Polychlorinated biphenyl (PCB) exposure has been associated with cancer risk. Vital status for 14,458 workers with >/= 90 days potential PCB exposure 1939-77 at two capacitor manufacturing plants was verified through 1998 to examine cancer mortality and excess liver and rectal cancer reported in a subcohort. Standardized mortality ratios (SMRs, U.S. rates) and Poisson regression rate ratios (RRs) were assayed; a job exposure matrix estimated PCB exposure. Liver and prostate cancer mortality was not elevated (liver, 21 deaths, SMR 0.9, 95% confidence interval (CI) 0.6-1.4; prostate, 34 deaths, SMR 1.0, CI 0.7-1.5), but rose with low to high cumulative exposure (liver RRs: 1.7, CI 0.3-10.3; 3.0, CI 0.6-15.4; 4.2, CI 0.9-19.8, p for trend 0.07; prostate RRs: 1.5, CI 0.4-5.6; 2.8, CI 0.8-9.6; 6.1, CI 2.0-18.2, p for trend 0.0001). Stomach cancer mortality among men (n = 24, SMR 1.5, CI 1.0-2.3) rose with cumulative exposure. Mortality from myeloma (n = 28, SMR 1.9, CI 1.2=2.7), intestinal cancer among women (n = 67, SMR 1.3, CI 1.0=1.7), and rectal cancer (n = 21, SMR 1.1, CI 0.7=1.7) were elevated but not clearly related to exposure. Ovarian cancer mortality, elevated among long term workers, rose with cumulative exposure. This study cannot associate intestinal or rectal cancer with PCB exposure but supports prior reports of elevated liver cancer mortality, is the first occupational cohort study to show a strong exposure-response relationship between PCBs and prostate cancer mortality, and suggests that myeloma, ovarian cancer, and male stomach cancer mortality may be related to PCB exposure.
American Journal of Epidemiology. Abstracts of the 2nd North American Congress of Epidemiology, June 21-24, 2006, Seattle, Washington