The People Who Age Well: What a New Study on Social Connection and Epigenetic Clocks Taught Me About Longevity

There is a particular finding I keep returning to. In a 2025 study published in Scientific Reports, researchers analysed data from 1,309 adults in the Midlife in the United States (MIDUS) cohort and found that people with a “net positive” ratio of social experiences — more positive interactions than negative over a lifetime — had GrimAge scores that were 4.63 years younger on average than those with net negative social experiences. After adjusting for demographics, health behaviours, and confounders, the difference held at 2.85 years.

Not 2.85 years from a pharmaceutical intervention. Not from an extreme dietary protocol. From the texture of your social life, measured in DNA methylation patterns at your CpG sites.

Longevity content almost never talks about this. The conversation is dominated by food, supplements, sleep, cold exposure, and exercise — all of which matter. But the social connection data has been quietly building for years, and it is now robust enough that ignoring it is a genuine gap in any serious biological age framework.

What the research actually shows

The JAGS 2026 study by Hu Z et al. — “Social Engagement and Epigenetic Age Acceleration in the Health and Retirement Study” — is the most recent and largest of a growing body of research connecting social life to biological age. Using data from the Health and Retirement Study, which tracks thousands of older American adults with repeated biological sampling, the researchers found that active social engagement — specifically community activities, physical activities done socially, cognitive engagement, and home-based creative activity done with others — was independently associated with reduced epigenetic age acceleration across multiple clock measures. The findings highlighted physical and community social activities as the highest-impact categories. Read the study abstract here.

This sits alongside the MIDUS study findings published in Scientific Reports in 2025 by Prigerson, Russell, and Maciejewski — the one that found the 4.63-year GrimAge difference between net positive and net negative social experiencers. Their analysis used both GrimAge and DunedinPACE and found that positive social experiences including regular social meeting attendance independently predicted decelerated epigenetic ageing, while negative social experiences — childhood adversity, social disruption — independently predicted acceleration. The effects were additive and independent of perceived social support levels and self-rated health. Read the full paper here.

And from the University of Minnesota, a 2024 study published in Psychology and Aging — Freilich et al., using the MIDUS cohort — found that loneliness was weakly but independently associated with higher epigenetic age acceleration in Horvath, DunedinPACE, and GrimAge measures, even after controlling for demographics and health behaviours. Critically, they also found a synergistic interaction: the health impacts of loneliness were significantly more pronounced in individuals with higher DunedinPACE scores. Social isolation and a fast pace of ageing appear to amplify each other. Read the study here.

The mechanism — why connection registers in your DNA

This is not mystical. The pathway from social experience to epigenetic clock is biological and reasonably well characterised.

Loneliness and social threat activate the sympathetic nervous system and the HPA (hypothalamic-pituitary-adrenal) axis — the same stress response system that regulates cortisol. Chronic activation of this system produces sustained elevations in cortisol and pro-inflammatory signalling molecules including TNF-α and IL-6. These are the same cytokines elevated in the low-grade chronic inflammation — inflammageing — that is one of the primary drivers of biological age acceleration.

Steve Cole’s research at UCLA has shown that loneliness produces a specific genomic signature: upregulation of pro-inflammatory transcription factors and downregulation of antiviral transcription factors. This pattern — called the Conserved Transcriptional Response to Adversity (CTRA) — appears to be directly written into DNA methylation patterns over time. It is the biological record of accumulated social experience, encoded at the epigenetic level.

Conversely, social connection and the subjective experience of safety within relationships appears to dampen HPA axis reactivity, reduce baseline cortisol and inflammatory cytokines, and, through those downstream effects, slow the rate at which unfavourable methylation patterns accumulate. This is why the signal shows up in the clocks.

Why this is particularly relevant in perimenopause

Perimenopause is a period when social connection is under more pressure than at almost any other life stage — and when the biological consequences of that pressure are highest.

The hormonal changes of perimenopause already dysregulate the HPA axis. Oestrogen normally buffers cortisol reactivity — it modulates glucocorticoid receptor sensitivity and supports the negative feedback loop that brings cortisol back down after a stressor. As oestrogen declines, this buffering weakens. The HPA axis becomes more reactive; cortisol responses are larger and slower to resolve. A woman in perimenopause is, from a neuroendocrine perspective, more susceptible to the biological consequences of loneliness and social stress than she was a decade earlier.

At the same time, the social landscape of midlife frequently contracts. Children leave home. Friendships that were built around school or early career become less proximate. Long-term relationships face the particular strain of two people navigating their own hormonal and identity transitions simultaneously. The social infrastructure of the twenties and thirties — built largely around shared circumstance — often requires more deliberate maintenance at this stage, precisely when there is less energy to spend on it.

