COSMOS study: Multivitamins and epigenetic clocks in over-70s

Frequently cited results from a US study suggest that taking a daily multivitamin may slightly slow certain ageing processes in people over 70. Behind the headlines, though, the picture is more complicated: the effects are small, the clinical meaning is still uncertain, and the study funding includes notable industry involvement.

What the COSMOS project actually studied

The findings come from the large COSMOS programme, short for the “Cocoa Supplement and Multivitamin Outcomes Study”. Researchers at Brigham and Women’s Hospital in Boston, a Harvard-affiliated hospital, set out to test whether supplementing with multivitamins or cocoa extracts produces measurable changes related to ageing.

For the part of the trial now being discussed, the team enrolled just under 1,000 women and men, all aged at least 70. Participants were randomly allocated to one of four groups and, for two years, took daily:

• a standard multivitamin plus 500 mg cocoa extract,
• cocoa extract plus a placebo,
• the multivitamin plus a placebo,
• or two placebos.

The trial was double-blind: neither participants nor researchers knew who received which capsules. In medical research, this approach is widely regarded as the gold standard because it largely removes expectation effects.

Epigenetic clocks: how ageing was measured

Rather than simply counting calendar years, the researchers used so-called epigenetic clocks. The underlying idea is that biological ageing can be inferred from changes in the genome-more specifically, from chemical tags on DNA known as methylation.

In practical terms, the team analysed five different “epigenetic signatures”. These are based on methylation patterns at specific DNA sites and are treated as statistical markers for:

• biological age,
• the pace of ageing,
• the risk of premature mortality.

Blood samples were taken at the start of the study, after 12 months and after 24 months. This made it possible to see whether the epigenetic clocks advanced at different rates across the four groups.

"The key message: multivitamins seem to slightly slow some epigenetic ageing markers-especially in older adults who were already ‘older’ on these measures at baseline."

What the researchers actually found

The analysis indicated that participants who received the multivitamin (with or without cocoa) showed a small slowing of epigenetic ageing across all five measured signatures. The clearest differences appeared in two clocks that are closely linked to mortality.

Across the two-year period, the effect corresponds mathematically to a “gain” of around four months of biological age compared with the placebo group. In other words, based on these markers, those taking multivitamins looked biologically slightly younger than those taking dummy pills.

Notably, the apparent benefit was larger in people who, at the start, already showed accelerated epigenetic ageing. Those whose markers suggested they were “older than their actual age” seemed to gain a little more.

For cocoa extract itself, the study did not identify a persuasive advantage. Whether combined with the multivitamin or taken on its own, cocoa did not produce clear positive shifts in the epigenetic clocks.

A warning against inflated expectations

The authors themselves only partially share the media enthusiasm around “multivitamins slow ageing”. In Nature Medicine, they explicitly describe “statistically significant, but limited effects” and stress the need for much larger, longer-term research.

"The study does not show a miracle cure for wrinkles or frailty, but rather a small statistical effect in laboratory markers."

The biggest unresolved issue is whether these shifts in epigenetic clocks translate into noticeable health outcomes. Does a slightly slower epigenetic ageing trajectory reduce the risk of heart attack, dementia or cancer? Or is this mainly an intriguing lab readout with limited real-world impact?

Epigenetic clocks are not yet an everyday tool

Although many epigenetic markers are considered useful for predicting longevity in large population studies, they are not used routinely in clinical practice. The tests are complex, expensive and not fully validated for day-to-day medical decision-making.

On top of that, the biology behind epigenetic change is often unclear. Exactly how diet, inactivity, stress, medicines or supplements influence these clocks can currently only be explained in part.

Industry influence: who paid for the COSMOS project

Another aspect that draws scrutiny is funding. The study was not supported solely by public research money. Alongside US health agencies, several companies and industry associations from the food and supplement sectors contributed.

Supporters included, among others:

• the food group Mars (via its nutrition arm Mars Edge),
• Pfizer Consumer Healthcare, now known as Haleon,
• Foxo Technologies, which provides epigenetic testing,
• American Pistachio Growers, a trade body for the pistachio industry,
• the Council for Responsible Nutrition, a lobbying group for supplement manufacturers.

The researchers say they worked independently and experienced no interference from funders. Even so, an uncomfortable question remains: would the same results be interpreted more calmly without financial participation from the supplement industry?

Pill or plate - what matters more for ageing?

Howard Sesso, who led the study, points out that multivitamins are only a small piece of a much larger puzzle. Overall dietary habits and lifestyle are what matter most. A direct comparison between a daily pill and consistently healthy eating is still lacking.

A realistic order of priorities for everyday life might look like this:

• stopping smoking and keeping alcohol to a minimum,
• daily movement and strength training within your own capabilities,
• a high-fibre diet rich in vegetables, fruit, wholegrains and pulses,
• enough sleep and active stress reduction,
• only then considering whether a multivitamin supplement is additionally worthwhile.

Many professional bodies still advise healthy people to obtain vitamins and minerals primarily through a balanced diet. Supplements can help where there are clear deficiencies or particular situations-such as vitamin D deficiency, pregnancy, or after certain operations. A blanket “fountain of youth” effect from multivitamins cannot be derived from the COSMOS study.

What older adults can take from the study

If you already take a multivitamin, these new data are not a reason to celebrate the tub-or to throw it in the bin in disgust. The findings are better treated as a matter for sober interpretation:

• There are signals of a slight slowing in certain ageing markers.
• The clinical relevance is still uncertain.
• More research with independent funding is needed.
• A healthy daily routine is likely to have much stronger effects on lifespan and quality of life than any pill.

Anyone considering starting a multivitamin should ideally discuss it with their GP-especially in later life, when multiple medicines are often taken in parallel. Some vitamins can be harmful at high doses or interact with prescribed drugs.

What a more realistic “anti-ageing check” at the GP might look like

Instead of relying on costly laboratory epigenetic tests, it is often more useful to begin with classic health factors. Many practices use straightforward, well-established measures:

Factor	Why it matters
Blood pressure	High readings damage blood vessels and raise the risk of heart attack and stroke.
Blood lipids & blood glucose	Provide insight into metabolism and the risks of diabetes and heart disease.
Body weight & muscle strength	Reflect activity patterns and fall risk in older age.
Diet and activity profile	Highlights areas that can make far more difference than a supplement.

Being in good shape on these fundamentals can significantly reduce personal disease risk-regardless of whether an epigenetic clock in a lab appears to tick a few months more slowly.

Where multivitamin-and-ageing research still needs to go

Future studies will likely need to link several layers of evidence: epigenetic markers, conventional clinical endpoints such as heart attacks or dementia, and lived quality of life. Only if changes in epigenetic clocks reliably align with fewer illnesses and more healthy years will such markers justify wider practical use.

Another open question is the role of combined effects: what happens to epigenetic clocks when someone improves diet, moves more, reduces stress and possibly takes a multivitamin at the same time? Trials focused on single pills offer only a very small slice of that bigger picture.