Time trends in mortality in forestry and construction workers in Finland 1970-85 and impact of adjustment for socioeconomic variables.

STUDY OBJECTIVE This study aimed firstly to describe the development of cause-specific mortality in forestry workers, farmer/forestry workers, and skilled and semiskilled construction workers between 1970 and 1985 in Finland, and to compare this with mortality in the total working male population. The second aim was to evaluate how well the cause-specific mortality differences between the occupations could be explained by differences in socioeconomic status, marital status, or in the region of residence. DESIGN AND SETTING This is a follow up study based on the 1970, 1975, and 1980 census records in Finland linked with all death certificates for 1971-75, 1976-80, and 1981-85 respectively. Log-linear regression analysis was used. SUBJECTS All economically active men in Finland aged between 35 and 64 years in 1971-85 were studied. The number of person-years in the period totals about 10 million. MAIN RESULTS Semiskilled construction workers had the highest mortality rate almost independent of the cause of death. The mortality of forestry workers was the second highest. Compared with the reference population, however, the differences were small with regard to neoplasms and cardiovascular diseases. With regard to other diseases, only the mortality of semiskilled workers was fairly high. Differences in mortality as a result of accidents were the highest. Both suicide and accidental death rates were high in semi-skilled construction workers and forestry workers. During the study total mortality fell by about 30% but mortality differences between groups did not decline. CONCLUSIONS The high mortality of forestry and semiskilled construction workers is partly explained by socioeconomic factors such as marital status and housing conditions. These factors do not, however, explain the high suicide and accident mortality rates of forestry workers or semiskilled construction workers. More research is needed to explain these findings.