Does frequency or diversity of leisure activity matter more for epigenetic ageing? Analyses of arts engagement and physical activity in the UK Household Longitudinal Study

Using data from the UK Household Longitudinal Study, this research demonstrates that both the frequency and diversity of arts engagement and physical activity are associated with slower epigenetic ageing, with comparable effect sizes between the two activities, particularly among middle-aged and older adults.

The Gist

Imagine your body is like a high-tech car. For a long time, we've only been able to check the odometer to see how many miles (years) the car has driven. But recently, scientists invented a new kind of dashboard gauge called an "Epigenetic Clock."

This new gauge doesn't just count miles; it measures the wear and tear on the engine. It tells you if your car is aging faster than it should (rusting out early) or slower than expected (running like a dream).

This study asked a simple but profound question: To keep your engine running smoothly, is it better to do one thing really often, or to do a wide variety of different things?

The researchers looked at two main ways people spend their free time:

1. Arts and Culture: Singing, painting, visiting museums, or seeing a play.
2. Physical Activity: Running, swimming, gardening, or playing sports.

They analyzed data from over 3,300 adults in the UK to see how these activities affected their "biological age." Here is what they found, translated into everyday terms:

1. The "Variety vs. Frequency" Showdown

Think of your leisure time like a diet.

• Frequency is eating the same healthy meal every day.
• Diversity is eating a colorful rainbow of different healthy foods.

The study found that both matter, but diversity (variety) might be the secret sauce.

• For the Arts: Doing a mix of different cultural activities (like painting and visiting a museum and singing in a choir) was just as good for slowing down your biological clock as doing one single activity very frequently. It's like giving your brain and body a "multivitamin" of different experiences.
• For Exercise: The same rule applied. Doing a variety of sports (running, swimming, and yoga) was better for your biological clock than just doing one type of exercise, even if you did that one type every single day.

2. The "New Dashboard Gauges" Work Best

The researchers used seven different "clocks" to measure aging.

• The Old Clocks (1st Generation): These were like the old odometer. They just counted years. The study found that hobbies didn't really change the reading on these old gauges.
• The New Clocks (2nd & 3rd Generation): These are the fancy new digital dashboards that measure how fast the engine is actually wearing down. This is where the magic happened.
  • People who engaged in arts and exercise showed up as "younger" on these new, sensitive gauges.
  • Specifically, the DunedinPACE clock (which measures the speed of aging) showed the biggest improvement. It's as if the participants were driving at a slower, more sustainable speed, preserving their engine for longer.

3. The "Middle-Age" Boost

The study noticed something interesting about when these benefits kick in. The "anti-aging" effects of hobbies were much stronger for people in their 40s and older.

• Analogy: Think of your body like a house. When you're young, the house is sturdy, and a little rain (stress) doesn't hurt. But as the house gets older, you need to start doing maintenance. The study suggests that picking up diverse hobbies in middle age is like installing a new roof and reinforcing the foundation—it stops the house from deteriorating as fast as it otherwise would.

4. Why Does This Work?

The researchers suggest a few reasons why doing "fun stuff" slows down aging:

• Stress Relief: Arts and exercise act like a pressure valve, releasing stress that otherwise rusts your biological engine.
• The "Active Ingredients": Different activities provide different "ingredients" for health. Singing might boost your mood hormones, while gardening gets your blood flowing. Doing a variety of things ensures your body gets a full menu of benefits.
• Inflammation: Both arts and exercise seem to calm down the body's internal "fire" (inflammation), which is a major cause of aging.

The Bottom Line

You don't have to be a marathon runner or a professional painter to slow down your biological clock.

• Don't just repeat the same thing: If you love running, try adding a dance class or a pottery workshop.
• Don't just do it once a year: Consistency helps, but variety is the key to unlocking the full potential.
• Start now: While it's never too late, these habits seem to be the most powerful "maintenance tools" for people entering middle age.

In short: A colorful, varied life isn't just fun; it's a powerful tool to keep your biological engine running longer and smoother.

Technical Summary

1. Problem Statement

The study addresses a critical gap in understanding how modifiable lifestyle factors influence biological ageing, specifically through the lens of epigenetic clocks. While previous research has established links between lifestyle (e.g., smoking, diet) and epigenetic age, two key areas remain underexplored:

• Arts and Cultural Engagement (ACEng): Despite evidence that ACEng affects gene regulation and reduces stress, no epidemiological studies have linked it to epigenetic ageing.
• Physical Activity (PA) and Clock Generations: Existing studies on PA and epigenetic clocks have yielded inconsistent results, often limited by small sample sizes, lack of adjustment for confounders (e.g., socioeconomic position), and a focus on only a few specific clocks.
• Frequency vs. Diversity: It is unclear whether the frequency of engagement or the diversity (variety) of activities is more critical for slowing biological ageing. Diversity may provide access to a wider range of "active ingredients" (cognitive, social, physical stimuli) that buffer stress and promote resilience.

2. Methodology

Data Source

• Cohort: The UK Household Longitudinal Study (Understanding Society), a nationally representative panel survey.
• Sample: 3,354 adults of white European ancestry with DNA methylation (DNAm) data.
• Timeline: Blood samples were collected during nurse visits in Waves 2 and 3 (2010–2012); leisure activity data was derived from the Wave 2 survey.

