COSMOS Study: Can Multivitamins Slow Biological Ageing After 70?

A new study has made headlines, but the reality is considerably more complicated.

Across much of the international media, multivitamin supplements are suddenly being portrayed as a kind of anti-ageing secret weapon for people over 70. The trigger is a recent paper in Nature Medicine, which suggests that certain multivitamins may measurably slow biological ageing. Yet the story behind the sober figures looks far less dramatic on closer inspection, especially once diet, lifestyle and funding sources are taken into account.

What the COSMOS study actually examined

The research is called COSMOS, short for the “Cocoa Supplement and Multivitamin Outcomes Study”. It was conducted by Brigham and Women’s Hospital in Boston, a respected research institution, and coordinated by preventive medicine specialist Howard Sesso.

Around 1,000 people took part, split roughly evenly between women and men. Every participant was at least 70 years old. Researchers randomly assigned them to one of four groups, each of which took a different daily supplement for two years:

• a commercially available multivitamin product plus 500 milligrams of cocoa, containing around 50 milligrams of epicatechin, a well-known antioxidant
• cocoa alone plus a placebo
• a multivitamin plus a placebo
• a placebo only

The central question was whether the genetic material could reveal if multivitamins or cocoa slowed ageing.

Epigenetic clocks: what they measure

Rather than assessing wrinkles or physical fitness, the researchers examined what are known as “epigenetic clocks”. These are patterns of DNA methylation: small chemical markers on genetic material that change with age and can provide clues about biological ageing.

During the study, the teams assessed five such epigenetic signatures at three points: at the outset, after 12 months and after 24 months. Some of these markers are regarded as being particularly closely associated with mortality and illness in later life, although the methods have not yet been fully established.

The paper’s main message is this: in the multivitamin group, the “epigenetic clock” slowed slightly - a measurable but small effect.

Among participants taking multivitamins, all five signatures assessed moved in the direction of “slower ageing”. Two epigenetic clocks with particularly strong links to mortality showed a notable response.

Four months of “biological advantage” sounds bigger than it is

To make the findings easier to interpret, the researchers converted the effect into time. Their result was that, over the two-year study period, multivitamin users had a biological “advantage” of about four months compared with the placebo-only group.

It is important to note the following:

• This concerns biological age as measured by an epigenetic clock, rather than additional calendar years of life.
• The results remain statistically significant, but their size is modest, meaning limited.
• The cocoa supplement showed no meaningful benefit for these markers.

The clearest benefit appeared in the group that already had accelerated epigenetic ageing at the beginning of the study. For these people, the clock seemed to slow somewhat more. That is plausible: those who are biologically “older” than expected may have greater scope for improvement.

The researchers themselves urge caution

While headlines proclaim that a “multivitamin stops ageing”, the authors’ conclusions are markedly more restrained. They stress that the effects observed are limited and that many questions remain unanswered.

The data are “encouraging”, but only a starting point, the researchers write. Whether these changes actually translate into noticeable health benefits in everyday life remains uncertain.

In an accompanying contribution in Nature, Howard Sesso explicitly highlights how closely the results are tied to diet and lifestyle. In essence, he says that a better understanding is needed of how the overall combination of eating habits, nutrient intake and everyday behaviours affects changes in epigenetic clocks.

One key issue is that COSMOS compares multivitamins with a placebo, not with an optimal diet. Whether a varied, plant-focused way of eating could have comparable or even stronger effects than a pill remains unknown - and it is precisely the sort of comparison many experts would like to see.

How reliable are epigenetic clocks?

Epigenetic markers are currently seen as promising tools in ageing research, but they are not yet the gold standard.

Several limitations remain:

• The methods are not fully validated for use in routine healthcare.
• The precise relationship between epigenetic age and specific diseases in old age is still incompletely understood.
• It remains unclear to what extent diet, exercise, medicines and supplements each influence these clocks.

Sesso himself also points out that too many details are still missing to bridge the gap between these laboratory values and real disease outcomes. In practical terms, a favourable epigenetic reading may sound reassuring, but it does not reliably predict whether someone will actually live longer or remain healthier.

