Selenium

A locus at 5q14, near BHMT, was the first genome-wide hit for blood selenium. A follow-up meta-analysis of toenail selenium — which reflects longer-term exposure than a blood draw — in cohorts totalling 4,162 people, replicated that locus and found a second, at 21q22.3, in CBS. Both genes sit in homocysteine metabolism, which is itself the kind of mechanistic thread that makes a statistical association more credible.

So the genetics of selenium levels is real and replicated. What it is not is a reason to raise your level with a supplement. Selenium had a plausible cancer-prevention story behind it — antioxidant enzymes need it, and lower selenium had tracked with more cancer in observational data — so the Selenium and Vitamin E Cancer Prevention Trial (SELECT) tested it directly: 35,533 men randomised to selenium, vitamin E, both, or placebo, followed for a median of 5.46 years. Hazard ratio for prostate cancer: 1.04 (99% CI 0.87–1.24) for selenium alone, 1.05 (0.88–1.25) for the combination — no benefit. There was a non-significant increase in type 2 diabetes in the selenium group (relative risk 1.07, 99% CI 0.94–1.22) and in prostate cancer itself in the vitamin E group (P = .06). The trial concluded selenium and vitamin E, at the doses tested, did not prevent prostate cancer.

Selenium also has a narrower margin between enough and too much than most nutrients sold as supplements — chronic overexposure causes selenosis, with hair and nail brittleness among its symptoms. A genotype that nudges a blood level is not a reason to push past a level that already sits in a normal range.