Variables

• Outcomes (Epigenetic Clocks): Seven clocks representing three generations of ageing metrics:
  • First-Generation (Chronological Age): Hannum, Horvath2013, Horvath2018, Lin.
  • Second-Generation (Phenotypic Age/Mortality): PhenoAge.
  • Third-Generation (Pace of Ageing): DunedinPoAm and DunedinPACE.
• Exposures:
  • ACEng: Measured via four categories (participatory arts, receptive arts, heritage sites, other cultural activities). Metrics included Frequency (categorized by highest frequency across activities) and Diversity (count of activities, split into quartiles).
  • Physical Activity (PA): Measured via vigorous, moderate, and mild activities. Metrics included Frequency, Diversity (count of activities), and Activeness (self-rated 0–10 scale).
• Covariates: A comprehensive set including demographics (age, sex, marital status), socioeconomic factors (income, education, area deprivation), behavioral factors (smoking, alcohol), health status (impairments), and technical covariates (blood cell composition estimates, time gaps between data collection).

Statistical Analysis

• Primary Method: Doubly Robust Estimation using the Inverse-Probability-Weighted Regression Adjustment (IPWRA) estimator.
  • This approach builds two models: one for the outcome and one for the treatment assignment (exposure).
  • It provides consistent estimates if either the outcome model or the treatment model is correctly specified, reducing bias from residual confounding.
• Sensitivity Analyses:
  • Adjusting for Body Mass Index (BMI) to test if effects are mediated by weight.
  • Restricting the sample to adults aged 40+ (where ageing acceleration is more pronounced).
  • Analyzing PA activeness levels separately.

3. Key Results

Association with Epigenetic Clocks

• First-Generation Clocks: No significant associations were found between ACEng or PA and the first-generation clocks (Hannum, Horvath2013, Horvath2018, Lin).
• Second- and Third-Generation Clocks: Significant associations were found with PhenoAge, DunedinPoAm, and DunedinPACE, indicating that leisure activities are linked to slower biological ageing and a decelerated pace of ageing.

Frequency vs. Diversity

• Arts and Cultural Engagement (ACEng):
  • Frequency: Weekly and monthly engagement was associated with lower PhenoAge and slower pace (DunedinPoAm/PACE) compared to low frequency.
  • Diversity: High and very high diversity were associated with slower pace (DunedinPoAm/PACE). For PhenoAge, the relationship was slightly clearer for frequency than diversity, though high diversity still showed benefits.
• Physical Activity (PA):
  • Frequency: Weekly engagement was associated with slower pace (DunedinPoAm/PACE).
  • Diversity: High diversity was associated with lower PhenoAge. For DunedinPoAm/PACE, even low and medium diversity showed associations with slower ageing compared to no activity.
  • Activeness: Very high self-rated activeness showed the strongest association with lower PhenoAge.

Comparative Effect Sizes

• The effect sizes for ACEng and PA were comparable. This suggests that engaging in arts and culture offers biological benefits similar to those of physical exercise regarding epigenetic ageing.
• Age Stratification: Associations were generally stronger and effect sizes larger in middle-aged and older adults (40+).

Robustness

• Results persisted after adjusting for BMI, suggesting the effects are not solely mediated by weight management.
• The use of doubly robust estimation minimized the risk of residual confounding from socioeconomic and behavioral factors.

4. Key Contributions

1. First Evidence for ACEng: This is the first epidemiological study to link Arts and Cultural Engagement with epigenetic ageing, demonstrating that non-physical leisure activities have measurable biological impacts.
2. Frequency vs. Diversity: The study clarifies that both frequency and diversity matter. Diversity is particularly important as it may provide access to varied "active ingredients" (social, cognitive, sensory) that buffer stress and reduce inflammation.
3. Clock Generation Specificity: The study highlights that second- and third-generation clocks (PhenoAge, DunedinPoAm/PACE) are more sensitive to lifestyle interventions than first-generation clocks. This supports the view that newer clocks better capture the "pace" of ageing and clinical health decline rather than just chronological time.
4. Methodological Advancement: By applying doubly robust estimation to a large, nationally representative cohort, the study overcomes limitations of previous small-scale studies that relied on simple regression adjustments, offering stronger causal inference.

5. Significance and Translational Implications

• Public Health Policy: The findings suggest that promoting diverse leisure activities, including arts and culture, is as vital for healthy ageing as promoting physical exercise. Interventions should encourage variety in engagement, not just repetition.
• Biological Mechanisms: The results support the hypothesis that leisure activities slow epigenetic ageing through mechanisms such as stress reduction (lowering cortisol and inflammatory markers) and cardiometabolic improvement. The "inflammageing" pathway appears to be a key mediator.
• Intervention Potential: Since epigenetic changes are potentially reversible, these findings provide a strong rationale for future intervention studies to test whether increasing the frequency and diversity of leisure activities can actively slow or reverse biological ageing in mid-to-late life.
• Healthspan: By decelerating the pace of biological ageing (DunedinPACE), these activities may extend "healthspan"—the period of life spent in good health—thereby reducing the burden on healthcare systems and improving quality of life for ageing populations.

Limitations Noted:

• Reliance on self-reported leisure data (potential recall bias).
• Use of whole blood DNAm, which may not reflect tissue-specific changes (e.g., in muscle for PA).
• Cross-sectional nature of the exposure-outcome relationship, though the use of robust statistical methods mitigates some causal inference concerns.