ABSTRACT: Selenium is needed by all living cells in order to ensure the optimal function of several enzyme systems. However, the selenium content in the soil in Europe is generally low. Previous reports indicate that a dietary supplement of selenium could reduce cardiovascular disease but mainly in populations in low selenium areas. The objective of this secondary analysis of a previous randomised double-blind placebo-controlled trial from our group was to determine whether the effects on cardiovascular mortality of supplementation with a fixed dose of selenium and coenzyme Q10 combined during a four-year intervention were dependent on the basal level of selenium.In 668 healthy elderly individuals from a municipality in Sweden, serum selenium concentration was measured. Of these, 219 individuals received daily supplementation with selenium (200 ?g Se as selenized yeast) and coenzyme Q10 (200 mg) combined for four years. The remaining participants (n = 449) received either placebo (n = 222) or no treatment (n = 227). All cardiovascular mortality was registered. No participant was lost during a median follow-up of 5.2 years. Based on death certificates and autopsy results, all mortality was registered.The mean serum selenium concentration among participants at baseline was low, 67.1 ?g/L. Based on the distribution of selenium concentration at baseline, the supplemented group was divided into three groups; <65 ?g/L, 65-85 ?g/L, and >85 ?g/L (45 and 90 percentiles) and the remaining participants were distributed accordingly. Among the non-treated participants, lower cardiovascular mortality was found in the high selenium group as compared with the low selenium group (13.0% vs. 24.1%; P = 0.04). In the group with the lowest selenium basal concentration, those receiving placebo or no supplementation had a mortality of 24.1%, while mortality was 12.1% in the group receiving the active substance, which was an absolute risk reduction of 12%. In the middle selenium concentration group a mortality of 14.0% in the non-treated group, and 6.0% in the actively treated group could be demonstrated; thus, there was an absolute risk reduction of 8.0%. In the group with a serum concentration of >85 ?g/L, a cardiovascular mortality of 17.5% in the non-treated group, and 13.0% in the actively treated group was observed. No significant risk reduction by supplementation could thus be found in this group.In this evaluation of healthy elderly Swedish municipality members, two important results could be reported. Firstly, a low mean serum selenium concentration, 67 ?g/L, was found among the participants, and the cardiovascular mortality was higher in the subgroup with the lower selenium concentrations <65 ?g/L in comparison with those having a selenium concentration >85 ?g/L. Secondly, supplementation was cardio-protective in those with a low selenium concentration, ?85 at inclusion. In those with serum selenium>85 ?g/L and no apparent deficiency, there was no effect of supplementation. This is a small study, but it presents interesting data, and more research on the impact of lower selenium intake than recommended is therefore warranted.Clinicaltrials.gov NCT01443780.