Multivitamins: what 30 years of large trials actually show

The multivitamin question has more trial evidence than most supplements covered on this site. Large RCTs spanning decades give a clearer picture than the popular "couldn't hurt" framing.

What the large trials show#

Physicians' Health Study II (PHS-II). 14,641 male US doctors aged 50+, randomised to a standard multivitamin or placebo, followed for 11 years. Modest 8% reduction in total cancer (driven mostly by reductions in specific cancers); no effect on cardiovascular events. Small reductions in age-related cataracts.

COSMOS / COSMOS-Mind. 21,442 adults; multivitamin vs cocoa flavanols vs placebo. The multivitamin arm showed small but statistically significant improvements in global cognition over 3 years, particularly memory. Effect sizes small.

VITAL. Tested vitamin D and omega-3 specifically, not a full multivitamin; null on most hard endpoints.

Older "negative" trials (SELECT, ATBC) tested individual high-dose antioxidants and found no benefit (and in some cases harm). These aren't the same as standard multivitamins.

The pattern across modern multivitamin trials: small but real signals on some endpoints (cancer in PHS-II, cognition in COSMOS-Mind), null on cardiovascular events, no harm signal in standard-dose products.

Who's likely to benefit most#

The trial signals are largest in:

• Older adults (most of the positive trials are in 50+ populations)
• People with restricted diets (vegans, those with food intolerances, older adults with low appetite)
• People with malabsorption conditions (coeliac, IBD, post-bariatric surgery)
• Pregnant or planning pregnancy (specific folic acid + iodine + iron benefit)

For these populations, multivitamin supplementation has clearer rationale beyond the general trial signals.

Who's unlikely to benefit much#

• Healthy younger adults eating a varied diet meeting reference intakes
• People already taking multiple targeted supplements (risk of duplicate over-supplementation)
• People with specific high-quality nutrition habits already

Dose-response and safety#

Standard once-daily multivitamins (≤100% of RNI for most micronutrients) have a very strong safety record in trial data.

High-dose "athletic" or "premium" multivitamins providing many times the RNI can cause harm:

• Vitamin A toxicity (>10,000 IU/day chronic) — particularly important in pregnancy where teratogenicity is a concern
• Iron overload in supplementation without deficiency
• Selenium toxicity: narrow therapeutic window
• Niacin (B3) flushing and hepatotoxicity at very high doses
• Vitamin D toxicity at very high doses (>10,000 IU/day chronic)

For UK adults, sticking close to RNI (and trial-tested products like Centrum or Boots A–Z) avoids these issues.

What the marketing claims that isn't supported#

• "Energy boost." Not in trial data unless you were specifically deficient in a B vitamin
• "Immunity booster." Modest and only in deficiency contexts
• "Mental clarity." COSMOS-Mind found small cognitive signals but didn't show acute mood/clarity effects
• "Anti-ageing." No good evidence for anti-ageing effects in non-deficient adults

Pregnancy specifically#

UK NHS recommends:

• Folic acid 400 µg/day from preconception through first 12 weeks (higher in specific risk groups)
• Vitamin D 10 µg/day throughout pregnancy
• Iodine in some diets (vegan, fish-avoiding)
• Iron if deficient

A prenatal multivitamin covering these is reasonable. Most off-the-shelf prenatals do; check for folic acid dose and avoid those with vitamin A (retinol) above 700 µg/day.

What we'd actually say#

For healthy adults under 50 eating a varied diet, the trial evidence for daily multivitamin use is small and inconsistent. The cost is low and the safety is good; whether it adds up to a worthwhile intervention is uncertain.

For adults 50+, the modest cancer and cognitive signals in PHS-II and COSMOS-Mind are real and replicated enough to take seriously. A standard once-daily multivitamin is a reasonable choice.

For specific populations (pregnancy, restricted diets, malabsorption), targeted supplementation has clearer rationale.

A pragmatic position: standard once-daily multivitamins are low-risk; "premium high-dose" products often have less evidence and more potential for over-supplementation. Stick to RNI-aligned products.

When to talk to a clinician#

• Pregnancy or planning pregnancy
• Suspected nutritional deficiency
• Significant dietary restrictions
• Malabsorption conditions (coeliac, IBD, post-surgical)
• Already on prescribed supplements
• High-dose specialised supplements you've been taking long-term