Which supplements should you avoid, and why?

1. Beta-carotene if you smoke, or used to. In a trial of 29,133 male smokers in Finland, 20 mg of beta-carotene a day raised lung cancer incidence by 18%². In a second trial, beta-carotene plus vitamin A raised the relative risk of lung cancer to 1.28 among 18,314 smokers, former smokers and asbestos workers, and the trial was stopped early³. The task force recommends against beta-carotene supplements¹. Food is another matter: diets rich in carotenoid fruits and vegetables were linked to less lung cancer, which is what prompted the trials²; see beta-carotene in food versus pills.

2. High-dose vitamin E. In a trial of 35,533 relatively healthy men, 400 IU a day of synthetic vitamin E raised prostate cancer by 17%⁴, and pooled trials of 400 IU a day or more found slightly higher death rates⁵. Two trials found more bleeding strokes with alpha-tocopherol, one of them at 50 mg a day, far below the upper limit⁶. See vitamin E.

3. Antioxidant megadoses in general. Across 78 trials with 296,707 people, antioxidant pills did not lower death rates; in the better-run trials, beta-carotene and vitamin E slightly raised them, and higher doses of vitamin A may too⁷. Vitamin C and selenium made no difference either way⁷.

4. Iron you do not need. Iron pills can upset the stomach at 45 mg a day or more and reduce zinc absorption at 25 mg or more⁸. About 1 in 10 white people carry a hemochromatosis gene variant, and people with hemochromatosis should avoid iron and vitamin C supplements⁸. Get ferritin checked first; see do you need a supplement and iron basics.

5. Calcium pills in place of calcium-rich food (debated). Calcium from food goes with fewer kidney stones, and calcium pills with more^11,13. Whether the pills raise heart risk is disputed: one meta-analysis found 27% to 31% more heart attacks⁹, another found no link within the upper limits¹⁰, and after more than two decades of follow-up in a large trial, calcium with vitamin D went with 6% more cardiovascular deaths but 7% fewer cancer deaths¹². If you need more calcium, the fact sheet's advice is to get it mainly from food and to stay within the limits¹⁴. See calcium.

6. Proprietary blends. A label may list a "proprietary blend" with only its total weight: the ingredients appear in order of weight, with no amount for each¹⁵. A dose you cannot see is a dose you cannot compare with the research or the upper limit. Our vitamin and mineral picks exclude blends that hide amounts.

7. "Detox" and weight-loss products. No randomized trial had tested a commercial detox diet as of a 2015 review¹⁶. A systematic review of 14 weight-loss supplements and alternative therapies found that only 52 of the trials were at low risk of bias with enough data, and only 16 of those found a significant weight difference, of 0.3 to 4.93 kg¹⁷. Weight-loss products are also among the most often found to contain hidden drugs¹⁸. See detox diets and teas and what works for weight loss.

8. Products with hidden drugs. From 2007 to 2016, the FDA warned about 776 supplements containing undeclared drugs, mostly sold for sexual enhancement, weight loss or muscle building¹⁸. Its Health Fraud Product Database lists more than 2,000 products cited for problems such as hidden ingredients or unproven disease claims, which the FDA says are only a small fraction of the hazardous products on sale¹⁹. See our guide to supplement purity.

9. Herbal extracts that can injure the liver. Green tea extract, turmeric products and kava have all been linked to serious liver injury^20,21,22. Some of the turmeric cases involved products that also contained piperine from black pepper²⁴. The cases involve concentrated products: green tea as a drink has been implicated only rarely, in large amounts²⁰. See turmeric, green tea catechins and kava.

10. Herbs and vitamins that clash with medicines. St John's wort can weaken many medicines in dangerous ways, including birth control pills, cyclosporine, some HIV and cancer drugs, digoxin, warfarin and some antidepressants²⁷. Vitamin K changes how warfarin works, so people on warfarin need a steady intake³⁰. Check pairs in our interaction checker, read which herbs interact with medications, and see St John's wort.

Before you add anything, check how much is too much, then use the supplement schedule for when to take what you keep.

