https://academic.oup.com/innovateage/article/10/6/igag038/8669801
Paper summary
The paper asks whether two forms of leisure activity — arts and cultural engagement and physical activity — are associated with slower epigenetic aging. It uses data from 3,556 adults in the UK Household Longitudinal Study, with blood DNA methylation data collected in 2010–2012. The authors analyse seven epigenetic clocks: Hannum, Horvath 2013, Horvath 2018, Lin, PhenoAge, DunedinPoAm and DunedinPACE. They use a doubly robust inverse-probability-weighted regression adjustment approach, adjusting for demographic, socioeconomic, sample timing, and technical blood-cell covariates.
The central finding is that both arts/cultural engagement and physical activity are associated with slower aging on PhenoAge, DunedinPoAm and DunedinPACE, but not on the first-generation clocks — Hannum, Horvath 2013, Horvath 2018 or Lin. The effect sizes for arts/cultural engagement were broadly comparable to those for physical activity.
For arts engagement, both frequency and diversity mattered. Monthly or weekly arts/cultural engagement was associated with lower PhenoAge, and more frequent or more diverse engagement was associated with slower DunedinPoAm and DunedinPACE. For physical activity, frequency, diversity and self-rated activeness all showed similar patterns: stronger associations with the newer clocks, especially DunedinPACE.
The findings were stronger in people aged 40 and above, and largely persisted after additional adjustment for smoking, alcohol, BMI, disability/illness and other health-related factors.
The authors interpret this as evidence that arts and cultural activities may be a biologically relevant health behaviour, potentially working through stress reduction, inflammation, social identity, cognitive stimulation and cardiometabolic pathways. They also argue that the lack of association with first-generation clocks is not surprising, because first-generation clocks are mainly trained to predict chronological age, whereas PhenoAge and DunedinPACE/PoAm are more related to morbidity, mortality and pace of physiological aging.
What is novel?
The main novelty is that this appears to be the first epidemiological study linking arts and cultural engagement to epigenetic aging. Prior work had examined physical activity, diet, stress, smoking and other lifestyle factors, but arts/cultural participation had not been studied in this context.
A second novelty is the comparison between arts engagement and physical activity in the same cohort, using the same methods and the same clock outcomes. This allows the authors to argue that arts engagement has associations of similar magnitude to a much more established health behaviour: physical activity.
A third novelty is methodological: the paper uses multiple clock generations and a doubly robust estimator rather than relying on one clock and simple regression. This is useful because the study shows a clear divergence between clock types: no signal in first-generation clocks, but a signal in PhenoAge and Dunedin pace-of-aging clocks.
A fourth useful feature is that leisure activity is not treated as a single binary variable. The study distinguishes frequency, diversity, and, for physical activity, activeness. This is important because “doing more varied things” may capture social, cognitive, emotional and physical stimulation better than simple frequency.
Critique
The main limitation is that this is an observational association study, not an intervention. Even with doubly robust estimation, the result cannot prove that arts engagement or physical activity slows biological aging. People who engage in many arts/cultural activities may differ in unmeasured ways: cognitive reserve, social networks, personality, baseline health, childhood socioeconomic status, wealth, mobility, urban access, diet quality, or healthcare engagement. These factors could partly drive both leisure participation and better epigenetic aging profiles.
A second issue is reverse causality. People with slower biological aging, better health, better mobility, lower fatigue, better cognition or fewer depressive symptoms may simply be more able to attend cultural events, visit heritage sites, dance, sing, exercise, travel or participate socially. The authors adjust for some health variables, but the exposure and methylation data are close in time, so directionality remains uncertain.
A third limitation is measurement. Arts/cultural engagement and physical activity are self-reported, and the arts measure is broad. Visiting a museum, singing in a choir, painting, dancing, going to a library and attending a concert may have quite different biological effects. Combining them into frequency/diversity scores is useful for population analysis, but it blurs mechanism.
A fourth point is that the biological material is whole blood DNA methylation. This may capture immune/inflammatory state well, but it may not reflect aging in brain, muscle, liver, vascular tissue or other organs. This is especially relevant for physical activity, where muscle tissue would be mechanistically important, and for arts engagement, where neural, endocrine and autonomic pathways may be central.
A fifth issue is that the largest claims lean slightly ahead of the evidence. The paper frames arts engagement as a potential public-health behaviour for healthy aging, which is plausible, but the evidence currently supports association, not prescription. An intervention trial assigning people to structured arts/cultural engagement and measuring longitudinal change in DunedinPACE or other biomarkers would be needed before making strong causal claims.
A sixth critique concerns the clocks themselves. The fact that only PhenoAge, DunedinPoAm and DunedinPACE respond could mean these clocks are more biologically relevant. But it could also mean they are more sensitive to current health, inflammation, immune-cell composition, socioeconomic patterning or general physiological status. The null results in first-generation clocks are not a problem, but they remind us that “epigenetic aging” is not a single thing.
Bottom line
This is a useful and interesting paper. Its strongest contribution is showing that arts and cultural engagement correlate with slower epigenetic aging to a similar degree as physical activity, especially on newer clocks linked to healthspan and pace of aging. The result is plausible and policy-relevant, but it should be read as hypothesis-generating rather than causal proof. The next step should be longitudinal and interventional work, ideally with objective activity measures, richer phenotyping, inflammatory/metabolic mediators, and repeated methylation sampling.