Unemployment and mortality among Finnish men, 1981-5

Abstract

Objective - To ascertain whether, after controlling for several relevant background variables simultaneously, unemployment is related to mortality and to assess whether this relation is causal or whether unhealthy people are more likely to become unemployed. Design - Prospective study of mortality in Finland during 1981-5 based on 1980 census data on 30-54 year old wage earner men and with particular attention to unemployment in the year before the census. Setting - Research project at the University of Helsinki. Subjects - All wage earner men in Finland aged 30-54 at the 1980 census. Main outcome measures - Causes of death during 1981-5 and duration of unemployment in the year before the census. Background variables controlled for were age, socioeconomic state, marital state, and health. The data were analysed by log linear regression models. Results - During the study period 1981-5, which covered almost 2·7 million person years, there were 9810 deaths. After controlling for all background variables relative total mortality among unemployed versus employed men was 1·93 (95% confidence interval 1·82 to 2·05). The excess mortality was highest in accidental and violent causes of death (relative mortality 2·51; 95% confidence interval 2·28 to 2·76). For circulatory diseases the relative death rate was 1·54 (95% confidence interval 1·40 to 1·70), but among neoplasms only lung cancer was associated with excess mortality. Selection for unemployment based on age, socioeconomic state, and marital state was evident but no such selection was detected based on health. Effects of unemployment on mortality were more pronounced with increasing duration of unemployment. Conclusions - The relative excess mortality of unemployed men in Finland cannot fully be explained by demographic, social, and health variables preceding unemployment. Unemployment therefore seems to have an independent causal effect on male mortality. Further studies are needed to elucidate the mechanisms between unemployment and mortality.