The short version
- In the Physicians' Health Study II, 14,641 men aged 50 and over took a daily multivitamin or placebo for a median of 11.2 years; total cancer incidence was 17.0 versus 18.3 events per 1,000 person-years, with no significant difference in cancer mortality.
- Pooling three US cohorts of 390,124 generally healthy adults followed for a median of 23.5 years found no longevity benefit from daily multivitamin use, with an adjusted hazard ratio for all-cause mortality of 1.04.
- The US Preventive Services Task Force concluded in 2022 that the evidence is insufficient to judge multivitamins for preventing cardiovascular disease or cancer, and recommended against beta carotene and vitamin E.
- In the COSMOS-Web trial, 3,562 older adults on a daily multivitamin scored significantly better on one memory test at one year, with no significant effects on any secondary outcome.
- For a healthy adult eating a reasonably varied diet, the honest expectation from a daily multivitamin is no detectable difference, at roughly $100 to $200 a year.
We do now sell a multivitamin, FOUNDATION, which makes this a harder article to write honestly than it used to be. We are going to write it anyway, and the following paragraphs cost usthing, so you can read them without wondering what we are steering you toward.
The multivitamin is the most-taken supplement in the world. Roughly a third of US adults report using one. It is also, for a general healthy adult, one of the weakest purchases in the category — and the evidence base is unusually large, which is what makes the conclusion sturdy rather than merely sceptical.
What the biggest trial found
Physicians' Health Study II randomised 14,641 male US physicians, aged 50 and over at entry, to a daily multivitamin or placebo, and followed them for a median of 11.2 years. Gaziano and colleagues published the cancer results in JAMA in 2012.
There was a statistically significant reduction in total cancer incidence: 17.0 events per 1,000 person-years on the multivitamin versus 18.3 on placebo, a hazard ratio of 0.92 (95% CI 0.86–0.998, P = .04). That is a real result from a well-run trial, and it deserves to be reported accurately — including the parts that did not move. There was no significant effect on prostate cancer, no significant effect on colorectal cancer, no significant effect on other site-specific cancers, and no significant difference in cancer mortality.
So: a modest signal on total incidence, in a cohort of older male physicians who are, as a group, unusually well-nourished and unusually health-engaged. Notice that the confidence interval's upper bound is 0.998. This is a result that squeaked past the line, not one that thundered past it.
What the biggest observational dataset found
Loftfield and colleagues pooled three large prospective US cohorts — 390,124 generally healthy adults, followed for a median of 23.5 years, during which 164,762 deaths occurred — to ask whether daily multivitamin use tracked with living longer.
It did not. The multivariable-adjusted hazard ratio for all-cause mortality among daily users was 1.04 (95% CI 1.02–1.07) in the first half of follow-up and 1.04 (95% CI 0.99–1.08) in the second. The authors' conclusion: they "did not find evidence to support improved longevity among healthy adults who regularly take multivitamins."
Two honest caveats, in both directions. This is observational, so it cannot establish cause. And a hazard ratio slightly above 1.0 almost certainly reflects people starting supplements because their health is declining, not the other way round — the authors went to some trouble to address that. What it does rule out is a large longevity benefit hiding in the data. There isn't one.
What the official read is
In 2022 the US Preventive Services Task Force updated its position after commissioning a fresh evidence review. On multivitamins for preventing cardiovascular disease or cancer, it issued an "I" statement: the evidence is insufficient to determine the balance of benefits and harms. On two specific supplements it went further and recommended against them — beta carotene, where it concluded with moderate certainty that the harms outweigh the benefits, and vitamin E, where it concluded there is no net benefit.
"Insufficient evidence" is not the same as "proven useless." It is a statement that after decades of study, the question is still open — which, for a product this widely taken, is itself informative.
The one place the recent news is better
The COSMOS-Web trial randomised 3,562 older adults to a daily multivitamin or placebo and tested memory annually over three years using an internet-based battery. The prespecified primary outcome — immediate recall on the ModRey test at one year — was significantly better on the multivitamin (P = 0.025), and the effect held on average across three years (P = 0.011). The authors estimated the difference as roughly equivalent to 3.1 years of age-related memory change.
Read the next sentence of the results too: multivitamin supplementation had no significant effects on any of the secondary outcomes, which included novel object recognition and executive function. And the "3.1 years" figure is an estimate derived by mapping the effect onto a cross-sectional age curve, not something the trial measured directly.
