Supplements to Avoid (or Use with Caution)

A handful of popular supplements don't just fail to help — at doses people routinely take, large studies found more deaths, more cancers, more fractures or more kidney stones. The pattern repeats: a pill looked good on observational data or a neat mechanism, then the randomized trials turned up harm.

Get vitamins E, A and C and beta-carotene from food rather than high-dose pills, don't take iron unless a blood test shows you need it, and skip red yeast rice. Keep total vitamin B6, food plus pills, under 12 mg a day and long-term zinc under 40 mg a day, and if you take calcium, supplement only the gap to 1,000–1,200 mg a day, no more than 500 mg at a time.

Why antioxidant pills backfire

Pooling 78 randomized trials in almost 300,000 people, the antioxidant supplements studied did not prevent death — and in the low-bias trials, beta-carotene, vitamin E and higher-dose vitamin A were linked to slightly higher mortality.[1]

The reactive oxygen species these pills mop up are not just damage but signals the body uses to switch on its own defences, so flooding the system with isolated scavengers blunts those adaptations. Antioxidants in fruit, vegetables, olive oil and tea don't carry this signal; the doses are far lower.

Exercise is the clearest case: the burst of reactive oxygen during a hard session helps drive the fitness gains. In one trial, 1,000 mg vitamin C plus 400 IU vitamin E a day abolished the training-induced gain in insulin sensitivity and stopped the body ramping up its own antioxidant enzymes.[2] But that trial ran four weeks in 39 healthy young men, and the insulin result has not replicated: in a 12-week trial of 21 men on half the vitamin C dose plus the same vitamin E, insulin-stimulated glucose uptake rose about as much with the vitamins (17.2%) as without (18.9%), even though blood vitamin levels rose.[3]

The case is thin, but caution costs nothing: don't take high-dose vitamin C, E or other concentrated antioxidants in the hours around training, especially if fitness or metabolic health is the goal. Antioxidant-rich food is fine.

Vitamin E and beta-carotene: the clearest harms

Vitamin E at 400 IU a day or more has one of medicine's clearest harm records, repeated across several trials. Keep supplemental vitamin E to about 30 IU — nuts, seeds and olive oil supply enough.

  • More deaths: a 2005 meta-analysis found about 39 extra deaths per 10,000 people at high doses.[4]
  • More heart failure hospitalizations, in a trial of people with vascular disease or diabetes.[5]
  • More prostate cancer: about 17% more in men taking vitamin E in a large prevention trial.[6]
  • A stroke trade-off: across nine trials at a range of doses, bleeding strokes were about 22% more common (only just statistically significant) while clot strokes were about 10% less common, and total stroke was unchanged where reported. The authors still advise against indiscriminate use, because bleeding strokes do more damage.[7]

Beta-carotene: avoid it entirely if you smoke or ever have. In a trial of male smokers it raised lung cancer by 18%,[8] and a second trial in smokers was stopped early after a 28% rise.[9] The US Preventive Services Task Force recommends against vitamin E or beta-carotene to prevent heart disease or cancer, its firmest "don't".[10] Carotenoids from orange and leafy vegetables carry no such risk.

Vitamins A and C: fine from food, risky as megadoses

Preformed vitamin A (retinol, from liver, supplements and multivitamins) is the concern, not the beta-carotene in vegetables, which the body converts only as needed. The usual ceiling is about 3,000 µg a day (counted as retinol activity equivalents), set on liver scarring and harm to a developing fetus. Pregnant women should avoid retinol-heavy multivitamins.

A hip-fracture signal appears lower than that. In a study following 72,337 postmenopausal women for 18 years, those taking in at least 2,000 µg of retinol a day had about 90% higher hip-fracture risk than those under 500 µg. Beta-carotene alone was not linked to fractures, and the retinol link was weaker in women taking postmenopausal estrogen.[11] Get vitamin A from vegetables, with liver, dairy and eggs in moderation.

Vitamin C is benign at normal intakes; the problem is the gram-plus "immune megadose". The adult upper limit is 2,000 mg a day.