Beta-carotene in smokers

The Alpha-Tocopherol, Beta-Carotene trial randomized 29,133 male smokers aged 50 to 69 in southwestern Finland to 50 mg of alpha-tocopherol, 20 mg of beta-carotene, both or placebo for five to eight years. Among 876 new lung cancers, incidence was 18% higher with beta-carotene (95% CI 3% to 36%), and total mortality was 8% higher (95% CI 1% to 16%), mainly from lung cancer and ischemic heart disease². The Beta-Carotene and Retinol Efficacy Trial (CARET) gave 30 mg of beta-carotene plus 25,000 IU of retinol a day, or placebo, to 18,314 smokers, former smokers and asbestos-exposed workers. After a mean of 4.0 years there were 388 lung cancers; the relative risk was 1.28 (95% CI 1.04–1.57), deaths from any cause rose (RR 1.17, 95% CI 1.03–1.33), and the trial was stopped 21 months early³.

Vitamin E

SELECT randomized 35,533 relatively healthy men in the United States, Canada and Puerto Rico to selenium, vitamin E (400 IU a day of synthetic all-rac-alpha-tocopheryl acetate), both, or placebo. With longer follow-up, 620 men in the vitamin E group developed prostate cancer against 529 on placebo (HR 1.17, 99% CI 1.004–1.36), an absolute increase of 1.6 cases per 1,000 person-years⁴. A dose-response meta-analysis of 19 trials with 135,967 participants found 39 more deaths per 10,000 people in trials of 400 IU a day or more (95% CI 3 to 74), with risk estimated to rise above 150 IU a day; many of those trials were small and in people with chronic disease⁵. SELECT's 400 IU of synthetic vitamin E is 180 mg, well under the 1,000 mg upper limit, and one of the stroke trials used 50 mg a day⁶: staying under a limit does not make a supplement safe.

Antioxidants pooled

A Cochrane review pooled 78 randomized trials with 296,707 participants. Overall, antioxidant supplements had no significant effect on mortality (random-effects RR 1.02, 95% CI 0.98–1.05), but in the 56 trials at low risk of bias, mortality was higher (RR 1.04, 95% CI 1.01–1.07)⁷. In those trials beta-carotene (RR 1.05, 95% CI 1.01–1.09) and vitamin E (RR 1.03, 95% CI 1.00–1.05) increased mortality, vitamin A, vitamin C and selenium did not significantly change it, and higher vitamin A doses were associated with more deaths⁷. The authors concluded that antioxidant supplements should be evaluated like medicines before they are marketed⁷. The USPSTF concluded with moderate certainty that the harms of beta-carotene outweigh its benefits and that vitamin E has no net benefit for preventing cardiovascular disease or cancer, and recommends against both¹.

Calcium: the debate in numbers

A meta-analysis of 15 placebo-controlled trials of calcium supplements (500 mg a day or more, without vitamin D) found more heart attacks: 143 versus 111 in five trials with patient-level data (HR 1.31, 95% CI 1.02–1.67), and a pooled relative risk of 1.27 (95% CI 1.01–1.59) in the trial-level analysis⁹. A later review, funded by the National Osteoporosis Foundation, covered 4 trials and 27 observational studies; the trials found no significant differences in cardiovascular events, and the authors concluded that calcium within the upper limits of 2,000 to 2,500 mg a day is not associated with cardiovascular risk in generally healthy adults, though heart outcomes were secondary end points in every trial¹⁰.

In the Women's Health Initiative, 36,282 postmenopausal women took 1,000 mg of calcium as calcium carbonate with 400 IU of vitamin D3, or placebo, for an average of 7.0 years: hip fractures did not fall significantly (HR 0.88, 95% CI 0.72–1.08) and kidney stones rose (HR 1.17, 95% CI 1.02–1.34)¹¹. After a median of 22.3 years, the supplement group had 7% fewer cancer deaths (HR 0.93, 95% CI 0.87–0.99) and 6% more cardiovascular deaths (HR 1.06, 95% CI 1.01–1.12), with no change in overall mortality; this post hoc analysis could not separate calcium from vitamin D¹². Among 91,731 nurses followed for 12 years, women with the most dietary calcium had fewer kidney stones (RR 0.65, 95% CI 0.50–0.83), while supplement users had more (RR 1.20, 95% CI 1.02–1.41)¹³. The fact sheet calls the cardiovascular findings mixed, and the calcium upper limit itself rests on the kidney-stone risk of supplements¹⁴.

Iron

Adults with normal digestion have very little risk of iron overload from food. Supplements are another matter: 45 mg a day or more can cause stomach upset and constipation, 25 mg or more can reduce zinc absorption, and an extreme overdose can cause intestinal necrosis, organ failure and death⁸. About 1 in 10 white people carry the most common hemochromatosis variant, and 4.4 per 1,000 have two copies and the disease; untreated, people with hemochromatosis typically show iron damage by their 30s, and liver specialists advise them to avoid iron and vitamin C supplements⁸. In homes with young children, an iron pill taken just in case is also a poisoning risk⁸.

