Skip to main content
NCBI home page
BMC Public Health logoLink to BMC Public Health
. 2026 Sep 8;26:2728. doi: 10.1186/s12889-026-29185-0

The mediating effects of depressive symptoms on the association between hobby engagement and functional disability: evidence from three cohort studies

Jia-min Yan 1,✉, Qi-qiang He 2,3
PMCID: PMC13628814  PMID: 42823764

Abstract

Background

Functional disability is a critical public health challenge in aging populations. While lifestyle factors have been extensively studied, the protective role of hobby engagement against functional decline remains underexplored. Thus, this study aimed to examine the association between hobby engagement and functional disability, as well as the potential mediating effects of depressive symptoms across three large international longitudinal cohort studies.

Methods

We analyzed data from the China Health and Retirement Longitudinal Study (CHARLS), the Survey of Health, Ageing and Retirement in Europe (SHARE), and the Mexican Health and Aging Study (MHAS). Hobby engagement and depressive symptoms were self-reported by participants. Functional disability was assessed using standardized Activities of Daily Living (ADL) and Instrumental Activities of Daily Living (IADL) measures. Cox proportional hazards models and General linear mixed models were employed to examine the associations between hobby engagement and functional disability incidence and progression. Mediation analysis was used to explore the role of depressive symptoms.

Results

Across all three cohorts, hobby engagement demonstrated consistent negative associations with both ADL and IADL disability. In the fully adjusted models, hobby engagement was associated with a reduced risk of ADL disability (CHARLS: HR = 0.77, 95% CI: 0.70–0.84; SHARE: HR = 0.73, 95% CI: 0.67–0.80; MHAS: HR = 0.82, 95% CI: 0.73–0.92) and IADL disability (CHARLS: HR = 0.70, 95% CI: 0.63–0.77; SHARE: HR = 0.65, 95% CI: 0.60–0.72; MHAS: HR = 0.68, 95% CI: 0.60–0.78). The mixed-effects models showed that hobby engagement was associated with a significantly slower rate of annual functional decline (For ADL, β for Time × Hobby interaction: -0.019 in CHARLS, -0.029 in SHARE, and − 0.025 in MHAS; For IADL, β for Time × Hobby interaction: -0.014 in CHARLS, -0.032 in SHARE, and − 0.027 in MHAS). The protective effect of hobby engagement was partially mediated by depressive symptoms, with the mediation proportion of 25.20% in CHARLS, 21.05% in SHARE, 32.25% in MHAS for ADL disability and 14.52% in CHARLS, 13.80% in SHARE, and 12.53% in MHAS for IADL disability.

Conclusion

Hobby engagement may protect against functional disability across diverse populations and cultural contexts. This effect was partially explained by the alleviation of depressive symptoms.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s12889-026-29185-0.

Keywords: Hobby engagement, Functional disability, Activities of daily living, Instrumental activities of daily living, Cohort study

Introduction

The global demographic landscape is undergoing profound change, characterized by a rapidly aging population [1, 2]. According to the United Nations, the number of individuals aged 65 and over is projected to more than double globally, from 761 million in 2021 to 1.6 billion by 2050 [3]. This shift has placed healthy aging, the maintenance of functional capacity to support well-being in later life, at the forefront of public health priority [4]. With this context, functional disability—manifested as limitations in essential daily activities, including Activities of Daily Living (ADL) and Instrumental Activities of Daily Living (IADL)—represents one of the most significant challenges facing aging populations worldwide [5, 6]. A recent meta-analysis estimated that the pooled prevalence of ADL disability was 26.07% among globally community-dwelling older adults, while IADL disability affected an even larger proportion at 46.15% [7]. With population age worldwide, the prevalence of functional limitations may continue to rise, imposing substantial burdens on individuals, families, and healthcare systems. The progressive loss of independence in essential daily activities not only compromises quality of life, but also precipitates increased healthcare utilization, institutionalization, and mortality risk [8]. Therefore, identifying modifiable factors that can prevent or delay functional decline has become a critical research priority.