  • Kidney stones — plausible, not settled. High doses can raise oxalate in urine, the main ingredient of most stones. The US National Institutes of Health calls the studies conflicting and places the best evidence in people who already have persistently high urinary oxalate, and in people with kidney disorders.[12] In one large cohort of Swedish men, those taking vitamin C tablets — typically 1,000 mg — were nearly twice as likely to get a first kidney stone, while multivitamin users were not.[13] The same link has not been found in women.
  • Iron overload. In hereditary hemochromatosis, chronic high-dose vitamin C could worsen iron loading and damage tissue; in healthy people, extra iron absorption isn't a concern.[14] The scarier "pro-oxidant" story — vitamin C feeding iron chemistry that makes the most damaging free radical — is a test-tube result with no convincing evidence it happens in people.[15]

Megadosing has no benefit; citrus, peppers, kiwi and broccoli cover your needs.

Calcium and iron: fill a real gap, nothing more

Calcium supplements above about 1,000 mg a day in people who aren't deficient carry a probable heart risk signal — probable, because the evidence points both ways. One review found no link between heart disease and total calcium from food and supplements up to the 2,000–2,500 mg a day upper limit — a ceiling on total intake, not a supplement allowance — but it was funded by an osteoporosis foundation and heart events were only a secondary endpoint in its trials.[16]

On the other side, calcium given without vitamin D was linked to about 31% more heart attacks: 143 against 111 among 8,151 adults over a median 3.6 years, with no significant rise in stroke or death.[17] Most shelf products add vitamin D. Calcium plus vitamin D, taken as 500 mg twice a day, did bring about 17% more self-reported kidney stones in a seven-year trial of 36,282 postmenopausal women — roughly 2.5% against 2.1%.[18] Food calcium shows neither pattern, probably because a single large pill spikes blood calcium for hours and food never does.

So: get calcium from dairy, leafy greens, sardines or calcium-set tofu; supplement only the gap to 1,000–1,200 mg a day total; keep each dose to 500 mg or less, taken with a meal, where carbonate absorbs as well as citrate; choose citrate, which needs no stomach acid, if you take acid-reducers or dose between meals. Postmenopausal women at risk of osteoporosis should work it out with a clinician. See Calcium.

Iron: don't supplement without a documented deficiency. Most multivitamins for men and older adults are iron-free on purpose. Iron fuels oxidative stress, and high stored iron is linked to disease: people in the highest band of ferritin (the blood marker of iron stores) had about 70% more risk of type 2 diabetes than the lowest, about 63% after accounting for inflammation, in pooled cohorts.[19] High stored iron is also linked to coronary artery disease.

About 1 in 150 people of European descent carry two copies of the main gene variant for hereditary hemochromatosis, an inherited iron-overload disorder. Not all get the disease — over about seven years it had been diagnosed in 21.7% of such men and 9.8% of such women — but 15.7% of those men developed a related condition, against 5.0% of men without the variant, and they were about four times as likely to have liver disease.[20]

The exception is premenopausal women with low ferritin: supplement to a ferritin above 30 µg/L (some clinicians aim for 40–70 µg/L for hair regrowth or athletic performance). Take it on alternate days: in iron-depleted women given 60 mg of ferrous sulfate, alternate-day dosing raised the share absorbed from 16.3% to 21.8%, because each dose leaves the iron-blocking hormone hepcidin raised for the next day.[21] That trial measured absorption, not anaemia recovery. Ferrous bisglycinate is a reasonable choice if sulfate upsets your stomach, but it was not the form tested.

Zinc and selenium: small doses fine, big doses not

Zinc above 40 mg a day for long periods depletes copper. That is the US upper limit;[22] the European Food Safety Authority sets a stricter 25 mg a day.[23]

Men have a specific reason for care. In a study of 46,974 men followed for 14 years, those taking more than 100 mg a day of supplemental zinc had more than double the rate of advanced prostate cancer — a result that only just clears chance and rests on 10 cases. Taking zinc for 10 years or more, at any dose, carried a separate signal of similar size. Total prostate cancer was not raised, and the authors could not rule out confounding.[24]

Daily, 8–15 mg matches the recommended intake. For a cold you already have, zinc lozenges may shorten it by about two days, but that benefit is low-certainty, and zinc used this way probably increases non-serious side effects.[25] The trials that showed anything used 75 mg a day or more for one to two weeks.

