Summary: For most healthy adults under 65, a daily multivitamin does not extend your life, prevent heart attacks, or do much beyond filling minor nutritional gaps you could close with a better diet. The exception: if you're over 65, evidence from three separate clinical trials consistently shows cog
Think of your cells like factories that each need specific raw materials to run their machines. Once a factory has enough zinc to run its enzymes, sending more zinc doesn't speed anything up — the excess just piles up in a loading bay or gets excreted. But if a factory is genuinely short on a material, restocking it gets the machines running at full capacity again — and for adults over 65, the body's supply chain for several key nutrients genuinely starts running low.
That's the general answer. Your stack is different.
Check your whole stackSupplement Engine
Insurance policy or waste of money?
Conditional Vitamins & Minerals / Combination Micronutrient SupplementThe Plain English Version
The world's most popular supplement protects your brain at 65+ — but does nothing for lifespan in healthy adults.
Think of your cells like factories that each need specific raw materials to run their machines. Once a factory has enough zinc to run its enzymes, sending more zinc doesn't speed anything up — the excess just piles up in a loading bay or gets shipped out. But if a factory is genuinely short on a material, restocking it gets the machines running at full capacity again — and for adults over 65, the body's supply chain for several key nutrients genuinely starts running low.
Want the full evidence? Keep scrolling
What People Claim
Multivitamins are sold on the premise that they act as a nutritional "insurance policy" — a catch-all solution to bridge dietary gaps, boost energy, sharpen immunity, and offset the health risks of an imperfect diet.
"Most people are deficient in key nutrients. A daily multivitamin closes those gaps and keeps you protected."
"Taking a multivitamin every day reduces your cancer risk, protects your heart, and slows cognitive decline as you age."
"New research shows multivitamins can actually make you biologically younger — slowing the aging process at the cellular level."
The bigger claims frame multivitamins as longevity tools. The newest frontier involves epigenetic aging — emerging research suggests multivitamins may decelerate biological clocks, making users biologically younger over time. These claims are newer, more mechanistically plausible, and critically, supported by actual randomised controlled trial data rather than just observational epidemiology.
What the Evidence Shows
| Claimed Benefit | Verdict | Key Study | Effect |
|---|---|---|---|
| All-cause mortality reduction DEBUNKED | Null — no benefit | Loftfield 2024 (N=390,124) | HR 1.04 (slightly worse in early follow-up) |
| CVD prevention DEBUNKED | Null | PHS II (N=14,641, 11.2yr) | No reduction in cardiac events |
| Cancer incidence reduction MODERATE | Conditional (older males only) | Manson/PHS II 2012 | HR 0.92 — 8% reduction in older men |
| Cognitive function in 65+ STRONG | Works for 65+ | COSMOS-Web 2023 (N=3,562) | Equivalent to 3.1 years preserved memory |
| Epigenetic aging attenuation EMERGING | Promising — needs replication | Sesso COSMOS 2026 (N=958) | PCGrimAge slowed 1.4 months/yr; PCPhenoAge 2.6 months/yr |
| Micronutrient deficiency correction STRONG | Definitively works | Blumberg 2017 (NHANES N=10,698) | Virtually eliminates inadequacy for Vit D, E, C, A, Zinc |
| Energy / general wellness WEAK | Unproven | — | No RCT data in non-deficient populations |
The largest study ever conducted on this question — Loftfield 2024, following 390,124 people for over 20 years across three US cohorts — found no mortality benefit from multivitamin use. There was a slight 4% early hazard signal, likely explained by "sick-quitter" bias: people often start taking multivitamins precisely when their health is already declining.
What would change this: A double-blind RCT with randomised multivitamin assignment (not self-selection) in a population confirmed deficient at baseline would test whether the null result holds when selection bias is eliminated.
The Physicians Health Study II — 14,641 men followed for 11.2 years in a double-blind RCT — showed no reduction in cardiovascular events, strokes, or cardiac mortality. The US Preventive Services Task Force issued an I-statement (insufficient evidence to recommend) in 2022.
What would change this: An RCT specifically enrolling adults with confirmed low baseline micronutrient status, measuring endothelial function and cardiovascular endpoints over 10+ years.
PHS II found a statistically significant 8% reduction in cancer incidence in older male physicians (HR 0.92, p=0.044). This is a genuine signal from a rigorous double-blind RCT. However, it has not been replicated in women or in younger or more diverse populations, which is why the USPSTF couldn't make a broader recommendation.
