Selenium

What it is

Selenium is an essential trace element — a mineral the body needs in small amounts — and it is built into proteins called selenoproteins, such as glutathione peroxidase, which help protect cells against oxidative stress [1][2]. A blood level of about 125 µg/L is enough to keep selenoproteins at their usual working level; higher intakes do not raise them further [3].

Sources supply different chemical forms. Brazil nuts contain organic selenium such as selenomethionine, and supplements use selenomethionine, selenium-enriched yeast or sodium selenite [4][5][3][6]. The recommended intake for adults is about 55 µg a day [7]. Brazil nuts are the richest known food source of selenium [4].

Selenium deficiency occurs where soil is selenium-poor; it causes Keshan disease, a heart condition that selenium supplements prevent [8]. People on long-term dialysis can also run low on selenium [9].

How it might work

Proposed mechanisms: selenium is a key component of selenoproteins, enzymes that protect against oxidative stress, and blood glutathione peroxidase activity rises when selenium intake rises in trials [1][2]. In the thyroid, selenium is required for thyroid hormone synthesis, and the gland's high oxidative activity is the proposed reason selenium might affect autoimmune thyroid disease [10]. Graves' eye disease is thought to involve oxygen free radicals and cytokines, which is why an antioxidant was tested in it [11]. These are proposed mechanisms; the trials are what test whether they translate into health effects.

What the research says

Graves' eye disease. A 2024 meta-analysis of four randomised trials found selenium improved the clinical activity score (a 0-10 measure of eye disease activity) by 1.27 points more than placebo at 6 months, improved Graves' eye quality-of-life scores (about two and a half times as many people improved), and changed eyelid aperture by about 1.5 mm more than placebo; there was no clear difference in proptosis, eye muscle movement or side effects [12]. The anchor trial, a 2011 study of 159 people with mild orbital disease, found better quality of life, less eye involvement and slower progression over 6 months, with no adverse events attributed to selenium, and the benefits persisted 6 months after the supplement stopped [11]. A 2024 study with 5 years of follow-up found early improvements but no difference between groups in long-term outcomes [13]. A small 2024 trial in inactive, moderate-to-severe eye disease found improved eyelid aperture but no difference in activity score or quality of life [14].

Selenium deficiency. In selenium-poor parts of China, giving selenium prevented Keshan disease: a 2018 meta-analysis of 41 community studies covering 1,983,238 people found supplementation cut the incidence by about 86% (RR 0.14, 95% CI 0.12 to 0.16) [8].

Hashimoto's thyroiditis. A 2024 meta-analysis of 35 studies found selenium lowered thyroid peroxidase antibodies substantially (standardised mean difference -0.96) and lowered TSH slightly in people not taking levothyroxine (-0.21), with no excess of side effects versus placebo [10]. An earlier meta-analysis found antibody falls in both treated and untreated people, said their clinical relevance was not established, and found selenium users complained of side effects more often (p = 0.036) [15]. An overview of six systematic reviews also reported antibody reductions, rated as very low certainty [16]. But a 2017 meta-analysis found no effect on TSH, quality of life or thyroid ultrasound appearance [17], and a 2013 Cochrane review could not pool the four trials it found and called the evidence incomplete [6]. In short: antibodies fall, but whether people feel better is not established [10][17].

Cancer. The strongest evidence is a 2018 Cochrane review of 10 randomised trials with 27,232 participants: selenium supplements did not reduce overall cancer incidence (RR 1.01, 95% CI 0.93 to 1.10, high-certainty evidence) or any site studied — prostate (RR 1.01), lung (1.16), colorectal (0.99), bladder (1.07) or breast cancer [18]. The largest included trial, SELECT, gave 35,533 men 200 µg a day of selenomethionine and found no reduction in prostate cancer (hazard ratio 1.04 in the first analysis; 1.09 in the final analysis) [5][19]. Observational studies point the other way — high blood or dietary selenium is linked to lower cancer risk (pooled OR 0.78 for the highest exposure category; RR 0.94 for dietary intake of at least 55 µg a day) — but these studies cannot show cause and effect, no dose-response pattern was found, and one pooled analysis found no protective effect from supplement use [20][7]. In one Chinese trial, a supplement combining vitamin C, vitamin E and selenium lowered stomach cancer incidence (OR 0.64) and mortality (HR 0.48), but the combination makes selenium's own contribution unclear [21].

Heart disease. A 2013 Cochrane review of 12 trials (19,715 people) found no effect of selenium on total deaths (RR 0.97, 95% CI 0.88 to 1.08), cardiovascular deaths (0.97) or all cardiovascular events (1.03), and its authors concluded the evidence does not support selenium for preventing heart disease [1]. A 2022 analysis of 884 supplement trials found selenium had no effect on cardiovascular disease or type 2 diabetes [22]. A meta-analysis combining 16 trials and 16 cohort studies found supplements gave no cardiovascular benefit (RR 0.91, 95% CI 0.74 to 1.10), while observational studies showed a U-shaped link — higher risk at very low blood selenium and possible benefit in a narrow 55 to 145 µg/L range [23]. A US preventive-services review of single-nutrient trials found no clear evidence of benefit or harm for selenium [24]. One indirect positive note: antioxidant mixtures that included selenium were linked to lower cardiovascular mortality (RR 0.77) and all-cause mortality (RR 0.90), but selenium alone showed no such effect [25].