Detox and weight-loss products

A 2015 review found very little clinical evidence for detox diets: a handful of small studies with flawed methods, and no randomized trials of commercial detox programs¹⁶. A systematic review screened 20,504 records and 1,743 full-text articles, 315 of which were randomized controlled trials of 14 weight-loss supplements and therapies; only 52 (16.5%) were at low risk of bias with sufficient data, and only 16 of those (31%) found significant differences in weight, ranging from 0.3 to 4.93 kg¹⁷.

Hidden drugs

Between 2007 and 2016, the FDA identified 776 adulterated supplements from 146 companies: 45.5% sold for sexual enhancement, 40.9% for weight loss and 11.9% for muscle building, with 20.2% containing more than one unapproved ingredient¹⁸. Sildenafil was in 47.0% of the sexual-enhancement products, sibutramine in 84.9% of the weight-loss products, and steroids or steroid-like ingredients in 89.1% of the muscle products¹⁸. Of 28 products named in more than one warning, 19 (67.9%) had new unapproved ingredients the second or third time¹⁸. The FDA's Health Fraud Product Database collects products cited in warning letters, recalls and public notifications; it held 2,217 entries on 5 November 2026, and the FDA says it covers only a small fraction of the hazardous products on sale¹⁹. Recent entries include products with undeclared diclofenac (a prescription painkiller) and dexamethasone (a steroid), and products with undeclared sildenafil or tadalafil, the erectile dysfunction drugs¹⁹.

Liver injury

The US Drug-Induced Liver Injury Network enrolled 839 people with liver injury from medicines or supplements between 2004 and 2013. Herbal and dietary supplements caused 130 cases (15.5%), their share rose from 7% to 20%, and injury from supplements other than bodybuilding products ended in death or liver transplant more often than injury from medicines (13% versus 3%)²³. LiverTox, the NIH's liver-injury database, records green tea extract, and more rarely large amounts of green tea, in cases of acute liver failure needing an urgent transplant or ending in death²⁰; several dozen cases of acute liver injury from turmeric products²¹; and kava products linked to severe, sometimes fatal, liver injury²².

In 10 turmeric cases in the network, all enrolled since 2011, five people were hospitalized and one died of acute liver failure; 3 of the 7 products tested also contained piperine, and the authors suggested that rising case numbers may reflect the combination with black pepper²⁴. A United States Pharmacopeia review found that green tea extract taken in bolus doses on an empty stomach raises blood levels of its main catechin, EGCG, linked liver injury to EGCG intakes from 140 mg to about 1,000 mg a day, and led to a label warning to take the extract with food and to stop if signs of liver trouble appear²⁵. An estimated 15.6 million US adults took at least one of six botanicals with known liver liability, led by turmeric and green tea, in a 30-day window²⁶.

Interactions with medicines

The National Center for Complementary and Integrative Health states that St John's wort can interact in dangerous, sometimes life-threatening ways with many medicines. It can weaken birth control pills, cyclosporine, some seizure, heart, HIV and cancer drugs, warfarin, some statins and some antidepressants, and combined with other drugs that raise serotonin it can cause serious side effects²⁷. In 12 healthy volunteers, 14 days of St John's wort (300 mg three times a day) halved exposure to the test drug alprazolam and cut its half-life from 12.4 to 6.0 hours, by inducing the CYP3A4 enzyme that handles at least half of all marketed medicines²⁸. The herb has not been convincingly linked to liver injury itself, but it can make other drugs more harmful to the liver by changing how they are broken down²⁹. Vitamin K can interact seriously with warfarin and similar anticoagulants; sudden changes in intake can strengthen or weaken the drug, so people taking them need a consistent intake from food and supplements³⁰. Our interaction checker covers a short list of combinations, so no warning there is not clearance; your pharmacist can check everything you take.

Bottom line (confidence 4 of 5)

Two large randomized trials found more lung cancer with beta-carotene in smokers, and a large trial found more prostate cancer with vitamin E; pooled trials point the same way, and the task force recommends against both. The evidence on calcium pills is genuinely mixed. The cautions about iron, blends, detox and weight-loss products, tainted products, liver-injury herbs and interactions rest on case series, registries and regulators' records rather than trials, but they describe real, documented harm. Before adding any supplement, ask your pharmacist or doctor, especially if you take prescription medicines.