Selenium at 200 µg a day or more in people who already get enough is not protective. In a trial of 35,533 men, 200 µg a day did not prevent prostate cancer, which was marginally and not significantly more common than on placebo,[26] and a later analysis suggests it may have harmed men who were already selenium-replete.[27] In another trial, in a low-selenium part of the eastern US, 200 µg a day — double the dose suggested below — brought more type 2 diabetes than placebo over 7.7 years, concentrated in those who started with the highest blood selenium.[28]

People in low-selenium regions, such as much of inland Europe, may benefit from a modest 50–100 µg a day. Stay below the European upper limit of 255 µg a day, which is set on hair loss, the earliest sign of excess.[29]

Vitamin B6 and niacin: vitamins at drug doses

Vitamin B6 in large doses damages nerves. Long-term excess causes tingling and numbness in the hands and feet. It usually stops if you quit as soon as symptoms appear,[30] but not always: in Australia's 2025 review, 103 people reported severe and sometimes permanent effects, including nerve damage and muscle weakness.[31]

Use the most conservative limit. The European upper limit is 12 mg a day from food and supplements combined.[32] Australia, whose own limit is 50 mg a day, has required a nerve-damage warning above 10 mg since 2022. In November 2025 it concluded that damage can occur at 50 mg a day or lower and decided that, from 1 June 2027, products above 50 mg a day will need a pharmacist's advice to buy.[33] The US limit of 100 mg a day is an outlier, not a target.[34] Many "stress" and "energy" formulas contain 50–100 mg or more, so check the label, and take B6 megadoses only to treat a deficiency under medical supervision. See B vitamins.

Niacin at 1 g a day or more lowers LDL ("bad") and raises HDL ("good") cholesterol, but two large trials found no reduction in heart attacks or strokes when it was added to a statin.[35][36] The larger, with 25,673 people, measured the harm: serious infections, new diabetes, and serious gut, muscle-and-bone, bleeding and skin problems were all more common on niacin, and in people who already had diabetes, serious loss of blood-sugar control rose 3.7 percentage points.[37] The niacin was combined with an anti-flushing drug not sold as a supplement, so not every excess is provably niacin's alone.

A 2025 endocrinology guideline strongly recommends against niacin for people with high triglycerides who have, or are at risk of, heart disease.[38] Niacin remains a prescription drug for specific uses and the treatment for pellagra; routine use for heart protection is what's out. Nicotinamide riboside and NMN are different molecules — see Geroprotectors.

Red yeast rice is an unlabelled statin

Red yeast rice, sold for lowering cholesterol, is no gentle botanical: its active ingredient, monacolin K, is chemically the prescription statin lovastatin, and supplements can deliver a therapeutic dose.[39] The EU caps a daily portion below 3 mg of monacolins and requires warnings against use in pregnancy, under 18, over 70, and alongside cholesterol medication;[40] products bought elsewhere carry no such limit. Even so, European regulators could not find any daily intake free of safety concerns: muscle breakdown and liver injury have been reported at as little as 3 mg a day.[41]

It also carries a manufacturing risk you can't check. After a 2024 recall of red yeast rice tablets in Japan, a nationwide survey collected 192 affected patients, nearly all with reduced kidney function, and 87% still had it at last follow-up, though the authors say causality still needs confirming.[42] The cause was traced to a mould contaminating the factory, not the red yeast itself.[43]

If you need lower LDL, take a prescribed statin at a known dose with monitoring. See Lipid management.

