Responses to selenium supplementation in healthy Americans

To determine dose‐response relationships of selenium (Se) intake and markers of Se status in healthy, non‐deficient subjects, we conducted a 12 mo. intervention using Se supplements (0, 50, 100 or 200 ug Se [as L‐selenomethionine]/d) in a group of 106 men and 155 women. At baseline, this cohort, none of whom took supplements with >50 ug Se/day, had plasma Se of 141.5±23.7 ng/ml, urinary Se of 58.2±21.5 ng Se/mg creatinine, and buccal cell Se of 10.2±6.5 ng Se/mg protein. Over the intervention, rates of drop‐out (7%) and compliance (98%) were very good, and each marker increased in a dose‐dependent manner. The 12 mo.‐increases in plasma and urinary Se in response to effective Se dose (corrected for body mass [kg 0.75 ]) were similar: change in plasma Se (ng/ml)=18.2 × dose (ug/kg 0.75 ) + 12.6 [r 2 =0.60]; change in urine Se (ng/mg creat.)=18.5 × dose (ug/kg 0.75 ) + 17.4 [r 2 =0.44]. Men and women showed comparable plasma Se responses; however, the urinary Se response of women (20.9 ng/mg creat./ug Se/kg 0.75 ) was nearly twice that of men (11.5 ng/ugSe/kg 0.75 ) (P<0.001). Baseline plasma Se varied by glutathione peroxidase (GPX1) genotype: 198Leu/Leu, 135.7±19.0 ng/ml; 198Leu/Pro, 139.2±23.8 ng/ml; 198 Pro/Pro, 145.9±24.4 ng/ml (P<0.05). The 12‐mo. plasma Se response was not related to GPXi genotype; however, the urinary Se response of 198Leu/Leu individuals was only 63% of that of 198Pro/Pro individuals (P<0.006).