Funding in the background: multivitamins are a multibillion-pound business

The issue of sponsors is particularly sensitive in this field. Multivitamin supplements and functional foods generate enormous revenues worldwide. Companies producing such products have a strong interest in positive studies that attract media attention.

The COSMOS paper lists public bodies such as the National Institutes of Health among its funders, alongside prominent companies and industry organisations, including:

• food company Mars, through its Mars Edge nutrition products division
• Pfizer Consumer Healthcare, now Haleon, a provider of over-the-counter products
• Foxo Technologies for the epigenetic tests
• American Pistachio Growers, an association of pistachio producers
• the Council for Responsible Nutrition, a lobbying organisation for the dietary supplements industry

The researchers state that they have no conflicts of interest and that their interpretation of the findings remains independent. Critics, however, point out that studies without direct links to manufacturers are generally easier to communicate to the public, simply because there is less suspicion of hidden influence.

What does this mean for people over 70?

Many readers will now wonder whether, at 70 or older, they should immediately start taking a multivitamin product. The COSMOS data offer only limited grounds for doing so.

A few points help put the findings in context:

• Over two years, the average observed benefit was around four months of biological ageing - measurable, but not dramatic.
• The study focused on people over 70, so its findings can barely be applied to younger people.
• Clinical outcomes such as heart attacks, cancer or the need for care are not the main focus here; the study is primarily about laboratory measures.

Someone whose diet is already highly unbalanced may benefit more from supplementation than a person who eats vegetables, wholegrains, pulses and enough protein every day. Ultimately, the study also supports a simple but robust principle: where genuine nutrient gaps exist, supplements may have an opportunity to improve something.

Multivitamin or a plate full of vegetables?

A direct contest would be particularly interesting: a multivitamin supplement versus a deliberately improved diet. Sesso identifies exactly this scenario as an important topic for future research. So far, COSMOS has compared the tablet only with a placebo, not with an “ideal” meal plan.

For everyday life, this means that improving one’s diet over the long term affects not only possible epigenetic markers but several areas of health at once: weight, blood sugar, blood pressure and the gut microbiome. A single pill is unlikely to replace all of that completely, even if it produces a small advantage on paper in one biological clock.

Practical advice: who may benefit and who should be careful?

Multivitamin supplements are generally considered relatively safe for healthy older adults, but they are not harmless substitutes for sweets. Problems can arise, particularly with high doses of certain vitamins or trace elements.

Typical examples of potential risks include:

• Too much vitamin A can place strain on the liver and weaken bones in older people.
• Excess vitamin B6 can damage nerves.
• High doses of antioxidants may, in individual cases, affect treatments such as cancer therapy.

It makes sense to speak to a GP, especially if several medicines are already being taken. Blood tests for important nutrients such as vitamin D, vitamin B12 or iron can also help people supplement more precisely rather than swallowing a broad combination product “just in case”.

How lifestyle and epigenetics influence one another

The COSMOS findings draw attention to an intriguing point: in a sense, our lifestyle “writes” on our genetic material without altering the DNA itself. Diet, exercise, sleep, stress and environmental toxins leave traces in the form of methyl groups and other modifications.

This leaves room for change. People who adjust their habits later in life may be able to shift some of these traces. Smaller studies have shown, for example, that:

• regular physical activity affects epigenetic patterns in muscle and fat cells
• smoking creates characteristic epigenetic signatures - and stopping smoking gradually weakens them again
• a strongly plant-based diet brings changes in genes linked to inflammation and metabolism

Viewed in this context, a multivitamin’s contribution to the epigenetic clock looks more like fine-tuning. The major levers remain exercise, diet, avoiding tobacco and maintaining a reasonably stable body weight.

Ultimately, the COSMOS study indicates that multivitamins can slightly shift certain biological ageing markers in older people. They will not, however, turn a 75-year-old into a 60-year-old. Anyone seeking to age more slowly over the long term cannot avoid the established foundations: a varied diet, regular activity and good medical monitoring, with or without an additional pill.