A nested, matched case-control study was conducted to assess the relationship between serum levels of copper and the subsequent risk of cancer. One hundred thirty-three cases of cancer were identified during 1974–1984 among 5000 members of a northwest Washington State employee cohort from whom serum specimens had been previously obtained and stored. Two hundred forty-one controls were selected at random from the cohort and were matched to the cases on the basis of age, sex, race, and date of blood draw. Serum copper levels were measured by atomic absorption spectrometry. Risk of a subsequent diagnosis of cancer was positively associated with serum copper levels, but only among those cases diagnosed within 4 years of the time the serum specimens were collected. Among cases diagnosed more than 4 years after specimen collection, there was no consistent association between serum copper levels and risk. Adjustment for age, sex, race, occupational status, cigarette smoking, family history of cancer, alcohol consumption, and, among females, use of exogenous hormones had no appreciable effect on these relationships. The findings suggest that the presence of cancer may increase serum copper levels several years prior to its diagnosis. They are less supportive of the hypothesis that serum copper levels affect cancer risk.
Supported in part by NIH Grants 1 R03 CA 39135-01, 1 P50 CA 34837, and AM-35816 to the Fred Hutchinson Cancer Research Center, Cancer Prevention Research Unit, and to the University of Washington Clinical Nutrition Research Unit.