The short version
The biggest hobby dataset in the literature links having a hobby to fewer depressive symptoms and higher self-rated health in adults aged 65 and over, across 16 countries. It is observational, so it cannot show that starting a hobby improves anything. The craft and arts trial literature is stronger on short-term effects and weaker on rigour. This article sorts the two, then helps you choose one hobby and give it a slot in the week.
The biggest hobby dataset there is
Researchers harmonised five longitudinal studies covering 16 countries and 93,263 adults aged 65 and over. Hobby engagement varied enormously by country - 51% of Spanish respondents had a hobby, against 96% in Denmark.
Across the pooled data, having a hobby was associated with fewer depressive symptoms (coefficient -0.10, 95% CI -0.13 to -0.07) and with higher self-rated health, happiness and life satisfaction. The associations survived models that used earlier measures to predict later outcomes, which is the strongest thing an observational dataset can do.
It is still observational. People who take up hobbies differ from people who do not in income, health, social support and a hundred unmeasured ways. The study is compatible with hobbies improving well-being, with well-being enabling hobbies, and with both.
[C01] [C02]
Frequency and variety versus hours
A meta-analysis of leisure engagement and subjective well-being found that measures of how often people engaged and how many different activities they did tracked well-being more closely than total time spent. The authors argued that pattern is a reason to treat leisure as something you can deliberately shape.
Practically, that means a thirty-minute session three times a week is a defensible starting point, and adding a second activity may matter more than extending the first. It also means that the usual excuse - I do not have a big enough block of time - is answering the wrong question.
[C03]
What the trial literature adds
The observational data is complemented by two reviews of hands-on and group activities. A systematic review of crafts-based interventions found positive short-term effects on depression, anxiety, stress and mental well-being across 19 studies, but most were small, several were not randomised, and follow-up was short.
Stronger evidence exists for group arts programmes in older adults: a 2025 meta-analysis of 39 controlled studies found moderate reductions in depression (d 0.70) and anxiety (d 0.76), with publication bias flagged in the depression analysis. That is the best experimental signal in this area, and it belongs to structured group programmes, not to the general idea of having a hobby.
Read together, the experimental evidence supports short-term improvements from doing something absorbing, especially in a group. It does not support a claim that any particular hobby changes long-term mental health.
[C04] [C05]
Choosing one that survives your week
Pick for repeatability first and interest second. The evidence is about engagement over time, and a hobby you cannot do on a Tuesday will not produce any. Filming something short, ten minutes of drawing, going to a class, repairing one thing, cooking one thing you have never cooked: all of these are engagement.
Prefer something with a session-length unit small enough to fit a bad week. Then give it a slot in a calendar, or attach it to an existing anchor - after dinner, before the news, on the walk home. This is the same implementation logic that goal-tracking research uses, and it is mundane on purpose.
If the thing you want is company, the group-arts evidence is the relevant one, and a class is the cheapest way to buy structure plus company. If the thing you want is absorbing time alone, the craft literature is the relevant one, and it is weaker, so calibrate your expectations to 'this feels better this week', not to a lasting clinical change.
Where this is not the answer
If you have persistent low mood, anxiety that interferes with functioning, or you are using alcohol or other substances to cope, a hobby is not the treatment. Get assessed and treated; then a hobby can sit alongside that, as the kind of thing that makes a week feel less thin.
The other failure mode is turning the hobby into a second job: an obligation to log progress, to be good at it, or to monetise it. Nothing in this evidence says the hobby has to be productive. The measure that showed the association was simply having one.
Claims, sources and what they can carry
Every claim links to the source(s) behind it. Weights rank source types
(Cochrane/systematic reviews 1.0, recognised guidelines 0.9, RCTs 0.8, observational or
preprint 0.5, mechanistic 0.2). A weight is not certainty and is not GRADE.
directionalc-01In 93,263 adults aged 65 and over across 16 countries, having a hobby was associated with fewer depressive symptoms (-0.10, 95% CI -0.13 to -0.07) and with higher self-rated health, happiness and life satisfaction.
Mak HW, Noguchi T, Bone JK, Wels J, Gao Q, Kondo K, Saito T, Fancourt D. Hobby engagement and mental wellbeing among people aged 65 years and older in 16 countries. Nat Med. 2023;29(9):2233-2240.
PMID 37696932 (verification gap: Observational cohort data in adults aged 65 and over: the association is consistent but does not show that starting a hobby causes the improvement, and the size of the association for younger adults is not established.)Harmonised analysis and multinational meta-analysis of five longitudinal cohorts (ELSA, JAGES, HRS, SHARE, CHARLS) covering 93,263 adults aged 65 or older in 16 countries, with fixed-effects and longitudinal regression models. · weight 0.5 · Having a hobby was associated with fewer depressive symptoms (pooled coefficient -0.10, 95% CI -0.13 to -0.07) and with higher self-reported health (0.06, 0.03 to 0.08), happiness (0.09, 0.06 to 0.13) and life satisfaction (0.10, 0.08 to 0.12). Longitudinal models testing direction were consistent, and hobby prevalence ranged from 51% in Spain to 96% in Denmark.↑ back to the article
uncertainc-02The hobby dataset is observational, so it cannot establish that hobby engagement causes better mental health; longitudinal models were consistent but not causal.
