A standard daily multivitamin, the kind sold at every pharmacy for roughly a dime a dose, slowed biological aging across every molecular measure tested in a two-year randomized clinical trial of nearly 1,000 older adults. The results, drawn from a prespecified ancillary study of the large COSMOS prevention trial and published in Nature Medicine, showed that participants randomized to Centrum Silver aged more slowly than those on placebo according to all five DNA-methylation clocks researchers tracked. The benefit was not evenly distributed: older adults whose cells already looked older than their birth certificates at the start of the study experienced the most pronounced slowing.
Five clocks, one consistent signal
The epigenetic analysis was embedded inside the COcoa Supplement and Multivitamin Outcomes Study (COSMOS), a registered, placebo-controlled trial that enrolled thousands of adults aged 60 and older. For the epigenetic substudy, 958 participants provided blood samples at enrollment and again two years later. Researchers extracted DNA-methylation data and ran it through five validated aging clocks, each built from a different set of chemical tags on DNA that shift predictably as people get older.
Four of those clocks, known as PCHannum, PCHorvath, PCPhenoAge, and PCGrimAge, estimate biological age: how old a person’s cells appear relative to their calendar age. The fifth, DunedinPACE, works differently. Instead of calculating a static age, it captures the speed of aging at the moment blood is drawn, functioning like a speedometer rather than an odometer. That the multivitamin group showed favorable shifts on both types of measurement is notable, because slowing the ongoing rate of molecular damage and dialing back accumulated biological age are distinct biological achievements.
GrimAge, the clock behind the PCGrimAge measure used here, deserves particular attention. Developed by Steve Horvath’s lab and validated in multiple cohorts, GrimAge was specifically engineered to predict mortality and time to disease onset, and it outperforms earlier-generation clocks on both counts. A slowing signal on GrimAge carries more clinical weight than a shift on clocks designed primarily to estimate chronological age, because its methylation signatures map more directly onto real-world health trajectories.
Who benefited most
The most striking subgroup finding involved participants who entered the trial with accelerated biological aging, meaning their epigenetic clocks already read older than their actual age at baseline. In this group, the multivitamin produced a larger reduction in the pace of aging than it did in participants whose biological and chronological ages were closely matched at enrollment.
This pattern makes biological sense. If part of the multivitamin’s effect comes from correcting marginal nutrient shortfalls that contribute to cellular stress and DNA-methylation drift, then people with the most drift at baseline have the most room for correction. It also aligns with a broader principle in preventive medicine: interventions tend to show the greatest absolute benefit in people at highest baseline risk.
Still, the finding raises a question the trial was not designed to answer definitively. Were the biologically older participants also more likely to have low levels of specific micronutrients such as vitamin D, B12, folate, or zinc? The published analysis did not stratify epigenetic outcomes by pre-randomization blood nutrient panels or dietary intake, so it remains unclear whether the multivitamin is primarily filling nutritional gaps or exerting a broader pharmacological effect on methylation machinery. That distinction matters enormously for public health guidance, because recommending routine supplementation for well-nourished adults is a different proposition from targeting people with measurable deficits.
What the trial can and cannot tell us
Several features of the study design strengthen confidence in the results. COSMOS is a registered, randomized, placebo-controlled trial, placing it near the top of the evidence hierarchy for preventive interventions. Participants and investigators were blinded, and the epigenetic analysis was prespecified rather than added after data collection, which limits the risk of cherry-picking favorable biomarkers from a larger pool. The convergence of results across five independently validated clocks, each built on different sets of CpG sites, makes it harder to dismiss the outcome as a statistical fluke tied to any single measure.
The intervention itself was deliberately ordinary. Centrum Silver contains a broad mix of vitamins and minerals at doses near recommended daily values. This is not a megadose regimen or a boutique longevity stack. The placebo was an identical-looking inert pill, and the two-year window gave researchers enough time to detect methylation shifts that typically accumulate slowly.
But the study has clear boundaries. Epigenetic clocks are surrogate endpoints. Slowing a methylation-based clock does not automatically mean a person will avoid heart disease, cancer, or dementia, or live longer. These clocks were built because they correlate with such outcomes in large observational datasets, but the history of medicine is full of cases where improving a surrogate marker through an intervention failed to produce proportional clinical gains. To close that gap, researchers will need to link clock changes directly to hard outcomes within COSMOS or similar trials, asking whether participants who showed the greatest slowing in DunedinPACE or GrimAge also experienced fewer cardiovascular events, cancers, or deaths over longer follow-up.
The two-year intervention window also leaves durability unanswered. Epigenetic clocks can shift in both directions, and no published data yet show whether the slowing persists after supplementation stops, continues to accumulate with longer use, or plateaus once underlying nutrient gaps are corrected. Only extended follow-up, ideally including a washout period, can resolve that question.
Generalizability is another consideration. COSMOS participants were predominantly white, with a mean age around 73, and were healthy enough and motivated enough to enroll in a prevention trial requiring daily pill-taking and periodic follow-up. Whether similar epigenetic benefits would appear in more diverse populations, or in people managing multiple chronic conditions or frailty, has not been tested.
Where this fits in the larger COSMOS picture
The epigenetic findings do not exist in isolation. Earlier analyses from the same parent trial reported that daily multivitamin use was associated with modest improvements in cognitive function, including memory, in older adults. Those cognitive results, published separately, generated significant public interest and debate. The epigenetic data now offer a potential biological mechanism: if the multivitamin is slowing molecular aging processes in blood, it is plausible that similar effects occur in brain tissue, though that connection remains speculative without direct measurement.
Funding and conflict-of-interest disclosures are worth noting. Centrum Silver is manufactured by Haleon (formerly part of GSK’s consumer health division), and the degree of industry involvement in supplying study product, underwriting costs, or shaping analysis plans is a standard concern in supplement research. The Nature Medicine paper includes formal disclosures, but readers evaluating the findings should weigh the potential for commercial interest alongside the study’s methodological strengths.
What this means for people weighing a daily multivitamin
As of July 2026, the practical picture is cautiously encouraging but incomplete. A daily multivitamin is inexpensive, widely available, and generally safe for most older adults. The COSMOS epigenetic data suggest it may offer a small but measurable benefit at the level of biological aging, particularly for people whose molecular age already outpaces their calendar age. That is a meaningful addition to the evidence base, because few interventions this accessible have shown consistent effects across multiple validated aging biomarkers in a rigorous trial design.
But the magnitude, durability, and clinical significance of the benefit remain open questions. A multivitamin is not a substitute for physical activity, adequate sleep, blood pressure management, smoking cessation, or a balanced diet. Those pillars of healthy aging rest on decades of outcome data that no single supplement trial can match. What the COSMOS findings do suggest is that basic micronutrient support may play a role in the biology of aging that previous research underestimated, and that the people most likely to benefit may be those whose bodies are already showing signs of accelerated wear.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
