Three hundred and ninety one male employees aged 35-65 in a Swedish pulp and paper company were followed up for 22 years; 151 deaths were recorded by 31 December 1983. On the basis of data from 1961, indices for job decision latitude, job support, and other work related psychosocial factors were constructed as were five indices for non-work related social network factors. All indices were checked by life table analysis in respect of mortality. Job decision latitude and a combined index for job decision latitude and job support showed significant associations with mortality. These two indices were investigated by multivariate analysis with scale for evaluation of neuroticism and known somatic risk factors such as smoking, cholesterol, and systolic and diastolic blood pressure. Age, educational level, occupational status, physically heavy work, and general health state were also included in the multivariate analysis. Age, systolic blood pressure, the combined index for job decision latitude and job support, smoking, and neuroticism were shown to be independent predictors of mortality.Work related psychosocial factors have shown an association with mental symptoms, psychosomatic disorders, and cardiovascular morbidity in cross sectional studies." Case-control studies have also shown an association between psychological work demands and lack of influence over the working situation on the one hand and cardiovascular cerebrovascular or cardiovascular mortality on the other.'Prospective studies have shown that hectic and psychologically demanding work, low decision latitude, and combinations of these factors are predictors of mental strain and cardiovascular morbidity and mortality.2"8In social epidemiology there is a growing interest in various aspects of social network and social support as determinants of health, including premature death."'5The theoretical basis for the action of work related psychosocial factors on health has its source in stress research dealing with principally adrenal hormones as mediators of ill health.' 1618 Karasek has hypothesised that the influence ofwork demands on health may be moderated by the degree of control the individual has over his work.2 This is called Accepted 18 April 1988 the job strain model. A combination of high work demands and low level of permitted discretion in the control ofones own work carries the highest risk for ill health.Johnson and Theorell have established the term isostrain to measure the combined exposure to psychological work demands, the level for work control, and the level for work related social support.3 The isostrain model was hypothesised to be the best predictor of cardiovascular morbidity and was shown to have the strongest cross sectional association with cardiovascular morbidity.The theory for social network and social support research is the "theory of general susceptibility," presented by Cassel in 1976'9 It emphasises a nonspecific influence of factors in the social environment on the resistance of man against different pathogen...