Se is a trace mineral that has antioxidant and anti-inflammatory properties. This study aimed to investigate the association between Se intake, diabetes, all-cause and cause-specific mortality in a representative sample of US adults. Data from 18 932 adults who attended the 2003-2014 National Health and Nutrition Examination Survey were analysed. Information on mortality was obtained from the US mortality registry updated to 2015. Multivariable logistic regression and Cox regression were used. Cross-sectionally, Se intake was positively associated with diabetes. When comparing the extreme quartiles of Se intake, the OR for diabetes was 1·44 (95 % CI 1·09, 1·89). During a mean of 6·6 years follow-up, there were 1627 deaths (312 CVD, 386 cancer). High intake of Se was associated with a lower risk of all-cause mortality. When comparing the highest with the lowest quartiles of Se intake, the hazard ratios for all-cause, CVD mortality, cancer mortality and other mortality were 0·77 (95 % CI 0·59, 1·01), 0·62 (95 % CI 0·35, 1·13), 1·42 (95 % CI 0·78, 2·58) and 0·60 (95 % CI 0·40, 0·80), respectively. The inverse association between Se intake and all-cause mortality was only found among white participants. In conclusion, Se intake was positively associated with diabetes but inversely associated with all-cause mortality. There was no interaction between Se intake and diabetes in relation to all-cause mortality.
Keywords: Adults; Diabetes; Mortality; Selenium intake; USA.