OBJECTIVE: To explore the relationship between serum selenium and cervical cancer. METHODS: We conducted a case-control study of cervical cancer in five areas around Birmingham, AL; Chicago, IL; Denver, CO; Miami, FL; and Philadelphia, PA. Community controls were selected by random-digit dialing and were matched to invasive cervical cancer cases by age, race/ethnicity, and telephone exchange. Serum selenium was determined by neutron activation analysis. Logistic regression analysis controlling for known risk factors of cervical cancer, including human papillomavirus (HPV) type-16 measured serologically, was performed on 227 invasive cases, 127 in-situ cases, and 526 controls. RESULTS: Values of serum selenium ranged from 67.5 to 185.0 ng/ml. Adjusted odds ratios for invasive cervical cancer by quintile were: 1.0 (highest selenium), 1.1, 1.0, 0.8, and 1.0 (lowest selenium), p for trend = 0.82. Similar patterns were observed for Stage I invasive, and Stages II-IV invasive cases, suggesting severity of disease did not influence the null results. Although no associations were seen among current or never smokers, a protective effect of selenium was suggested among former smokers. Effect modification was not evident for other variables examined. CONCLUSIONS: This study does not support a relationship between serum selenium and invasive cervical cancer at typical serum selenium levels in the US.
OBJECTIVE: To explore the relationship between serum selenium and cervical cancer. METHODS: We conducted a case-control study of cervical cancer in five areas around Birmingham, AL; Chicago, IL; Denver, CO; Miami, FL; and Philadelphia, PA. Community controls were selected by random-digit dialing and were matched to invasive cervical cancer cases by age, race/ethnicity, and telephone exchange. Serum selenium was determined by neutron activation analysis. Logistic regression analysis controlling for known risk factors of cervical cancer, including human papillomavirus (HPV) type-16 measured serologically, was performed on 227 invasive cases, 127 in-situ cases, and 526 controls. RESULTS: Values of serum selenium ranged from 67.5 to 185.0 ng/ml. Adjusted odds ratios for invasive cervical cancer by quintile were: 1.0 (highest selenium), 1.1, 1.0, 0.8, and 1.0 (lowest selenium), p for trend = 0.82. Similar patterns were observed for Stage I invasive, and Stages II-IV invasive cases, suggesting severity of disease did not influence the null results. Although no associations were seen among current or never smokers, a protective effect of selenium was suggested among former smokers. Effect modification was not evident for other variables examined. CONCLUSIONS: This study does not support a relationship between serum selenium and invasive cervical cancer at typical serum selenium levels in the US.