Does a daily multivitamin help you live longer?
Claim attributed to The supplement industry and the roughly one in three US adults who take a daily multivitamin , Marketed widely as an "insurance policy" for health. Industry-promoted and culturally entrenched rather than tied to one named advocate.
For living longer, it fails: a 390,000-person cohort and two large RCTs show no mortality or cardiovascular benefit. But it is not nothing, since randomized trials show a small, replicated memory benefit in older adults and a multivitamin does fill genuine nutrient gaps.
It fills nutrient gaps and nudges memory a little in older adults; it does not make a well-fed person live longer, which is the part being sold.
What it’s supposed to target
- Micronutrient 'insurance'
- Antioxidant vitamins (C, E, beta-carotene)
- Correcting subclinical deficiency
- Oxidative-stress theory of aging
The pitch rests on the insurance hypothesis: modern diets may leave small gaps in vitamins and minerals, so a daily multivitamin tops them up, and via antioxidant vitamins (C, E, beta-carotene) fighting oxidative damage it might slow aging and prevent disease. Correcting a genuine deficiency (B12, folate, or vitamin D in at-risk groups) is real medicine; the longevity claim extends that logic to everyone.
The logic breaks at the population level. Most well-nourished adults are not deficient, so topping up adds little, and high-dose antioxidants can even backfire (beta-carotene raised lung-cancer risk in smokers). Large trials and a very large cohort found no mortality benefit from daily multivitamins in generally healthy people, and a whole-food diet delivers micronutrients in a matrix a pill cannot copy. A reasonable-sounding mechanism, not a longevity effect.
Mechanism is theory, not proof. A plausible pathway explains why something might work, not whether it does. The verdict rests on the evidence below, not the elegance of the theory.
What would have to be true
The longevity half: a multivitamin would lower mortality in healthy adults. FAILS, no benefit in a 390,000-person cohort or in two RCTs.
The health half: it would improve outcomes that matter. PARTLY HOLDS, a small replicated memory benefit and nutrient-gap filling, but not heart disease, cancer death, or lifespan.
What the evidence actually shows
Living longer: tested directly, and it fails
The strongest test of the longevity claim is Loftfield 2024 (JAMA Network Open), which pooled 390,124 generally healthy US adults across three cohorts over up to 27 years: daily multivitamin users did not live longer, with a slightly higher early all-cause mortality (adjusted HR 1.04, 95% CI 1.02-1.07) the authors attribute to residual confounding. Two large RCTs agree: Physicians' Health Study II found no cardiovascular benefit (major events HR 1.01, 95% CI 0.91-1.10) and no mortality benefit, and the COSMOS trial echoed the null (CVD HR 0.98). For living longer or protecting the heart, the multivitamin does nothing.
Improving health: a small but real cognitive signal
The other half is not empty. Three randomized trials in the COSMOS program, synthesized in a pre-specified meta-analysis (Vyas 2024), show a small but statistically robust and replicated benefit on cognition (global cognition SMD 0.07, 95% CI 0.03-0.11) and episodic memory, roughly the equivalent of slowing cognitive aging by about two years; the effect did not reduce dementia and used one specific formulation. PHS II also found a modest 8% drop in total cancer incidence (HR 0.92, 95% CI 0.86-0.998), though COSMOS did not replicate it and cancer mortality was unchanged. And a multivitamin reliably fills nutrient gaps in people whose diets fall short. Real, specific, modest benefits, not the broad longevity payoff the marketing implies.
Studies, graded, and who paid
Refuted: a 390,000-adult cohort (Loftfield 2024) and two RCTs show no mortality benefit, with a slightly higher early HR 1.04.
No benefit: PHS II major-event HR 1.01 (0.91-1.10) and COSMOS HR 0.98 are both null.
Small but replicated RCT benefit (COSMOS-Mind, COSMOS-Web, Vyas 2024 meta-analysis); roughly two years of cognitive aging, with no drop in dementia.
One positive RCT (PHS II HR 0.92) not replicated by COSMOS, and cancer mortality is unchanged.
| # | Study | Type | Size | Funding / COI | Key limitations |
|---|---|---|---|---|---|
| 1 | Loftfield 2024, NIH/NCI 3-cohort mortality analysis | Prospective cohort (3 pooled US cohorts) | 390,124 adults, up to 27 years | Independent NIH Intramural; no conflicts reported. | Observational; early excess hazard likely reverse causation, not harm. |
| 2 | Gaziano 2012, PHS II (cancer) | Randomized, double-blind, placebo-controlled trial | 14,641 male physicians, median 11.2 years | Mixed NIH grants; manufacturer-supplied agents. Primary endpoints null. | Male physicians 50+, well-nourished; limited generalizability. |
| 6 | Sesso 2012, PHS II (cardiovascular) | Randomized, double-blind, placebo-controlled trial | 14,641 male physicians, median 11.2 years | Mixed NIH grants; manufacturer-supplied agents. | No CVD or all-cause mortality benefit; same narrow population. |
| 3 | USPSTF 2022 recommendation | Guideline / evidence synthesis | Systematic-review-based | Independent Government-appointed panel; no industry funding. | Insufficient evidence verdict; recommends against beta-carotene, vitamin E. |
| 5 | Yeung 2023, COSMOS-Web (memory) | Randomized trial, ancillary sub-study | 3,562 older adults, 3 years | Industry-funded Mars Edge; tablets donated by Pfizer/Haleon (Centrum maker). | Web-based testing; cognition not longevity; sponsor conflict. |
| 7 | COSMOS-Mind (Baker 2023), cognition RCT | Randomized controlled trial | Older adults, 3 yr | Mixed NIH-funded; supported by Mars Edge, Centrum donated by Pfizer/Haleon. | Small effect; single formulation; telephone-based cognitive testing. |
| 8 | Vyas 2024, pre-specified COSMOS cognition meta-analysis | Meta-analysis of 3 RCTs | Three COSMOS cognitive studies | Mixed COSMOS program (Mars Edge; Pfizer/Haleon). | Global-cognition SMD about 0.07; no dementia-incidence reduction. |
The claim bundles two different things: a longevity and heart benefit the best evidence refutes, and a small cognitive benefit that randomized trials genuinely support.
