[A retrospective cohort study on long-term health effects of shift work]

A retrospective cohort study was conducted to examine the long-term effects of shift work including night work upon sick leave utilization. The cohort consisted of all blue-collar workers aged 16 to 28 years, who entered a steel production plant between 1967 and 1970, who were then divided into a shift worker group (790 persons) and a day worker group (309 persons) according to their initial baseline work pattern. A 20 year follow-up study of the health administration files of the two groups was performed in November 1991. The main results were as follows: 1. Reasons for and time trends for retirements in the shift worker group were very similar to those in the day worker group. 2. With respect to illnesses needing hospitalization or medically certified absence lasting more than one month, the cumulative incidence rates during a 5-year period and a 10-year period in the shift worker group were 17% and 28% respectively, and 16% and 28% in the day worker group. No significant effect of shift work upon occurrence of illnesses could be found after controlling for factors at entrance and history of engagement in hazardous work although there were differences between the two groups. 3. No significant differences in incidence rates of any diagnostic categories were detected between the groups. 4. Among the workers still employed in 1990, we found no differences in prevalence rates of any diagnostic categories between the groups. These results lead to the conclusion that shift work appears to have no severe adverse effects upon morbidity when there is a high level of health and labor administration.