Recent decades have witnessed a growing interest in the role of lifestyle factors in promoting healthy aging. While physical activity and dietary habits have been extensively studied, hobby engagement remains a broader and less clearly bounded construct that overlaps with leisure activities, social participation, and even sport, and that has been examined under various terms across studies. Hobbies represent voluntary, personally meaningful activities that individuals pursue for enjoyment and personal fulfillment [9, 10]. Unlike formal exercise programs or structured interventions, hobbies are self-selected, intrinsically motivated activities that may be more sustainable over time and across diverse populations [11]. Emerging research has documented the protective effects of hobby engagement on mental health, cognitive function, and mortality [12–14]. However, relatively few studies have examined its relationship with functional disability. One study of Japanese older adults reported that having hobbies was associated with a lower risk of ADL and IADL disability, but it was limited by a relatively short follow-up, a modest sample size, and a single national context [15]. Thus, it remains unclear whether this protective association is robust and generalizable across diverse populations and cultural contexts, and how hobby engagement relates to the long-term trajectory of functional decline.

Therefore, we analyzed the data from three large, nationally representative aging cohorts to examine the association between hobby engagement and incident ADL and IADL disability and the trajectory of functional decline over time, as well as the potential mediating role of depressive symptoms.

Methods

Study population

Data were obtained from three established longitudinal cohort studies: the China Health and Retirement Longitudinal Study (CHARLS), the Survey of Health, Ageing and Retirement in Europe (SHARE), and the Mexican Health and Aging Study (MHAS). All cohorts were nationally representative of middle-aged and older adults and collected a wide array of harmonized data on sociodemographic characteristics, health behaviors, physical and cognitive function, and history of chronic diseases. Specifically, CHARLS focused on individuals aged 45 and older in China, and both SHARE and MHAS focused on those aged 50 and older in Europe and Mexico, respectively. Comprehensive descriptions of the cohort profiles and study design have been published previously [16–18]. In this study, we used similar time ranges for all three cohorts: 2011–2020 for CHARLS (wave 1–5), 2010–2020 for SHARE (wave 4–8), and 2012–2021 for MHAS (wave 3–6).

For all three cohorts, participants were included if they were aged 50 or older, and the exclusion criteria included: (1) had a pre-existing functional disability, or lacked data on hobby engagement or other key covariates at baseline. (2) did not participate in any follow-up wave or information on functional disability was unavailable during the follow-up period. The detailed flowchart for participant selection is presented in Figure S1-S3. After applying these criteria, the final analytic samples comprised 7,058 participants in CHARLS, 33,142 in SHARE, and 8,756 in MHAS.

Hobby engagement assessment

There were slight variations for the hobby engagement assessment across the three cohorts [12–14]. In CHARLS, participants were asked whether they had engaged in any of the following five activities during the past month: leisure games (e.g., ma-jong, chess, cards), club participation (sports or social), involvement in community-related organizations, volunteer or charity work, and educational or training programs. In SHARE, hobby engagement was assessed by asking whether participants had participated in any of seven activities over the past 12 months: playing cards or chess, engaging in word or number games (e.g., crosswords, Sudoku), reading newspapers or books, club participation (sports or social), involvement in political or community-related organizations, volunteer or charity work, and educational or training programs. In MHAS, participants were asked whether they had engaged in any of eight activities during the past 12 months: playing cards, dominoes, or chess; engaging in word or number games (e.g., crosswords, jigsaw puzzles, Sudoku); reading newspapers or books; club participation (sports or social); volunteer or charity work; educational or training programs; engaging in crafts; and gardening. To ensure comparability across cohorts, hobby engagement was operationalized as a binary variable, with participants classified as engaged if they reported participation in at least one of the listed activities.

Functional disability assessment

Functional disability was assessed using standardized ADL and IADL measures across all three cohorts [19, 20]. The specific items in the three cohorts varied due to differences in cultural context and study design. To enhance cross-cohort comparability, we focused on the overlapping items: five core ADL tasks (dressing, bathing, eating, getting in and out of bed, and using the toilet) and four IADL tasks (preparing meals, shopping, taking medications, and managing money). Participants were defined as having ADL or IADL disability if they reported difficulty with at least one task in the corresponding domain. Disability scores were then calculated as the sum of reported difficulties, ranging from 0 to 5 for ADL and 0 to 4 for IADL.

