A 2024 analysis of three COSMOS cognitive substudies involving 5,203 older US adults found a 0.07-standard-deviation difference in global cognition and a 0.06-standard-deviation difference in episodic memory between multivitamin and placebo groups. Those test-score changes were small, and the analysis did not show that multivitamins prevent dementia.

The parent trial produced a less favorable result on its main disease outcomes. It found no reduction in total cancer, total cardiovascular disease or death from any cause, leaving little support for treating a daily multivitamin as broad protection against chronic disease.

Suspected overdose requires immediate poison advice

A child or adult who may have swallowed much more than the labeled amount needs immediate advice from a local poison service or medical professional, even if no symptoms have appeared. Australia's government-funded Healthdirect service says suspected poisoning needs medical help immediately and identifies vomiting, drowsiness, seizures and breathing problems as warning signs.

Collapse, loss of consciousness, seizures or breathing difficulty require the local emergency service. The product container or a photograph of its label can help clinicians identify the ingredients. Vomiting should not be induced unless a poison specialist or clinician instructs it.

COSMOS found no reduction in its main disease outcomes

The randomized, placebo-controlled COSMOS trial enrolled 21,442 US adults and followed them for a median of 3.6 years. Women were at least 65 and men at least 60 when they entered the trial, and participants did not have major cardiovascular disease or a recent cancer diagnosis at baseline.

Total invasive cancer occurred in 518 participants assigned the multivitamin and 535 assigned placebo, producing a hazard ratio of 0.97 with a confidence interval that crossed no effect. The corresponding hazard ratios were 0.98 for the composite cardiovascular outcome and 0.93 for all-cause mortality; neither difference was statistically conclusive.

A chart compares cancer, cardiovascular and mortality results from a multivitamin trial (illustrative image)

The study tested one daily multivitamin formulation in older US adults. It did not test every formula, high-dose single-nutrient products or treatment for a diagnosed deficiency. Its median follow-up may also have been too short to detect some long-term cancer or cardiovascular outcomes.

The cognition result is measurable but modest

The 2024 paper combined three nonoverlapping COSMOS cognitive substudies. The pooled estimate favored the multivitamin for global cognition and episodic memory, and the authors described the global result as equivalent to about two years less cognitive aging.

That equivalence is a statistical interpretation of changes on cognitive tests, not evidence that each participant became two years younger or noticed an improvement. The estimated differences were 0.07 standard deviation units for global cognition and 0.06 for episodic memory. The clinic substudy alone found a 0.06-unit global-cognition difference whose confidence interval included no effect, while its episodic-memory result favored the multivitamin.

Two evidence panels compare neutral disease outcomes with a small cognitive test difference (illustrative image)

The cognition findings apply most directly to adults aged 60 and older in the United States who resembled the study participants. They do not establish benefit for younger people, and cognitive test performance is not the same endpoint as a diagnosis of mild cognitive impairment or dementia.

US guidance separates multivitamins from high-dose nutrients

The US Preventive Services Task Force concluded in 2022 that evidence was insufficient to assess the balance of benefits and harms of multivitamins for preventing cardiovascular disease or cancer in community-dwelling, nonpregnant adults. An insufficient-evidence finding is not proof of benefit or proof of no effect; it means the task force could not determine the balance.

The same recommendation advises against beta-carotene or vitamin E supplements for those prevention goals. It reports increased lung-cancer risk from beta-carotene in people who smoke or had occupational asbestos exposure and notes that excessive vitamin doses can cause harm. Evidence about a basic multivitamin therefore cannot be transferred to concentrated single-nutrient supplements.

A nutrient gap and a disease-prevention claim are different questions

The US National Institutes of Health Office of Dietary Supplements says multivitamins may help people obtain recommended nutrient amounts when food alone is not enough, but they cannot replace a varied diet. It also notes that many products contain only small amounts of calcium and magnesium, so the presence of a nutrient on the label does not show that the product closes a particular shortfall.

Population averages cannot diagnose an individual deficiency. Diet history, symptoms and laboratory testing address different questions, and the appropriate assessment depends on the nutrient and clinical setting. The COSMOS disease-prevention results do not determine how a confirmed deficiency should be treated.

Pregnancy, childhood and regular medicines need separate decisions

Pregnancy

The COSMOS prevention findings do not apply to people who are pregnant or planning pregnancy. Prenatal nutrient needs and product formulations differ from those of other adults, and excessive preformed vitamin A during pregnancy can cause birth defects. Selection of a prenatal supplement belongs with local clinical guidance rather than the COSMOS result.

Children

Children were not enrolled in COSMOS, and adult formulations should not be treated as evidence for pediatric use. Iron-containing products can be especially dangerous after accidental large ingestion, which is why supplements should remain secured and out of children's reach.

Chronic illness and regular medicines

The US National Institutes of Health warns that vitamin K in a multivitamin can reduce the effectiveness of warfarin and that formulas should be considered in relation to age, sex and life stage. People with chronic illness, known nutrient deficiencies, previous surgery affecting absorption or regular medicines fall outside the healthy-population question addressed by COSMOS and need an individual clinical assessment.

What the current evidence supports

The strongest direct trial evidence does not support a multivitamin as a general strategy to prevent cancer, cardiovascular disease or death in otherwise healthy older adults. The cognitive substudies offer a narrower positive finding: a small average difference on tests among older US participants, without proof of dementia prevention.

That leaves two questions that should not be merged. One is whether a broad supplement prevents major disease in a generally well-nourished population; current evidence does not establish that benefit. The other is whether a particular person has a dietary gap, life-stage need or diagnosed deficiency, which requires evidence specific to that nutrient and circumstance.