It is a genuine positive finding on a prespecified endpoint in a large trial. It is also one endpoint, in older adults, of modest size.
So who is a multivitamin actually for?
The evidence supports thinking about specific gaps rather than general insurance. The people most likely to benefit are the people most likely to have one:
- Anyone eating a genuinely restricted diet — very low calorie, highly limited variety, or excluding whole food groups without planning.
- Older adults with reduced appetite or intake.
- People after bariatric surgery or with a condition affecting absorption, where this is a clinical matter and should be managed as one.
- Anyone with a documented low level on a blood test.
- Anyone pregnant or planning a pregnancy — where nutrition is a conversation with a clinician, not a blog article, and prenatal formulations exist for reasons that have nothing to do with the trials above.
If none of those describe you, and you eat a reasonably varied diet, the honest expectation for a daily multivitamin is: no detectable difference, at about $100–$200 a year.
What we'd actually tell you
Measure before you supplement. A basic panel — vitamin D, B12, ferritin, thyroid function — tells you more in one morning than five years of a broad-spectrum tablet. If something is low, address that specific thing at a dose that matches it, and retest. If nothing is low, you have your answer and you have saved yourself a subscription.
None of the trials above are a reason to take or avoid anything for disease prevention; that is a conversation with your doctor. But if you were taking a multivitamin because you assumed the evidence was strong, it is worth knowing that it isn't — and that most of the good the supplement aisle can do you sits in a few specific, measurable places rather than in a tablet that covers everything a little.
Good questions
Is taking a multivitamin a waste of money?
For a healthy adult eating a reasonably varied diet, probably yes. The honest expectation is no detectable difference at roughly $100 to $200 a year. We sell one, and this answer still argues most people will not notice it. The evidence base here is unusually large, which is what makes the conclusion sturdy rather than merely sceptical.
Do multivitamins help you live longer?
No. Pooling three large US cohorts of 390,124 generally healthy adults, followed for a median of 23.5 years with 164,762 deaths, found no mortality benefit among daily users; the adjusted hazard ratio was 1.04. That is observational and cannot establish cause, but it does rule out a large longevity benefit hiding in the data.
Who actually should be taking a multivitamin?
People with a specific gap rather than a general worry. That includes anyone on a genuinely restricted diet, older adults with reduced appetite or intake, people after bariatric surgery or with a condition affecting absorption, anyone with a documented low level on a blood test, and anyone pregnant or planning a pregnancy, where prenatal formulations exist for separate reasons.
Are there any vitamins I should actively avoid?
The US Preventive Services Task Force recommends against two. It concluded with moderate certainty that the harms of beta carotene outweigh the benefits, and that vitamin E has no net benefit, for preventing cardiovascular disease or cancer. On multivitamins generally it issued an insufficient-evidence statement, which is not the same as proven useless. The question is still open.
Didn't a study show that multivitamins improve memory?
One did, partly. COSMOS-Web randomised 3,562 older adults and found significantly better immediate recall on one memory test at one year, holding on average across three years. Read the next line too: there were no significant effects on any secondary outcome, including novel object recognition and executive function. The quoted 3.1-years figure is a modelled estimate, not a measurement.
Should I get blood tests instead of taking a multivitamin?
Measure first, then supplement whatever is actually low. A basic panel covering vitamin D, B12, ferritin and thyroid function tells you more in one morning than five years of a broad-spectrum tablet. If something is low, address that specific thing at a dose that matches it, and retest. If nothing is low, you have your answer.
Sources
- Gaziano JM, et al. Multivitamins in the prevention of cancer in men: the Physicians' Health Study II randomized controlled trial. JAMA, 2012. View study
- Loftfield E, et al. Multivitamin Use and Mortality Risk in 3 Prospective US Cohorts. JAMA Netw Open, 2024. View study
- US Preventive Services Task Force. Vitamin, Mineral, and Multivitamin Supplementation to Prevent Cardiovascular Disease and Cancer: US Preventive Services Task Force Recommendation Statement. JAMA, 2022. View study
- Yeung LK, et al. Multivitamin Supplementation Improves Memory in Older Adults: A Randomized Clinical Trial. Am J Clin Nutr, 2023. View study
These statements have not been evaluated by the Food and Drug Administration. This product is not intended to diagnose, treat, cure, or prevent any disease.