Mak HW, Noguchi T, Bone JK, Wels J, Gao Q, Kondo K, Saito T, Fancourt D. Hobby engagement and mental wellbeing among people aged 65 years and older in 16 countries. Nat Med. 2023;29(9):2233-2240.
PMID 37696932 (verification gap: Observational cohort data in adults aged 65 and over: the association is consistent but does not show that starting a hobby causes the improvement, and the size of the association for younger adults is not established.)Harmonised analysis and multinational meta-analysis of five longitudinal cohorts (ELSA, JAGES, HRS, SHARE, CHARLS) covering 93,263 adults aged 65 or older in 16 countries, with fixed-effects and longitudinal regression models. · weight 0.5 · Having a hobby was associated with fewer depressive symptoms (pooled coefficient -0.10, 95% CI -0.13 to -0.07) and with higher self-reported health (0.06, 0.03 to 0.08), happiness (0.09, 0.06 to 0.13) and life satisfaction (0.10, 0.08 to 0.12). Longitudinal models testing direction were consistent, and hobby prevalence ranged from 51% in Spain to 96% in Denmark.↑ back to the article
uncertainc-03In a meta-analysis of leisure and subjective well-being, measures of the frequency and diversity of engagement were more strongly associated with well-being than measures of time spent.
Kuykendall L, Tay L, Ng V. Leisure engagement and subjective well-being: A meta-analysis. Psychol Bull. 2015;141(2):364-403.
PMID 25602273 (verification gap: Correlational evidence base, so it says more about which measures track well-being than about whether adding a hobby changes it.)Meta-analysis of the relationship between leisure engagement and subjective well-being, comparing how different ways of measuring leisure change the association. · weight 0.5 · Measures of the frequency and diversity of leisure engagement were more strongly associated with subjective well-being than measures of time spent in leisure, which the authors read as support for treating leisure as a target for improving well-being.↑ back to the article
uncertainc-04A systematic review of 19 crafts-based intervention studies found positive short-term effects on depression, anxiety, stress and well-being, but most studies were small and only some were randomised.
Bukhave EB, Creek J, Linstad AK, Frandsen TF. The effects of crafts-based interventions on mental health and well-being: A systematic review. Aust Occup Ther J. 2025;72(1):e70001.
PMID 39956657 (verification gap: Most included studies were not randomised, follow-up was short, and the review could not pool results; the positive direction is the authors' summary, not an effect size.)Systematic review of 19 craft-intervention studies (3 randomised controlled trials, 1 pilot randomised trial, 7 non-randomised pre-post trials, 6 mixed-methods, 1 qualitative, 1 registry study) in adults with mental health disorders or distress secondary to a chronic illness. · weight 0.5 · All included studies reported short-term improvements in the outcomes measured, including anxiety, stress, depression, mood, well-being, self-esteem and life satisfaction. The review states that wide variation in interventions, designs, outcomes and research quality made firm conclusions about effectiveness difficult, and that most studies had fewer than 50 participants.↑ back to the article
directionalc-05A meta-analysis of 39 controlled studies found group arts interventions reduced depression (d 0.70) and anxiety (d 0.76) in older adults, with publication bias flagged in the depression analysis.
Quinn EA, Millard E, Jones JM. Group arts interventions for depression and anxiety among older adults: a systematic review and meta-analysis. Nat Ment Health. 2025;3(3):374-386.
PMID 40084230 (verification gap: Publication bias in the depression analysis, and the evidence covers structured group programmes in adults over 55 - it does not test hobby engagement in general.)Systematic review and meta-analysis of controlled studies of group arts interventions in adults aged over 55 without dementia: 50 studies identified, 39 included, 36 contributing depression outcomes (n 3360) and 10 anxiety outcomes (n 949). · weight 1.0 · Group arts interventions reduced depression (d 0.70, 95% CI 0.54 to 0.87) and anxiety (d 0.76, 0.37 to 1.52) in older adults, with a larger reduction in care homes (d 1.07) than in the community (d 0.51). The authors report publication bias in the depression analysis.↑ back to the article
Evidence ledger for this entry
| Source | Type | Weight | PubMed |
|---|
| src-hb-01 | observational meta analysis | 0.5 | 37696932 |
| src-hb-02 | observational meta analysis | 0.5 | 25602273 |
| src-hb-03 | observational systematic review | 0.5 | 39956657 |
| src-hb-04 | systematic review meta analysis | 1.0 | 40084230 |
Verification in this run was record/abstract level via NCBI E-utilities. Full text
was not read and no retraction database was queried. Sources with a recorded verification gap are
flagged in place.