What would change this: Replication in a diverse, mixed-sex population would elevate this to STRONG. Failure to replicate would weaken it to LOW.
This is the headline finding — and it's real. Three independent arms of the COSMOS trial (COSMOS-Web 2023, COSMOS-Clinic, COSMOS-Mind) have now consistently shown that adults over 65 taking a standard daily multivitamin preserved memory function equivalent to having a brain 1.8–3.1 years younger. Three independent arms converging on the same signal is strong evidence.
What would change this: Failure to replicate in a non-US population, or evidence that the benefit is driven by correcting specific baseline deficiencies rather than the multivitamin itself.
The COSMOS 2026 trial (Nature Medicine) found that Centrum Silver slowed biological aging as measured by two epigenetic clocks: PCGrimAge slowed by 1.4 months per year, and PCPhenoAge by 2.6 months per year over a 2-year RCT in adults over 60. This is mechanistically plausible — the triage theory of micronutrients predicts exactly this. But it's a single trial and needs independent replication before confidence is high.
What would change this: Independent replication in a larger, more diverse cohort would elevate this to STRONG. The mechanistic story is already compelling.
NHANES biochemical data on 10,698 Americans shows that multivitamin use virtually eliminates blood-level inadequacy for Vitamin D (94% of Americans have suboptimal status), Vitamin E, Vitamin C, Vitamin A, and Zinc. This isn't a health outcome — it's confirming the supplement does what it says on the label. Whether correcting those deficiencies translates to downstream health outcomes is the more contested question.
What would change this: Nothing — the biochemistry of deficiency correction is not in dispute. The question is what those corrections mean for health outcomes.
There is no randomised controlled trial showing that well-nourished, non-deficient adults experience any measurable improvement in energy, mood, or general wellbeing from taking a multivitamin. The threshold model of micronutrient function explains why: if you're already at saturation, topping off doesn't do anything.
What would change this: An RCT specifically in adults with confirmed adequate baseline status showing a subjective or objective wellness improvement would be a genuine surprise.
How It Works
Most people assume that more nutrients = more benefit. Micronutrients don't work that way. They operate on a threshold model: up to the point of cellular saturation, each nutrient performs its biochemical job — acting as an enzyme cofactor, antioxidant, hormone precursor, DNA repair signal, or immune activator. Once the threshold is met, additional intake provides no extra benefit. For water-soluble vitamins, excess is simply excreted. For fat-soluble vitamins (A, D, E, K) and trace minerals, excess accumulates and can eventually cause toxicity.
In well-nourished populations — like the male physicians in PHS II — most micronutrient "tanks" are already full. The supplement tops them off past saturation and adds nothing. The body's biochemical factories are already running at capacity. More raw materials doesn't increase output when you're already at full capacity.
When nutrients are scarce, the body makes a triage decision: prioritise acute survival functions (immune response, energy metabolism) over long-term maintenance (DNA repair, oxidative damage regulation). This chronic low-level shortfall — subclinical deficiency that doesn't cause obvious symptoms — accumulates as accelerated biological aging. This is the mechanistic backbone of the COSMOS epigenetic data: resupplying the maintenance pathways slows how fast the biological clock ticks.
As we age, the body's ability to absorb and utilise key micronutrients declines. Intrinsic factor production decreases (reducing B12 absorption from food), vitamin D conversion in the skin becomes less efficient, and overall gut absorption capacity can fall. For this age group, the supplement is genuinely correcting real biochemical gaps — not topping off an already-full tank.
The Debate
Loftfield 2024 — N=390,124 — 20yr follow-up
No mortality benefit; slight 4% increase in early hazard among multivitamin users.
Sesso COSMOS 2026 — N=958 — 2yr RCT
Epigenetic aging slowed by 2.7–5.1 months in a double-blind randomised controlled trial measuring molecular biomarkers.
Why they disagree: Loftfield is observational — "sick-quitter" bias means people often start multivitamins when health is already declining, confounding the analysis. COSMOS is a double-blind RCT measuring molecular biomarkers, immune from selection bias. They also measure different outcomes at different timepoints. These studies aren't in direct conflict — they're measuring different things.
Grodstein 2013 (PHS II) — N=5,947 — 12yr RCT
No cognitive benefit in male physicians over 12 years of follow-up, measured via telephone cognitive tests.
