Digital health, fitness, and longevity platforms have focused on the constant measurement and relentless optimization of biometric parameters, an often-valuable effort but one that has systematically left behind vital components of health and flourishing – such as social connection — that are more difficult to measure. 

The Health Value of Social Connection

Robust longitudinal studies, including the Harvard Study of Adult Development, and Northwestern’s Super Agers study, have demonstrated that a critical determinant of longevity is warm human relationships.  The Harvard research, for example, “found that the people who stayed healthiest and lived longest were the people who had the strongest connections to others,” according to a Harvard School of Public Health summary of the work.

Moreover, a 2021 review by Jaime Vila of 23 meta-analyses of studies conducted over the last sixty years, involving over a billion subjects, found a “significant association of functional and structural measures of social support with health and longevity.” 

Vila continues, “The strength of this association is equivalent to that documented for other well-documented risk factors such as smoking or obesity.”

The health importance of social connection is also a key theme of Why Brains Need Friends, a new book by neurobiologist (and social media phenomenon) Ben Rein that I recently reviewed for the Wall Street Journal. 

Rein’s bottom line: “I really think you should spend more time socializing.”  He adds, “Your brain is counting on you.”

This message is reinforced by a pair of just-published papers that used “aging clocks” to explore the impact of social connection.  One of these papers was highlighted by geroscience researcher Steve Horvath at the 2025 Biomarkers of Aging conference held this week at Harvard Medical School.

Before we get to these papers, I thought it might be useful to share a few thoughts from this meeting with TR readers.

Biomarkers of Aging Conference: Reflections

As a first-time attendee (and I only was able to make it to some of the sessions), I was struck by the remarkable (and welcome!) degree of energy, hope, and optimism among so many of the speakers, including many (like Horvath) who were toiling in the field long before it became so trendy. 

Much of the work presented involved various measures of aging, so-called “aging clocks” that relate specific patterns of biological or structural measurements to estimates of age and health.  For example, patterns of DNA methylation accumulate as individuals age, and seem to build up faster in less healthy people; thus, this measurement can reveal someone is “older” or “younger” than their chronological age would otherwise suggest.

The first multi-tissue aging clock, based on DNA methylation, was developed by Horvath back when he was a geneticist at UCLA; today, he works at Altos Labs.  (This 2024 interview of Horvath by Dr. Eric Topol on the Ground Truths podcast explores the fascinating origin story, rich in serendipity, of aging clocks, and is highly recommended.)  Perhaps not surprisingly, a colleague referred to Horvath as the “King of Clocks” in a discussion at the conference.

The attraction of ever more researchers to geroscience is easy to understand.  As Nobel Laureate Venki Ramakrishnan writes in Why We Die, “aging is connected intimately with so many biological processes,” and captures the attention of scientists in so many disciplines.

These connections weren’t always so apparent.  Ramakrishnan, an RNA biologist (see his wonderful first book, Gene Machine), explains “when I started my career, I had no idea that anything I did would be connect with the problem of why we age and die.” 

But more recently, Ramakrishnan says, he’s learned that his life’s research on “the problems of how proteins are made” is highly relevant to aging, since “much of aging has to do with how our body regulates the production and destruction of proteins.”

As palpable as the attendees’ excitement, however, was a frustration, perhaps even impatience, with a nagging sense that their work is still not quite taken seriously by the biomedical “establishment” – even as they appreciate that many patients (or more precisely, individuals who are hoping to forestall becoming patients) are drawn to the field looking for answers. 

Among the promising initiatives to integrate longevity into medicine were:

• “Healthy Longevity Medicine Society,” discussed by professor Andrea Meier of the National University of Singapore. Meier was the founding President of the organization, which was established in 2022 “to build a clinically credible framework and platform for longevity medicine that promotes the highest standards of interdisciplinary collaboration in the field.”
  

  Dr. Evelyne Bischof

• “Sheba Longevity Center,” discussed by medical director Evelyne Bischof. Sheba, located in the Tel Aviv metropolitan area in Israel, describes itself as “the first longevity center within an academic medical institution,” and is focused on “integrating the latest research and innovations in longevity into patient care.”

Other prominent centers in geroscience include the Buck Institute in Novato, California, “an independent biomedical research institute that researches aging and age-related disease,” and was founded in 1999; and the Abu Dhabi-based Institute for Healthier Living, which characterizes itself as “the world’s first specialized healthy longevity medicine center.”

On many levels, I especially appreciated the presentation from Duke’s Terrie Moffitt, a psychologist who first pursued the roots of antisocial behavior, and who — by following the Dunedin birth cohort across five decades — has helped redefine how we understand biological aging. 

Her group’s DunedinPACE clock, now widely regarded as one of the most rigorous measures of biological aging, is even featured in the Rejuvenation Olympics (although, as Open Evidence reminds us, Dunedin aging measures “are not yet established for routine clinical practice”).

In her talk, Moffitt described a range of new clocks the team is developing, including one based on neuroimaging (the DunedInPACENI) that aims to predict overall aging rate (not the aging rate of the brain, specifically).