The data suggest this is not merely a quality-of-life concern. It is a biological age concern.

What the research suggests actually works

The JAGS study’s finding about which categories of social engagement had the highest biological impact is useful here. It was not large social events or broad social networks. It was recurring, purposeful activities — community groups, physical activity with others, creative work shared with people you see regularly. The mechanism appears to be the regularity and predictability of positive social contact, rather than its intensity or breadth.

This maps onto what we know about the cortisol pathway. What dampens HPA reactivity is not a single positive social experience — it is the accumulated signal that social threat is low and social safety is high, built up through repeated, reliable contact with people who are genuinely glad to see you. That is a different thing from a busy social calendar.

A few practical framings that emerge from the evidence:

• Recurring weekly contact matters more than occasional larger gatherings. A regular walk with one friend, a weekly group class, a book group that meets monthly — these low-stakes, predictable points of connection appear to have more epigenetic impact than sporadic high-intensity social events.
• Shared physical activity compounds the benefit. The JAGS study specifically highlighted physical activities done socially as high-impact. This makes mechanistic sense: exercise independently reduces cortisol and inflammatory markers; social connection independently reduces them; the combination appears to be additive.
• Community and purposeful activity were the other high-impact category. Volunteering, creative groups, any activity that involves being in a community of shared purpose. This connects to the identity literature: women who maintain a strong sense of purposeful engagement outside the domestic sphere show lower epigenetic age acceleration than those who do not, independent of physical health.
• The ratio of positive to negative experiences is what moves the clock. The MIDUS data did not find that you need a perfect social life — it found that a net positive ratio was protective. That framing is useful: the goal is not to maximise social activity but to ensure the balance tilts towards safety and warmth over time.

The cortisol connection — and where support helps

If social connection is partly mediated through the cortisol and HPA axis pathway, then anything that supports cortisol regulation also creates more biological capacity to benefit from whatever social connection is available. For women in perimenopause where HPA dysregulation is near-universal, this is not a trivial consideration.

Ashwagandha (KSM-66) has the strongest clinical evidence among adaptogens for reducing cortisol and restoring diurnal rhythm in stressed adults — which is the same system through which loneliness does its biological damage. Sleep quality, which is where cortisol regulation is partly reset overnight, is also directly relevant here. These are not alternatives to social connection — they are supports that make the body more capable of registering its benefits. More on this in the invisible work of ageing well post.

The gut microbiome is a third intersection point. The gut-brain axis is bidirectional: social stress disrupts microbiome diversity through the cortisol-gut pathway, and a depleted microbiome feeds back into mood and stress reactivity through the vagus nerve. The estrobolome work covered in the estrobolome post — and the fermented foods protocol from the Friday reset — sits inside this same system.

A note on what this is not

This is not an instruction to be more sociable if that is not who you are. The introversion-extraversion dimension is real, and forced social engagement that feels depleting is not the same thing as genuine connection — and may not produce the same biological signal. What the research points to is the quality of felt safety within relationships, not the quantity of social activity.

Nor is it a suggestion that social isolation is a personal failing with health consequences that you should feel anxious about. That would be counterproductive in the most literal sense — anxiety activates the same HPA pathway that social threat does. The evidence is offered in the same spirit as the dietary evidence: as information that expands the range of levers you know are available.

The people who age well, in the data, are not necessarily the most extroverted, the most disciplined, or the most optimised. They are, more often, people who have managed to maintain a consistent experience of warmth and safety in their lives — through relationships, through purposeful activity, through community — alongside the dietary and lifestyle foundations that the rest of this series covers.

That combination — food, movement, sleep, and genuine human connection — is what the ageing clocks appear to be measuring when they run slow.

Sources

• Hu Z et al. Social Engagement and Epigenetic Age Acceleration in the Health and Retirement Study. Journal of the American Geriatrics Society, 2026. https://pubmed.ncbi.nlm.nih.gov/41697905/
• Prigerson HG, Russell D, Maciejewski PK. Associations between positive and negative social experiences and epigenetic aging. Scientific Reports, 2025. https://www.nature.com/articles/s41598-025-07222-z
• Freilich CD et al. Loneliness, Epigenetic Age Acceleration, and Chronic Health Conditions. Psychology and Aging, 2024. https://pmc.ncbi.nlm.nih.gov/articles/PMC11283257/
• Cole SW et al. Social regulation of gene expression in human leukocytes. Genome Biology, 2007.