COSMOS-Web/Mind 2023/24 — N=3,562–573 — 3yr RCT
Consistent cognitive protection across three independent arms — 1.8–3.1 years of preserved memory function vs placebo.
Why they disagree: PHS II used an older Centrum Silver formulation with lower vitamin D (400 IU) and no lutein or zeaxanthin; cognition was measured via telephone — a relatively crude tool. COSMOS used an updated formulation and sensitive web-based neuropsychological tests (ModRey). The measurement tool and formulation differences likely drove the discrepancy as much as any true biological difference.
USPSTF 2022 — Systematic review
Issued an I-statement (insufficient evidence): cannot recommend MVMS for cancer or CVD prevention across the general population.
Manson/PHS II 2012 — N=14,641 — Double-blind RCT
8% statistically significant reduction in cancer incidence in older male physicians (HR 0.92, p=0.044).
Why they disagree: The USPSTF requires replication across diverse populations before making a public health recommendation. One double-blind RCT in exclusively male physicians cannot anchor a recommendation for all adults. No equivalent RCT has been conducted in women or diverse demographics — which is the gap, not a contradiction of the finding itself.
Current direction: The emerging consensus is population-segmented rather than all-or-nothing. Evidence is converging that adults 65+ benefit cognitively, deficient populations benefit from micronutrient correction, and the epigenetic aging data is becoming harder to dismiss. The null result for longevity in healthy well-nourished adults remains robust.
Real World vs Lab
Exactly How to Use It
| Population | Dose & Key Nutrients | Timing | Form |
|---|---|---|---|
| General adults under 40 | 100% RDA across B-complex, C, D3 (1000 IU — one standard capsule), Zinc; avoid iron unless menstruating | With morning meal containing fat | Tablet or capsule |
| Adults 40–65 | Add B12 50mcg, higher D3 (2000 IU — two standard capsules), K2 (MK-7); full B-complex | Split AM/PM dose if including minerals | Capsule |
| Older adults 65+ | High B12, D3 2000 IU+, Lutein/Zeaxanthin, balanced B-complex — the actual trial-tested formulation (Centrum Silver or equivalent) | With largest meal | Tablet (Centrum Silver was used in COSMOS) |
| Vegans / Vegetarians | B12, Iron (with Vitamin C), D3 (lichen-derived), Zinc; fills systematic dietary gaps | Iron away from tea and coffee | Capsule |
| Athletes | B-complex, Magnesium, Zinc; avoid high-dose antioxidants (C, E blunt training adaptations) | Away from training window | Capsule |
Safety & Interactions
Vitamin K directly antagonises warfarin's blood-thinning mechanism. Fluctuating vitamin K intake from multivitamins can destabilise your INR (blood clotting measure). Action: maintain stable vitamin K intake; inform your prescriber; monitor INR regularly.
Divalent minerals (calcium, iron, magnesium, zinc) in multivitamins bind to these antibiotics in the gut, drastically reducing how much antibiotic actually gets absorbed. Action: space multivitamin by at least 2 hours from antibiotic dose.
Iron and calcium in multivitamins bind to levothyroxine in the gut, reducing how much thyroid medication is absorbed. Action: space levothyroxine dose by 2–4 hours from multivitamin.
Vitamin B6 (pyridoxine) enhances peripheral breakdown of levodopa, preventing the drug from crossing into the brain where it's needed. Action: avoid high B6 if on levodopa alone; carbidopa co-formulation mitigates this interaction.
Long-term acid suppression reduces absorption of B12, iron, magnesium, and calcium from food and supplements. Action: monitor B12 and magnesium levels on chronic PPI use; consider higher B12 dosing.
Calcium competes for the same intestinal transporter as iron, reducing non-heme iron absorption by up to 60% when taken together. Action: do not co-administer; space by at least 2 hours.
Men and postmenopausal women generally do not need supplemental iron and should choose iron-free formulations. Individuals with hemochromatosis (an iron overload condition) must strictly avoid iron-containing products — iron in standard multivitamins bypasses normal absorptive regulation and accelerates cardiac and liver iron deposition.
Smokers or individuals with asbestos exposure: beta-carotene at supplemental doses paradoxically increases lung cancer incidence via pro-oxidant mechanisms in lung tissue (CARET trial). The USPSTF specifically recommends against beta-carotene supplementation in these populations.