The conference presenters included several leaders from several emerging biotechs focused on aging, and with established pharma partnerships. These include:

• BioAge, represented by CEO and co-founder Kristen Fortney.  The company has ongoing relationships with Lilly (focused on therapeutic antibody discovery) and Novartis (focused on target discovery, with an eye towards therapeutics that can mimic some of the benefits of exercise).
  

  Alex Zhavoronkov, CEO Insilico Medicine

• Insilico Medicine, represented by founder and CEO Alex Zhavoronkov. The company’s active partnerships include a target discovery relationship with Sanofi, earning a milestone payment in 2024.

The only speaker from an established and comparatively large biopharma was David Glass,  a world expert in muscle biology who is now Vice President, Research, Aging/Age-Related Disorders, at Regeneron.  Glass thoughtfully discussed important (and easily-overlooked) nuances of the molecular study of sarcopenia in rats vs mice.

A final observation from this conference is simply that as I considered the ever-increasing number of biological clocks available, their remarkable ease of use, and the tendency to use multiple clocks in single studies, I found it all too easy to imagine p-hacking efforts (whether deliberate or just self-deception) where, if you test enough clocks in enough conditions, eventually you find some association, and then can persuade yourself (or others) that you’ve found a way to reverse aging – or have discovered a factor that accelerates it.  If most everything can be tied to alterations in some biomarker of aging, then it will be increasingly challenging to identify meaningful signal beneath the volume of noise.

Fortunately, while one of studies we’re about to discuss does consider multiple measures of aging, it was performed rigorously, and accounts for multiple comparisons in their statistical analysis.

Social Connection and Biological Aging

Social connection, as we’ve discussed, is good for your health.  Two research groups have recently published findings reflecting their efforts to relate measures of biological aging to social connection and dissect the most impactful components.  Interesting, the teams both analyzed the same basic dataset — a well-characterized cohort, the Midlife in the United States (MIDUS) project that includes extensive self-report data as well as biobanked samples — but approached their analyses in slightly different, fairly complementary fashions.

One team — Anthony Ong of Cornell, Frank Mann of Stony Brook, and senior author Laura Kubzansky of Harvard – analyzed data from over 2000 MIDUS subjects, with an average age of about 55.

The authors utilized a measure – “cumulative social advantage” (CSA) – they previous developed, intended to capture and integrate social resources across four broad domains:

• Religious and faith-based support
• Parent-child relationship quality
• Community Engagement
• Extended emotional support.

These were assessed by self-report, using several questions for each domain.

Then, the researchers analyzed samples by using a range of stress and aging, including several biological clocks, serum measures of inflammation, and urine measures of neuroendocrine function. The numbers were then crunched, comparing the singular, composite assessment of connection (CSA) to each of these measures of stress and aging, explicitly correcting for the multiple comparisons. 

The result: higher CSA scores were linked in statistically significant manner to slower biological aging, and also with lower levels of IL-6 (a marker of inflammation), even after accounting for a comprehensive set of sociodemographic (eg socioeconomic status) and health-related (eg smoking history) confounders.

Within the CSA construct, the component of social integration–measured by the self-reported response to ‘I feel close to other people in my community’–accounted for much of the variability in CSA scores. The contribution of perceived community closeness to the observed association between CSA and slower biological aging and lower inflammation was inferred from its central role within the CSA composite.

The second team of scientists – Holly Prigerson and Paul Maciejewski of Cornell, together David Russell of Appalachian State University – examined data from 1309 MIDUS subjects, average age about 51.

This team explored the relationship between (presumptive) positive social experiences (such as volunteering for organizations) and negative social experiences (eg dropping out of school), on the one hand and two measures of biological aging on the others.  The intention was to determine, after adjusting for known confounders (critical!), which experiences most impact aging, and also to explore whether there was a relationship between the net ratio of positive/negative experiences and aging.

The key findings: engaging in social meetings had the strongest (and highly significant) positive effect on aging; also scoring well: marriage and giving assistance to others.  On the negative side, a strikingly deleterious effect on aging was associated with dropping out of school and with the subject’s parents experiencing drug problems. 

Perhaps not surprisingly, a respondent’s total number of positive experiences were associated with slower aging, and total number of negative experiences were associated with accelerated aging.  Comparison of subjects with “net positive” ratio of experiences to those with “net negative” revealed a profound difference between these two groups, with the “net negative” group aging much faster.

Interestingly, perceived social support was not found to be associated with aging in this publication.

While both studies link aspects of social connection with aging, they don’t speak directly either to causation (the social factors measured may reflect a deeper underlying driver of stress and aging, say) or to the biological mechanisms by which social connectivity exerts its putative effect on the biology. 

Also not addressed: the potential impact on aging of interventions focused on enhancing social connectivity – a critical question for those seeking to improve health rather than just measure it.

Bottom Line

Social connection exerts a powerful effect on health and aging, with robust evidence linking greater social engagement and support to reduced morbidity, mortality, and slower biological aging. These relationships deserve greater attention in longevity-promoting efforts, including digital health and wellness platforms. Geroscience research is experiencing a remarkable intensification of interest, partly due to the availability of biological aging measures, though consumer enthusiasm for these tools often exceeds their current scientific validation. The ability to connect social connection to biological aging, and to use these measures to dissect and optimize social interventions, represents an exciting but still nascent approach to developing meaningful, targeted strategies for healthy aging.