The New Ageing Clocks: What GrimAge, DunedinPACE and OMICmAge Actually Measure — and Which Foods Move Them

If you have spent any time reading about biological age in the last two years, you will have encountered the names. GrimAge. DunedinPACE. OMICmAge. The Horvath clock. PhenoAge. They appear in longevity podcasts, research summaries, and the dashboards of commercial testing companies, often without much explanation of what they are actually measuring — or whether food can meaningfully change them.

This piece is an attempt to answer that honestly. What each clock measures, where the diet evidence is strongest, where it is weakest, and what a woman in her forties or fifties can reasonably do with this information.

What these clocks actually are

Epigenetic clocks measure biological age by reading patterns of DNA methylation — chemical tags on your DNA that change predictably as you age. The clocks are trained on large populations to identify which methylation patterns best predict a given outcome: how old you are chronologically, how quickly your body is ageing, or how likely you are to develop age-related disease.

They are not all measuring the same thing, and this distinction matters enormously when you are trying to understand what diet can do.

GrimAge — the mortality predictor

GrimAge (and its updated version GrimAge2) was trained to predict lifespan and healthspan — specifically, how close a person is to death from age-related disease. It incorporates plasma protein levels alongside methylation data, which makes it more biologically comprehensive than earlier clocks. A higher GrimAge score means a higher predicted mortality risk relative to your chronological age.

GrimAge is not a speedometer. It is more like a snapshot of where you currently sit on a risk curve. It can shift with dietary intervention — clinical evidence suggests it does — but it tends to move more slowly than pace-of-ageing clocks, and changes require sustained behaviour over months.

DunedinPACE — the speedometer

DunedinPACE is different in design and purpose. It was developed by researchers at Duke University using 20 years of longitudinal data from the Dunedin Birth Cohort — tracking 19 biomarkers of system integrity across two decades to create a measure of how fast someone is currently ageing, expressed as a decimal. A score of 1.0 means you are ageing at exactly the average rate for your age. A score of 1.15 means you are ageing 15% faster. A score of 0.85 means 15% slower.

This is the clock that is most responsive to dietary change, and the one with the strongest and most consistent intervention evidence. The Moffitt-Caspi lab at Duke describes it plainly: DunedinPACE is “the most sensitive to intervention to date, slows most in intervention trials, and is the most consistently responsive across studies.” Read the Duke lab’s summary here.

OMICmAge and multimodal clocks — the new frontier

A landmark 2026 paper in Cell by Li J et al. — “Multimodal clocks of human aging” — took the clock concept significantly further. Rather than using DNA methylation alone, the researchers built clocks from multiple data layers simultaneously: genomics, proteomics, metabolomics, microbiome data, and clinical measurements. The result was a multimodal clock that predicts chronological age with a mean absolute error of just 3.87 years, outperforming all single-omics clocks that preceded it. Read the Cell paper here.

The significance for nutrition is this: the multimodal framework confirmed that different biological systems age at different rates within the same person, and that the gut microbiome is one of the most informative data layers for predicting overall ageing trajectory. As explored in the estrobolome post, the gut is not a passive digestive system — it is an active participant in how fast every other system ages.

What the diet evidence actually shows

The honest answer is: diet moves these clocks, but the effect sizes are modest, the time required is longer than most people expect, and not every clock responds equally to the same intervention. Here is what the strongest trials have found.

The CALERIE trial — caloric restriction and DunedinPACE

The Comprehensive Assessment of Long-term Effects of Reducing Intake of Energy (CALERIE) trial is the most rigorous diet intervention trial to use epigenetic clocks as an outcome. 220 adults without obesity either restricted their caloric intake by 25% or ate freely for two years. The caloric restriction group showed a significant slowing in DunedinPACE — approximately 0.02 units — while showing no significant change in second-generation clocks including GrimAge.

This finding has two important implications. First, DunedinPACE is genuinely responsive to dietary intervention — this is not a theoretical possibility but a measured outcome in a two-year randomised trial. Second, the mechanism appears to be related to reducing the pace of cellular damage accumulation rather than reversing accumulated risk, which is why the mortality-prediction clocks like GrimAge did not shift significantly over the same period.

The methylation diet — GrimAge and methyl adaptogens

An eight-week randomised controlled trial published in Aging — and reviewed more recently in a 2025 Aging-US paper — tested a diet specifically designed around methyl donors and methyl adaptogen foods. Participants who followed the protocol showed a mean reduction of 3.23 years in Horvath DNAmAge. The intervention included specific foods rather than caloric restriction.

The methyl adaptogen foods identified as most consistently associated with reduced epigenetic age across studies are: green tea, turmeric, garlic, blueberries, rosemary, oolong tea, and cruciferous vegetables. These are foods that influence the methylation machinery itself — not just providing methyl donors, but modulating the enzymes that place and remove methylation tags on DNA.