Unproven ≠ disproven
This is not merely untested. Mortality and cardiovascular benefit were measured in large RCTs and a 390,000-person cohort and did not appear, while the cognitive benefit is shown but small.
Where claim and evidence diverge
Marketing sells the multivitamin as broad health insurance and a longer life; the evidence supports neither, and the genuine wins (a little memory benefit, filling nutrient gaps) are narrower than the pitch.
The money trail
The decisive mortality null (Loftfield 2024) is independent NIH work with nothing to sell. The positive cognition trials (COSMOS) were supported by Mars Edge and used Centrum tablets donated by Pfizer and Haleon, a multivitamin maker, a conflict to weigh, though they were randomized with pre-specified endpoints.
The honest read
For a longer life or a healthier heart, a daily multivitamin will not deliver, and a whole-food diet does more. For an older adult worried about memory there is modest randomized evidence of a small benefit, and correcting a real deficiency is worthwhile. Useful in specific cases, not the general longevity insurance it is sold as.
What would change this verdict
A large independent RCT showing daily multivitamins lower all-cause mortality in healthy adults.
Independent, non-industry replication of the COSMOS cognitive benefit that also lowers dementia or mild-impairment rates.
Sources
- Loftfield E, et al. Multivitamin Use and Mortality Risk in 3 Prospective US Cohorts. JAMA Network Open. 2024;7(6):e2418729.
- Gaziano JM, et al. Multivitamins in the Prevention of Cancer in Men: PHS II RCT. JAMA. 2012;308(18):1871-1880.
- US Preventive Services Task Force. Vitamin, Mineral, and Multivitamin Supplementation to Prevent CVD and Cancer. JAMA. 2022;327(23):2326-2333.
- O'Connor EA, et al. Vitamin and Mineral Supplements for Primary Prevention: USPSTF Evidence Report. JAMA. 2022;327(23):2334-2347.
- Yeung LK, et al. Multivitamin Supplementation Improves Memory in Older Adults (COSMOS-Web). Am J Clin Nutr. 2023;118(1):273-282.
- Sesso HD, et al. Multivitamins in the Prevention of CVD in Men: PHS II RCT. JAMA. 2012;308(17):1751-1760.
- Baker LD, Manson JE, Rapp SR, et al. Effects of cocoa extract and a multivitamin on cognitive function: a randomized clinical trial (COSMOS-Mind). Alzheimers Dement. 2023;19(4):1308-1319. PMID 36102337.
- Vyas CM, Manson JE, Sesso HD, et al. Multivitamin-mineral supplementation and cognitive function: COSMOS clinic subcohort and meta-analysis of three COSMOS cognitive studies. Am J Clin Nutr. 2024;119(3):692-701. PMID 38244989.
People also ask
- Do daily multivitamins actually help you live longer?
- No, not for healthy, well-nourished adults. The largest trial (Physicians' Health Study II) and the largest cohort (an NIH analysis of about 390,000 adults) both found no all-cause mortality benefit. A whole-food diet outperforms the pill for longevity purposes.
- Do multivitamins prevent heart disease or cancer?
- No clear benefit. Physicians' Health Study II found no effect on major cardiovascular events, and the USPSTF rates the evidence insufficient (Grade I). The trial did find a small drop in total cancer incidence (HR 0.92), but cancer death was not significantly reduced.
- When does it make sense to take a multivitamin?
- Supplementing makes sense to correct a documented deficiency, not as insurance for everyone. The evidence against multivitamins applies to healthy, well-fed adults. Long trials in less well-nourished or non-Western populations are largely missing, so a benefit in deficient groups remains genuinely open.
- Is the COSMOS multivitamin memory study trustworthy?
- Treat it with caution. COSMOS-Web (3,562 older adults) met its prespecified primary cognitive endpoint, but it is an industry-funded ancillary sub-study (Mars Edge; Centrum tablets donated by Pfizer/Haleon, a multivitamin maker) using web-based testing. It measured cognition, not lifespan.
Part of our guide: Longevity supplements, fact-checked
Caveat is journalism, not medical advice. We check public claims against published evidence; we don’t diagnose, treat, or tell you what to take.