Abstract
Kidney cancer has been a suspected occupational disease in petroleum workers. Health conditions that are linked to kidney cancer may prompt termination or change of work and thereby restrict occupational exposures in high-risk individuals, creating a healthy worker survivor bias (HWSB). We examined associations between occupational exposures and kidney cancer among males in the Norwegian offshore petroleum workers cohort using a case-cohort design, with 169 incident cancers identified by linkage to national registry data (1999-2021) and a subcohort of 2090 noncases, all employed between 1965 and 1998. Relative risks (hazard ratios [HRs]) by cumulative exposure to benzene, crude oil, chlorinated degreasing agents (CDA), asbestos, welding fumes, or surface treatment (priming, painting) were estimated by weighted Cox regression. Inverse exposure-response trends suggested HWSB, reinforced by analyses of necessary components of HWSB. Bias was partly alleviated by adjustment for total employment duration and by 20-year lagging of cumulative exposure to benzene, crude oil, or CDA. Workers in surface treatment (ever vs never) showed increased HR = 2.22, 95% confidence interval [CI], 1.04-4.72 (9 cases, only). For asbestos and welding fumes, the initial inverse trends largely remained after adjustment. In sum, we could neither confirm nor exclude an occupational impact on kidney cancer.