Prospective study of toenail selenium levels and cancer among women

Miriam Garland, J. Steven Morris, Meir J. Stampfer, Graham A. Colditz, Victoria L. Spate, Connie K. Baskett, Bernard Rosner, Frank E. Speizer, Walter C. Willett, David J. Hunter

Research output: Contribution to journalArticlepeer-review

188 Scopus citations


Background: Inverse associations between selenium status and cancer risk have been observed in animal studies, ecologic studies, and some case-control and prospective studies. Whereas results of some prospective studies have suggested an overall inverse relationship between selenium levels and cancer, other prospective studies have failed to confirm this finding. Prospective data on women are particularly limited because fewer women than men have been studied prospectively. Purpose: The aim of this study was to prospectively examine the relationship between selenium levels in toenails (previously shown to reflect selenium in take and incidence of cancer among women. Methods: The Nurses'Health Study cohort began in 1976 with 121 700 female nurses aged 30-55 years living in 11 U.S. states. In 1982, we requested toenail clippings from the members of the cohort, and 62 641 participants with no history of cancer returned these clippings. During 41 months of follow-up, 503 cases of cancer other than breast cancer (results previously reported) or nonmelanoma skin cancer were analyzed. For each case patient, a control subject was chosen from women who remained free of diagnosed cancer, matched by age and by date of nail return. Results: No inverse association was observed between selenium levels in toenails and cancer risk. The age- and smoking-adjusted relative risk (RR), comparing the highest with the lowest quintile of toenail selenium level, was 1.44 (95% confidence interval (CI) = 0.97-2.13), and the trend across quintiles was marginally significant (two-sided P =.06). Comparing the highest with the lowest decile, the RR (age- and smoking-adjusted) was 1.77 (95% CI = 1.04-3.02). When these data were combined with the data from 434 breast cancer case patients and their matched control subjects identified in parallel from this same cohort, the RR comparing the highest with the lowest quintile was 1.24 (95% CI = 0.93-1.65). Toenail selenium level was not inversely associated with cancer at any major site, including uterine cancer, colorectal cancer, melanoma, ovarian cancer, or lung cancer (after adjusting for smoking); in fact, nonsignificant positive associations were observed at several sites. Conclusions: Toenail selenium levels were not inversely associated with cancer risk in this study. Implications: These data, in conjunction with previous findings of no association between toenail selenium status and breast cancer risk, strongly suggest that higher selenium intake within the range consumed by most U.S. women (as reflected by toenail selenium levels) is not protective against overall cancer incidence in women. (J Natl Cancer Inst87: 497-505, 1995).

Original languageEnglish
Pages (from-to)497-505
Number of pages9
JournalJournal of the National Cancer Institute
Issue number7
StatePublished - Apr 5 1995


Dive into the research topics of 'Prospective study of toenail selenium levels and cancer among women'. Together they form a unique fingerprint.

Cite this