Early pregnancy: Vitamin A in retinol form is highly harmful to the developing baby above 3,000 mcg RAE. Beta-carotene is safe. Always use pregnancy-specific prenatal formulas — they use beta-carotene, not retinol, for Vitamin A.
| Side Effect | How Common | Cause | Fix |
|---|---|---|---|
| Nausea / GI upset | 5–10% | Iron and Zinc on an empty stomach | Always take with food; switch to iron-free if not deficient |
| Constipation / dark stools | Common | Iron | Increase hydration; switch to iron-free if not needed |
| Headache / flushing | Occasional | High-dose B3 (niacin) | Look for inositol hexanicotinate form, which reduces flushing |
| Metallic taste | Occasional | Iron, Zinc | Typically resolves; try split dosing with meals |
The Nuance
The "just take a multivitamin" vs "multivitamins are useless" debate both miss the same thing: context is everything. The supplement is not the variable — the nutritional baseline of the person taking it is the variable.
At £5–8/month for a standard tablet (the type actually used in clinical trials), multivitamins are one of the cheapest interventions available when the evidence supports their use. For the populations that benefit — 65+, deficient individuals, vegans, pregnant women — the value-to-cost ratio is strong.
For healthy adults under 65 with a solid diet seeking specific outcomes: targeted individual supplements are more cost-effective and precise. D3 (2000 IU) + Magnesium (300–400mg) + B12 (if dietary intake is low) costs approximately £10–15/month and delivers therapeutic doses where a multivitamin's quantities are often pharmacologically underpowered.
Overall Conviction
MODERATEThe cognitive protection in 65+ and micronutrient deficiency correction are HIGH conviction. The cancer signal in older men is MODERATE. The epigenetic aging data is genuinely exciting but needs replication. For the populations that matter — 65+, deficient, vegan, pregnant — the evidence is clear.
Sources
1. Loftfield E et al. (2024) — JAMA Network Open — N=390,124
Multivitamin Use and Mortality Risk in 3 Prospective US Cohorts. Key finding: No mortality benefit; HR 1.04 in early follow-up. Largest observational study to date.
2. Sesso HD et al. (2026) — Nature Medicine — N=958
Multivitamin Use and Epigenetic Age Attenuation in COSMOS. Key finding: PCGrimAge slowed 1.4 months/yr; PCPhenoAge slowed 2.6 months/yr over 2-year double-blind RCT.
3. Yeung LK et al. (2023) — American Journal of Clinical Nutrition — N=3,562
Multivitamin Supplementation Improves Memory in Older Adults (COSMOS-Web). Key finding: Memory improvement equivalent to 3.1 years of age-related recovery vs placebo.
4. Vyas CM et al. (2024) — American Journal of Clinical Nutrition — N=573
Cognitive effects of multivitamin supplementation in older adults (COSMOS-Clinic). Key finding: Improved global cognition and episodic memory vs placebo.
5. Manson JE et al. (2012) — JAMA — N=14,641 (PHS II)
Multivitamins in the Prevention of Cancer in Men. Key finding: 8% reduction in cancer incidence (HR 0.92, p=0.044) in older male physicians. Null for CVD and mortality.
6. Blumberg JB et al. (2017) — Nutrients — N=10,698 (NHANES)
Contribution of Multivitamin Supplements to Micronutrient Intakes. Key finding: Virtually eliminates inadequacy for Vitamin D, E, C, A, and Zinc at the population level.
7. Grodstein F et al. (2013) — Annals of Internal Medicine — N=5,947 (PHS II)
Long-term Multivitamin Supplementation and Cognitive Function in Men. Key finding: No cognitive benefit in male physicians over 12 years of follow-up.
8. U.S. Preventive Services Task Force (2022)
Vitamin Supplementation to Prevent CVD and Cancer. Key finding: I-statement (insufficient evidence) for multivitamin supplementation to prevent CVD or cancer across the general adult population.
How strong is the evidence for the claims in this review? Higher = more confidence the claims are supported. This does not measure how large the effect is or how important it is compared with other levers.
Is this worth your time, money, effort, risk, and trust for this goal? Different from Verdict Score (evidence strength) and Leverage Map (relative importance) — Action ROI is the worth-it call once friction is priced in.
Evidence-scored dosing, timing, forms, and who should skip it. One page, no fluff.
Get the protocolConviction-scored verdicts on supplements, nutrition, training, physio, and recovery.