists and sociomedical researchers to investigate other types of variables, including psychosocial factors, such as personal predispositions, social support networks, and work-related stress. A series of studies, in a variety of disciplines, suggested that occupational stress may contribute to the development of cardiovascular disease.

A major source of data was the Swedish Survey of Living Conditions (ULF). The ULF was drawn from a sample which included the total occupational and demographic variation of an entire society. The survey contained information on age, sex, marital status, occupational class background and nationality of respondents and parents, residence, education, past employment history and income data.

DESIGN NARRATIVE:

In this non-concurrent prospective study, ten years of follow-up data were linked with five years of data from the Survey of Living Conditions collected between 1976 and 1980. Total and cause-specific mortality data were obtained from the Swedish National Death Registry and morbidity incidence data from the Registry of Hospitalizations. The relationship between the duration of exposure time and disease risk was examined utilizing occupational history information which was combined with Theorell's Occupational Psychosocial Scoring System. Logistic regression analysis was used to examine the relative risks associated with adverse work characteristics and to test for potential interactions and confounding effects. Cox's proportional hazards model was used as an approximation of life table methods. Analysis was performed separately for men and women.