Over-the-counter sleeping pills

Diphenhydramine and doxylamine block acetylcholine, a key nerve-signalling chemical, causing daytime sedation and cognitive effects. In a study of 3,434 adults aged 65 and over followed about seven years, the heaviest use of drugs like these — roughly a daily dose for more than three years — carried about 54% higher dementia risk, with risk rising with exposure; lighter use showed no significant increase.[44]

That is observational, and the size has shrunk: a study about seventeen times larger broke the signal down by drug class,[45] and pooled data from more than 600,000 people put the association well below 54%.[46] Early dementia may itself drive both poor sleep and the pill-taking.

Avoid chronic use at any adult age. The standard list of medicines older people should usually avoid names both drugs from age 65, a strong recommendation.[47] Below 65 there is no age cut-off in the evidence; the reason to skip them is that they work poorly.

Detox, adrenal, memory, immune and fat-burner products

Cleanses and "liver support" do nothing for a normal liver or kidneys, which detox themselves — and some herbs damage the liver. Kava is a well-known, rare but occasionally fatal cause of liver injury, typically two to twenty-four weeks after starting, and is banned or restricted in several countries.[48] Oral comfrey probably damages the liver's small blood vessels and is banned or restricted in most countries.[49] A few Chinese herbal medicines, such as Jin Bu Huan, Shou Wu Pian and Ma Huang (ephedra), have caused liver injury, and such products are often poorly labelled and can be contaminated with metals, bacteria or prescription drugs.[50] Green-tea-extract pills and high-absorption curcumin are two more documented causes; see Tea and Curcumin. If you're worried about your liver, get liver function tests rather than self-treating.

"Adrenal fatigue" is not a recognized diagnosis. A systematic review of 58 studies found no support for it, and no endocrinology society recognizes it.[51] Adrenal glandulars and megadose vitamin C or B5 protocols have no evidence of benefit. Real adrenal insufficiency needs medical evaluation and proper hormone replacement.

"Memory", "immune" and "fat-burner" blends are mostly unproven mixtures with poor quality control. Between 2007 and 2021 the US Food and Drug Administration found 1,068 supplement products spiked with undeclared prescription drugs — mostly sexual-enhancement and weight-loss products, one carrying as many as five drugs, and the commonest weight-loss adulterant being sibutramine, a drug withdrawn for heart harm.[52] That counts only products caught, not the share of the shelf, and the FDA's categories don't include memory or immune products. For immune support, vitamin D, zinc lozenges during a cold, and enough sleep beat any "immune booster".

Further reading

  • Bjelakovic G et al. Antioxidant supplements for prevention of mortality in healthy participants and patients with various diseases. Cochrane Database Syst Rev 2012.[53]
  • Ristow M et al. Antioxidants prevent health-promoting effects of physical exercise in humans. PNAS 2009.[54]
  • Miller ER 3rd et al. Meta-analysis: high-dosage vitamin E supplementation may increase all-cause mortality. Ann Intern Med 2005.[55]
  • Klein EA et al. Vitamin E and prostate cancer (SELECT). JAMA 2011.[56]
  • Bolland MJ et al. Effect of calcium supplements on risk of myocardial infarction and cardiovascular events: meta-analysis. BMJ 2010.[57]
  • Leitzmann MF et al. Zinc supplement use and risk of prostate cancer. J Natl Cancer Inst 2003.[58]
  • Gray SL et al. Anticholinergics and incident dementia. JAMA Intern Med 2015.[59]
  • ATBC: Beta-carotene and vitamin E in male smokers. NEJM 1994.[60]
  • CARET: Beta-carotene and retinol in smokers and asbestos workers. NEJM 1996.[61]
  • HPS2-THRIVE Collaborative Group. Effects of extended-release niacin with laropiprant in high-risk patients. NEJM 2014.[62]
  • Schürks M et al. Effects of vitamin E on stroke subtypes: meta-analysis of randomised controlled trials. BMJ 2010.[63]
  • Pilling LC et al. Common conditions associated with hereditary haemochromatosis genetic variants: cohort study in UK Biobank. BMJ 2019.[64]
  • Shinzawa M et al. A nationwide questionnaire study evaluated kidney injury associated with Beni-koji tablets in Japan. Kidney Int 2025.[65]

— § —