Type 2 diabetes. Pooled trial evidence suggests a small increase in risk. A 2018 meta-analysis of 5 randomised trials found an 11% higher risk of type 2 diabetes with selenium 200 µg a day (RR 1.11, 95% CI 1.01 to 1.22) [26]. Two earlier pooled analyses found a similar direction but did not reach statistical significance (RR 1.09, 95% CI 0.99 to 1.20; OR 1.18, 95% CI 0.95 to 1.47), and Cochrane found a small non-significant increase (RR 1.06) [27][28][1]. Observational studies show a stronger association (OR 2.03, 95% CI 1.51 to 2.72) that is likely to be partly explained by other differences between people [28].

Doses used in studies

The cancer trials used 200 µg a day of L-selenomethionine: SELECT ran for a planned 7 to 12 years [5], and the Graves' eye disease trial also used 200 µg a day (100 µg twice daily) for 6 months [11]. The Danish PRECISE trial tested 100, 200 and 300 µg a day of selenium-enriched yeast for 5 years [3]. Thyroid antibody trials commonly used 200 µg a day or less [6], and the SERENA pregnancy trial used 83 µg a day of L-selenomethionine [29]. Brazil nut trials used one to two nuts a day; in the New Zealand trial two nuts were intended to supply 100 µg but actually supplied 20 to 84 µg (average 53 µg) [2]. These are doses used in research, not recommendations; the European tolerable upper intake level is 255 µg a day [30].

Safety and interactions

Side effects. At 200 µg a day, selenium increased hair loss (RR 1.28, 95% CI 1.01 to 1.62) and mild dermatitis (RR 1.17, 95% CI 1.00 to 1.35) compared with placebo in SELECT [1][18]; hair loss is the best-documented effect of excess selenium intake [30]. In thyroid trials, side effects were broadly similar between selenium and placebo [10], though one meta-analysis found selenium users reported more adverse effects (p = 0.036) [15].

Upper limits. European authorities set a tolerable upper intake level of 255 µg a day for adults, including during pregnancy and breastfeeding, based on hair loss starting at about 330 µg a day [30]. In a Danish trial, 300 µg a day for 5 years was followed by higher all-cause mortality over the next decade (HR 1.59, 95% CI 1.02 to 2.46), and its authors advised avoiding total intakes above 300 µg a day [3].

Brazil nuts and dosing surprises. The selenium in Brazil nuts is unpredictable: two nuts bought for a trial supplied 20 to 84 µg a day instead of the expected 100 µg [2], while a Brazilian study using naturally high-selenium nuts measured about 1,261 µg in one daily portion and found increased activity of inflammatory genes at that dose [31]. People eating a selenium-rich Amazon diet had very high blood levels (median 228 µg/L) without signs of selenium excess, so how selenium is packaged in food may matter, but official advice names regular Brazil nut eaters and high-dose supplement users as the groups most likely to exceed the limit [32][30].

Interactions. Reviews of antioxidant supplements during cancer treatment report mixed effects on treatment side effects and no clear data on whether selenium changes treatment outcomes, so people having chemotherapy or radiotherapy should ask their cancer team before taking selenium [33][34]. In the thyroid trials, selenium was given alongside standard treatment, including levothyroxine, without reported interactions [10].

Pregnancy and breastfeeding. The 255 µg a day upper level applies to pregnant and breastfeeding women [30]; trials in pregnancy used much lower intakes — 60 µg a day for pre-eclampsia risk markers and 83 µg a day for thyroid antibodies — without reported harm [35][29].

Kidney disease. People on dialysis often have low selenium: in one small study a daily Brazil nut raised plasma selenium from below normal (17.3 µg/L) to normal (106.8 µg/L), and the effect faded after the nuts stopped [9][36]. Whether selenium supplements help people with kidney disease is not established [9].

Caveats and open questions

Most trial participants have been men — Cochrane's cancer review notes 88% were male — so effects in women are less certain [18]. Observational and randomised findings disagree sharply: blood selenium tracks with lower cancer and heart disease risk in cohort studies, while supplements show no effect in trials, which suggests the observational links reflect other differences between people [18][23]. The Graves' eye disease benefit comes from a handful of small trials, mostly in mild, active disease; a 5-year follow-up found no difference in long-term outcomes [12][13]. Whether different selenium forms (selenomethionine, selenite, yeast) behave differently is unresolved [18]. The mortality signal at 300 µg a day comes from a single small trial and needs replication [3]. Ongoing trials listed below are testing selenium in heart failure and in thyroid nodules.