Selenium and prostate cancer: the SELECT trial long-term follow-up

The Selenium and Vitamin E Cancer Prevention Trial (SELECT) was the largest US-funded cancer chemoprevention trial of the 2000s and was designed to confirm a prostate cancer benefit from selenium supplementation that had been suggested by smaller secondary endpoints. The trial was stopped early in 2008. Sixteen years of follow-up later, the story is a near-textbook case of why surrogate signals do not equal endpoints — and why supplementing a nutrient above sufficiency is rarely benign.

The hypothesis that drove SELECT

The Nutritional Prevention of Cancer trial published in 1996 found a 63% reduction in prostate cancer incidence among men randomised to 200 µg/day selenium (as selenised yeast) versus placebo, as a secondary endpoint (PMID: 8971064).¹ The signal was strong enough to drive supplementation campaigns through the late 1990s and to motivate the NCI to fund SELECT, a 35,000-man trial of 200 µg L-selenomethionine, 400 IU vitamin E, both, or placebo with a planned 7–12 year follow-up.

What SELECT actually found

SELECT was stopped early in October 2008 for futility — no protection against prostate cancer — and for a non-significant safety signal of increased diabetes in the selenium-only arm (PMID: 19066370).² The follow-up analysis at a median 5.5 years post-randomisation found a 17% increase in prostate cancer incidence in the vitamin E arm and no effect in the selenium arm overall (PMID: 22039753).³ A nested analysis stratifying by baseline selenium status delivered the most consequential finding: men in the highest baseline selenium quartile (toenail selenium above approximately 1.4 µg/g) who took selenium supplements had a 91% higher risk of high-grade prostate cancer than placebo (PMID: 24563517).⁴

Extended follow-up changed the picture

Long-term passive follow-up through 2024 confirmed and amplified the pattern. The decade-plus analysis published in 2023 reported that the vitamin E excess risk of prostate cancer persisted (HR 1.14), and the selenium-by-baseline-status interaction held: men with high baseline selenium who took selenium had elevated long-term cancer risk that did not attenuate over time (PMID: 37856029).⁵ The data argue that selenium supplementation in a selenium-replete US population was actively harmful for a meaningful subgroup.

Why the early NPC trial signal was misleading

The original 1996 trial enrolled patients in low-selenium regions of the southeastern US — a population genuinely deficient by toenail and serum selenium standards. SELECT enrolled across the US with a mean baseline serum selenium of approximately 135 ng/mL, well above the threshold (around 120 ng/mL) at which selenoprotein P expression plateaus. Adding more selenium to a population already past biological saturation produced no benefit and unmasked a U-shaped harm curve. This pattern — benefit in deficiency, harm in sufficiency — recurs across nutrient trials (beta-carotene, vitamin E, folate) and is the most reliable mechanism by which a "healthy" supplement becomes a hazard.

What current guidelines now say

The American Urological Association and the European Association of Urology no longer recommend any selenium or vitamin E supplementation for prostate cancer prevention. The US Preventive Services Task Force has explicitly recommended against beta-carotene and vitamin E supplementation for cardiovascular or cancer prevention, citing SELECT and CARET (PMID: 35420709).⁶ The European Food Safety Authority in 2023 lowered its tolerable upper intake level for selenium to 255 µg/day in adults, partly citing the SELECT subgroup data (PMID: 36710901).⁷

What this means for the supplement aisle

An adult with a normal mixed diet in any high-income country is almost certainly selenium-sufficient. Adding a 200 µg selenium tablet to an already-adequate intake offers no plausible benefit and carries a documented risk for a meaningful minority. The legitimate indications for selenium supplementation in 2026 are narrow: confirmed deficiency (uncommon outside specific geographies and parenteral nutrition), some autoimmune thyroid protocols at 100–200 µg under specialist supervision, and Keshan-disease endemic regions. A "longevity stack" that includes 200 µg selenium is borrowing from a clearly negative trial and should be reconsidered.