This connects directly to the seed cycling and epigenetic clock work covered in the seed cycling post: flaxseed lignans and sunflower B6 work through the same methylation pathway that these methyl adaptogens act on.

Diet quality and DunedinPACE — the NIH longitudinal study

A 2025 NIH National Institute on Aging study by Shen B et al., published in GeroScience, followed 421 adults over five years and found a clear longitudinal relationship between diet quality and DunedinPACE. A higher Dietary Inflammatory Index score — indicating a more pro-inflammatory diet — was independently associated with a higher (faster) DunedinPACE score. A higher Healthy Eating Index score was associated with a lower (slower) DunedinPACE. These relationships held after adjusting for age, race, poverty status, BMI, and smoking. Read the GeroScience paper here.

The practical translation: consistently eating in a way that reduces dietary inflammation — more vegetables, more oily fish, more whole grains, less ultra-processed food — is measurably associated with a slower pace of ageing over time, independent of other lifestyle factors.

What this means in practice

The clocks give us a framework, but the day-to-day application is simpler than the terminology suggests. Based on the strongest available evidence, here is what actually moves them.

For DunedinPACE (slowing your pace of ageing)

• Reduce dietary inflammation consistently — this is the single most evidenced lever. Ultra-processed foods, refined carbohydrates, and seed oils are the primary drivers of a high Dietary Inflammatory Index score.
• Increase diet quality score — more vegetables across more colours, legumes, oily fish, whole grains, nuts, and seeds. This is what the CALERIE and NIH longitudinal data show moving the clock.
• Caloric moderation — not aggressive restriction, but avoiding consistent overconsumption. The CALERIE trial used 25% restriction, which is extreme; even modest reductions in overall energy intake may contribute to slowing DunedinPACE.

For GrimAge (reducing mortality risk acceleration)

• Methyl adaptogen foods daily — green tea (two to three cups), turmeric (with black pepper for absorption), garlic, blueberries, and rosemary. These are the foods most associated with favourable GrimAge outcomes in observational and intervention studies.
• Adequate methyl donors — folate from dark leafy greens, B12 from oily fish and eggs, B6 from chickpeas and salmon. The methylation machinery needs substrate to function, and deficiency accelerates GrimAge.
• Sustained behaviour over months — GrimAge responds to intervention, but more slowly than DunedinPACE. Eight weeks is a floor, not a ceiling.

The cortisol clock — where ashwagandha fits

One aspect of clock biology that dietary intervention alone cannot fully address is the chronic stress axis. Cortisol dysregulation — particularly the flattened diurnal cortisol rhythm common in perimenopause — is independently associated with accelerated epigenetic ageing via GrimAge. The mechanism involves glucocorticoid receptor signalling that directly modifies DNA methylation patterns at stress-response genes.

Ashwagandha (KSM-66, the most clinically validated extract) has the strongest evidence base among adaptogens for reducing cortisol and restoring diurnal rhythm in stressed adults. It does not replace sleep, movement, or dietary intervention — but for women in perimenopause where cortisol dysregulation is near-universal, it is a meaningful supporting tool. Time Health Ashwagandha KSM-66 uses the extract form with the most consistent clinical backing.

What the clocks cannot tell you

It would be a disservice to present these tools without their limitations.

Commercial epigenetic testing is expensive and not yet standardised across laboratories. A result from one provider is not directly comparable to a result from another. Single measurements are useful for orientation but not for tracking change — you need at least two measurements from the same lab, ideally six months apart, to draw any meaningful conclusions from your personal data.

The clocks also cannot tell you which organ or system is driving your biological age acceleration. The multimodal approach from Li et al. (Cell 2026) is beginning to solve this — by combining data layers, it can identify whether the brain clock, the immune clock, or the metabolic clock is running ahead. But this level of specificity is not yet available in commercial testing.

What the clocks can do — and this is genuinely useful — is confirm that the dietary and lifestyle choices you make today are registering at a molecular level. The evidence that diet moves DunedinPACE is not theoretical. It has been measured in randomised trials. That is a meaningful finding, and it is worth acting on.

Sources

• Li J et al. Multimodal clocks of human aging. Cell, 2026. https://pubmed.ncbi.nlm.nih.gov/42105758/
• Shen B et al. Lower diet quality accelerates DNA methylation-based age. GeroScience, 2025. https://pmc.ncbi.nlm.nih.gov/articles/PMC12972488/
• Belsky DW et al. DunedinPACE, a DNA methylation biomarker of the pace of aging. eLife, 2022. Duke lab summary
• Dietary associations with reduced epigenetic age. Aging-US, 2025. https://www.aging-us.com/article/206240/text
• Caloric restriction and DunedinPACE — CALERIE trial analysis. Nature Aging, 2023. https://pmc.ncbi.nlm.nih.gov/articles/PMC10148951/