Higher Serum Selenium Concentration is Associated with a Higher Prevalence of Esophageal Squamous Dysplasia in Bomet, Kenya

15 Background: East Africa is a region of high esophageal squamous cell carcinoma (ESCC) incidence, and is known to have low soil selenium levels. Low selenium status measured in blood has been associated with ESCC risk. We measured serum selenium in a cross-sectional study of subjects from Bomet, Kenya, a region with a high incidence of ESCC and esophageal squamous dysplasia (ESD), its precursor lesion.Methods: An age and sex-stratified sample of 294 asymptomatic adult residents completed questionnaires and underwent endoscopy with Lugol's iodine and biopsy for detection of ESD in this community-based study. Serum selenium concentrations were measured using instrumental neutron activation analysis. Odds ratios (OR) and confidence intervals (95% CI) for associations between serum selenium and ESD were calculated using unconditional logistic regression, adjusting for multiple potential confounders. Ethical approval was obtained from the National Cancer Institute, Tenwek Hospital, and Kenyatta National Hospital.Results: Higher serum selenium (quartile 4) was associated with prevalence of ESD (OR: 3.03; 95% CI:1.05, 8.74) relative to lower serum selenium (quartile 1). This association remained significant after adjusting for potential confounders (Q4 vs Q1: OR: 4.24; 95% CI: 1.16, 15.46). Serum selenium concentration did differ significantly by geographic location (p=0.001); however, this did not explain the association between selenium and ESD.Conclusions: We found a significant adverse association between higher serum selenium levels and risk of ESD. These results are in contrast to our hypothesis derived from previous observations of an inverse association between selenium and ESCC. The discrepancy may be due to the difference in endpoint of ESD rather than ESCC. Further work is needed to better understand the potential role of selenium in esophageal cancer prevention and control strategies. Funding: This work was supported by the Intramural Research Program, Division of Cancer Epidemiology & Genetics, National Cancer Institute, NIH.AUTHORS' DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST: Natalie R. Pritchett No relationship to disclose Michael M. Mwachiro No relationship to disclose Stephen L. Burgert Employment: Centers for Gastroenterology, Fort Collins, Colorado, USA Research Funding: Covidien Gwen Murphy No relationship to disclose John D. Brockman No relationship to disclose Christian C. Abnet No relationship to disclose Sanford M. Dawsey No relationship to disclose