Depressive symptoms

Depressive symptoms were assessed using validated instruments: the 10-item Center for Epidemiologic Studies Depression Scale (CESD-10) in CHARLS, the EURO-D scale in SHARE, and the 9-item Center for Epidemiologic Studies Depression Scale (CESD-9) in MHAS. The depression scores range from 0 to 30 for CHARLS, 0–12 for SHARE (EURO-D) and 0–9 for MHAS (CESD-9), with higher scores indicating more severe depressive symptoms. Despite the differences in the instruments used, all scales have been extensively validated in older adults [21–23]. In our baseline analytic sample, the depression instruments showed acceptable internal consistency, with Cronbach’s α of 0.78 for the CESD-10 (CHARLS), 0.65 for the EURO-D (SHARE), and 0.78 for the CESD-9 (MHAS). The lower value for the EURO-D is in line with previous reports, in which its Cronbach’s α has ranged from 0.62 to 0.78 across SHARE countries [24].

Covariates

Covariates were selected based on previous literature and included sociodemographic characteristics, health behavior, and health status. Sociodemographic variables included age, sex (male/female), education level (less than high school education and high school education or above), and marital status (married/partnership or single/divorced/widowed). Health behaviors comprised smoking status (current smoker, non-smoker), alcohol consumption (yes or no), body mass index (BMI, calculated as weight in kilograms divided by height in meters squared), and physical activity (vigorous or other). Health conditions included hypertension and diabetes mellitus (yes/no). Because the physical-activity questionnaires differed in detail across cohorts, physical activity was harmonized as a binary variable [25]. In CHARLS, participants were classified as vigorous if they reported any vigorous physical activity for at least 10 min per week. In SHARE, participants were classified as vigorous if they engaged in vigorous activity at least occasionally. In MHAS, participants were classified as vigorous if they reported hard physical activity, including sports, heavy household chores, or other physically demanding work, on average at least three times per week during the last two years. All remaining participants were classified as “other.”

Statistical analysis

Descriptive statistics were calculated for baseline characteristics, stratified by hobby engagement status within each cohort. Continuous variables were presented as means  (standard deviations), while categorical variables were presented as frequencies and percentages. Chi-square tests and t-tests were used to examine differences between groups, where appropriate.

Cox proportional hazards regression models were used to investigate the relationship between hobby engagement and functional disability. Follow-up time was calculated from the initial interview to the first occurrence of functional disability, loss to follow-up, or the end of the study period, whichever occurred first. Three sequential models were created: Model 1 adjusted for age and sex; Model 2 additionally adjusted for education, marital status, smoking, alcohol consumption, and BMI; and Model 3 additionally adjusted for physical activity, hypertension, and diabetes. To examine the relationship between changes in hobby engagement and functional disability risk, we defined four patterns based on hobby status at baseline and the first follow-up wave: persistent non-engagement (no hobbies at both baseline and the first follow-up), initiation of engagement (no hobbies at baseline but engagement at the first follow-up), cessation of engagement (engagement at baseline but discontinued at the first follow-up), and persistent engagement (engagement at both baseline and the first follow-up). Functional disability risk was then assessed from the second follow-up wave onward. Cox proportional hazards regression models were applied to estimate hazard ratios and 95% confidence intervals (CIs) for each pattern, with persistent non-engagement serving as the reference category. The assumption of proportional hazards was tested using the Schoenfeld residuals and was met for all models.

To investigate the trajectory of functional decline over time, we used generalized linear mixed-effects models with random slopes for participants. The outcome variables were ADL and IADL at each follow-up wave, treated as continuous measures. The primary fixed effects included time (a continuous variable representing years since the baseline interview), baseline hobby engagement status, and a Time × Hobby interaction term. This interaction term was specifically included to estimate the difference in the annual rate of change in disability scores between participants with hobby engagement and those without hobby engagement. We fitted three sequential models using the same covariate adjustment strategy as described previously.

