The importance of employment status in occupational cohort mortality studies.

Person-years at risk in occupational cohort mortality studies may be defined as "active" (when a person is working) or "inactive" (after a person has left employment at the plant under study). To investigate the effects of employment status (active/inactive) both across studies and within them, we have analyzed ten large cohort studies conducted by the National Institute for Occupational Safety and Health in which no occupational risk had been observed. These ten data sets included 89,376 workers, 1,984,505 person-years, and 18,840 deaths. In these ten studies, the SMR for all causes was positively correlated with the percentage of inactive person-years in the study (r = 0.57, p = 0.08). Considering only inactive person-years, the all-causes SMR was 1.12 (approximately 1.25 before age 65, dropping to 1.00 after age 65). Stratification of inactive person-years by time-since-last-employment showed markedly increased mortality during the first year following employment. The all-causes SMR during active person-years was 0.40 and was fairly constant across age categories. With active and inactive person-years combined, a strong negative trend in SMRs with duration of employment was observed for all causes and for heart disease. These trends were not apparent when person-years were stratified by employment status. These results indicate that investigators should evaluate the effects of employment status when comparing SMRs between multiple cohorts or when interpreting trends in rate ratios within cohorts.