Is there a healthy worker effect for cancer incidence among women in Sweden?

Background Our aim was to evaluate whether there is a healthy worker effect (HWE) for cancer incidence among women. HWE is a bias found in occupationa l studies that compare rates of disease among employed people to disease ra tes for the general population, which includes unemployed people (who may b e less healthy than those who are employed). Methods Data from the 1960 and 1970 Swedish censuses were used to identify all 1,659, 940 Swedish women who were employed in either year: They were fo llowed during 1971-1989 through linkages to the national cancer and death r egisters. Standardized incidence ratios (SIRs) were computed comparing empl oyed women to the 1,627,873 women who were not employed in either 1960 or 1 970. Results For the 545,857 women employed in both 1960 and 1970, the SIR for a ll cancers combined was 1.05 (1.04-1.06). When specific cancer sites were a nalyzed separately, the highest cancer risks were for cancers of the lung a nd bladder (SIR = 1.2) and reproductive organs (breast, ovary, endometrium, and cervix SIR = 1.1). Overall cancer risks were highest among full-time w orkers, younger workers, urban workers, and workers with the highest socioe conomic status (based on the woman's job Title). Conclusions These results show no general HWE for cancer incidence among em ployed Swedish women. Am. J. Ind. Med. 36:193-199, 1999. Published 1999 Wil ey-Liss, Inc.dagger