Vitamin Pills vs. Food: What Trials Show About Cancer Risk

The short answer

No. In the large trials we could open, vitamin and antioxidant pills did not work like eating the same nutrients in food. They did not clearly lower cancer risk. A few pills raised risk in some groups.

The National Cancer Institute (NCI) looked at nine large randomized trials of antioxidant pills for cancer prevention. It said they "did not provide evidence that dietary antioxidant supplements are beneficial" for preventing cancer.

Multivitamins are less clear. Two large trials found a small drop or no drop in cancer. The US Preventive Services Task Force (USPSTF) says the evidence is not enough to say if the benefits outweigh the harms.

The idea as it circulates

The idea goes like this. Fruits and vegetables are linked to lower cancer risk. They are full of vitamins and antioxidants. So a pill with the same vitamins should give the same protection, with less effort.

Part of this is reasonable. NCI explains that antioxidants can block free radicals in lab and animal studies. Free radicals are unstable molecules that can damage cells, including DNA. The question is whether a pill does this in people. That is what the trials tested.

What the trials show: single antioxidant pills

NCI's summary covers nine randomized trials. Here are the results that matter most.

• Beta-carotene in smokers. In the ATBC trial in Finland, male smokers who took beta-carotene had more lung cancer. In the CARET trial in the US, people at high risk of lung cancer who took beta-carotene plus vitamin A had more lung cancer and more deaths from all causes. The trial ended early. NCI says these higher risks persisted later.
• Vitamin E and selenium (SELECT). SELECT enrolled 35,533 men. At first, neither vitamin E (400 IU a day) nor selenium (200 micrograms a day) lowered prostate cancer (Lippman 2009). With longer follow-up, 620 men on vitamin E got prostate cancer, versus 529 on placebo. That was a hazard ratio of 1.17 (Klein 2011). Per 1,000 person-years, that was 1.6 more cases with vitamin E.
• Other trials. NCI lists no benefit or no clear effect for most of the rest. One French trial (SU.VI.MAX) found lower cancer in men only, and more skin cancer in women only.

NCI notes one possible reason for the lack of benefit. Pills have a purified chemical. Food has a mix of vitamins, minerals and other compounds that may work together. That is an explanation under study, not a proven one.

What the trials show: multivitamins

A multivitamin has many nutrients at lower doses. Two large trials tested one.

• Physicians' Health Study II. 14,641 male US doctors, about 11 years. Total cancer was 17.0 per 1,000 person-years on a multivitamin and 18.3 on placebo. That is a hazard ratio of 0.92 (95% range 0.86 to 0.998), a small drop right at the edge of chance. There was no clear effect on prostate cancer, colorectal cancer or cancer deaths (Gaziano 2012).
• COSMOS. 21,442 older US adults, about 3.6 years. Invasive cancer occurred in 518 people on a multivitamin and 535 on placebo, a hazard ratio of 0.97 (95% range 0.86 to 1.09). That is no clear effect. Lung cancer was lower (hazard ratio 0.62), but the trial reported this as one of many secondary results (Sesso 2022).

The USPSTF review pooled four multivitamin trials. It found a small drop in any cancer (odds ratio 0.93, range 0.87 to 0.99). The review said the evidence had "important limitations." Its summary was "little or no benefit," apart from that small cancer finding.

What the trials show: vitamin D

The VITAL trial gave 25,871 adults vitamin D (2,000 IU a day) or placebo for a median of 5.3 years. Cancer was diagnosed in 793 people on vitamin D and 824 on placebo. That was a hazard ratio of 0.96 (range 0.88 to 1.06). It did not lower cancer (Manson 2019). Our page on vitamin D and cancer has more.

What official advice says

The USPSTF (2022) recommends against beta-carotene or vitamin E pills to prevent cancer or heart disease. It says harms of beta-carotene outweigh benefits, and that vitamin E has no net benefit. For multivitamins and for single or paired nutrients, it says the evidence is "insufficient." That is not a green light, and it is not a warning. It means we do not know.

The Dietary Guidelines for Americans say that nutritional needs should be met mainly from foods and drinks. They also say that fortified foods and supplements can be useful when it is not possible to meet a need otherwise.

What about cancer?

For people without cancer, the trials above do not support taking vitamin or antioxidant pills to prevent it. For smokers and people at high risk of lung cancer, the beta-carotene results are a clear warning.

NCI says the research in humans "has not demonstrated convincingly" that antioxidant pills lower the chance of getting or dying from cancer. It adds that some studies showed more risk for some cancers.

For people in cancer treatment, NCI says some trials found worse outcomes in people who took antioxidant supplements, especially smokers. NCI's advice is to use them with caution and to tell your doctors about any supplement you take. Ask your care team before you start or stop any pill. Our pages on antioxidants and treatment and vitamins during chemo have more.

None of this says that a pill treats or prevents cancer. Nor does it say that a pill is needed to make up for a bad diet.

What this story cannot tell you

• Most trials tested one nutrient at a high dose, not a typical multivitamin. They cannot tell you what the same nutrient does in food.
• We found no randomized trial that compared a pill to the same nutrient in food for cancer. The "complex mixture" idea is an explanation, not a result.
• Follow-up in these trials ranged from a few years to about 12. Cancer can take longer to develop.
• The multivitamin trials were in older adults, and PHS II was in male doctors only. Results may differ for others.
• The COSMOS lung cancer result was one of many secondary checks. It needs to be repeated before anyone relies on it.
• These trials cannot speak to people with a true shortage of a nutrient, who may have a medical reason to take a pill.
• We could not open the NIH Office of Dietary Supplements fact sheets (the site blocked our access), so we did not use them.

Sources

• NCI, Antioxidants and Cancer Prevention: https://www.cancer.gov/about-cancer/causes-prevention/risk/diet/antioxidants-fact-sheet
• NCI, Risk Factors: Diet: https://www.cancer.gov/about-cancer/causes-prevention/risk/diet
• USPSTF 2022, Vitamin, Mineral, and Multivitamin Supplementation to Prevent Cardiovascular Disease and Cancer (JAMA): https://pubmed.ncbi.nlm.nih.gov/35727271/
• Evidence report for the USPSTF, JAMA 2022: https://pmc.ncbi.nlm.nih.gov/articles/PMC13207626
• Sesso et al., American Journal of Clinical Nutrition 2022, COSMOS: https://pmc.ncbi.nlm.nih.gov/articles/PMC9170475
• Gaziano et al., JAMA 2012, Physicians' Health Study II: https://pmc.ncbi.nlm.nih.gov/articles/PMC3517179
• Lippman et al., JAMA 2009, SELECT: https://pmc.ncbi.nlm.nih.gov/articles/PMC3682779
• Klein et al., JAMA 2011, SELECT follow-up: https://pmc.ncbi.nlm.nih.gov/articles/PMC4169010
• Manson et al., NEJM 2019, VITAL: https://pmc.ncbi.nlm.nih.gov/articles/PMC6425757
• Dietary Guidelines for Americans 2020 to 2025: https://www.dietaryguidelines.gov/sites/default/files/2020-12/Dietary_Guidelines_for_Americans_2020-2025.pdf

How this article was prepared

An AI-assisted editorial system helped prepare this page. No named medical reviewer has reviewed it unless one is listed. We opened the sources above, except where a limit is stated in the text.

See an error, old source, or unclear wording? Tell us.