To investigate the mediating role of depression underlying the association between hobby engagement and functional disability, we conducted mediation analyses using the CMAverse package in R, which implements the regression-based approach with survival outcomes [26]. The effect of baseline hobby engagement was assessed with depression scores measured at baseline as the mediator and subsequent functional disability as the outcome. Bootstrap resampling (n = 1000) was employed to construct 95% percentile CIs for all mediation parameters. Because the outcome was modeled on the hazard-ratio scale, the total effect (TE) was decomposed into a natural direct effect (NDE) and a natural indirect effect (NIE), all expressed as hazard ratios, and the proportion mediated was calculated following the regression-based approach for survival outcomes implemented in CMAverse package as Inline graphic. Several sensitivity analyses were conducted to evaluate the robustness of our findings. First, functional disability was redefined using the complete set of ADL and IADL items available in each cohort, thereby capturing cohort-specific operationalization of disability (Table S1). Second, participants with a history of stroke or physician-diagnosed arthritis at baseline were excluded to minimize potential reverse causation. Third, inverse probability of treatment weighting (IPTW) was applied to construct a pseudo-population in which baseline covariates were balanced between different hobby engagement groups to reduce the confounding effects by observed characteristics [27].

Results

Baseline characteristics of participants

The baseline characteristics of participants stratified by hobby engagement status across the three cohorts are presented in Table 1. The final samples included 7,058 participants from CHARLS, 33,142 from SHARE, and 8,756 from MHAS. The mean age of participants was 61.10 years (SD = 7.61) in CHARLS, 64.26 years (SD = 8.99) in SHARE, and 63.06 years (SD = 8.43) in MHAS. The proportion of hobby engagement varied substantially across the three cohorts: 26.4% (n = 1,862) in CHARLS, 88.8% (n = 29,439) in SHARE, and 88.5% (n = 7,747) in MHAS. In CHARLS, participants with hobby engagement were more likely to be male (58.5% vs. 50.2%, p < 0.001), have a high level of education (19.0% vs. 8.4%, p < 0.001), married or in a partnership (89.3% vs. 86.7%, p = 0.004), current smokers (38.0% vs. 32.2%, p < 0.001), and consume alcohol (40.6% vs. 33.1%, p < 0.001), compared with those without hobby engagement. They also had a higher prevalence of diabetes (7.6% vs. 5.4%, p < 0.001) and hypertension (28.1% vs. 24.9%, p = 0.008), as well as a higher BMI (24.09 kg/m2 vs. 23.11 kg/m2, p < 0.001). In SHARE, participants with hobby engagement were slightly younger (64.20 years vs. 64.76 years, p < 0.001) and more females (54.6% vs. 50.0%, p < 0.001). They were also more likely to have completed high school education or above (67.8% vs. 33.1%, p < 0.001), consume alcohol (91.9% vs. 75.0%, p < 0.001), and engage in vigorous physical activity, whereas those without hobby engagement had a higher prevalence of diabetes (13.4% vs. 10.1%, p < 0.001) and a higher BMI (27.28 kg/m2 vs. 26.69 kg/m2, p < 0.001). In MHAS, participants with hobby engagement were significantly younger (62.81 years vs. 64.98 years, p < 0.001), more likely to have high level of education (18.4% vs. 5.2%, p < 0.001), current smoker (13.6% vs. 10.1%, p = 0.002) and consume alcohol (28.1% vs. 21.3%, p < 0.001). They also had higher rates of vigorous physical activity (45.2% vs. 32.6%, p < 0.001) and a higher BMI (27.57 kg/m2 vs. 27.04 kg/m2, p = 0.001) than those without hobbies.

Table 1.

Baseline characteristics of study participants in CHARLS, SHARE, and MHAS

CHARLS SHARE MHAS
Variable All participants (N = 7058) No (N = 5196) Yes (N = 1862) p-value All participants (N = 33142) No (N = 3703) Yes (N = 29439) p-value All participants (N = 8756) No (N = 1009) Yes (N = 7747) p-value
Age (years) 61.10 (7.61) 61.13 (7.57) 60.99 (7.74) 0.500 64.26 (8.99) 64.76 (9.47) 64.20 (8.93) < 0.001 63.06 (8.43) 64.98 (9.30) 62.81 (8.28) < 0.001
Sex, n (%) < 0.001 < 0.001 0.362
 Male 3,696 (52.4) 2,607 (50.2) 1,089 (58.5) 15,222 (45.9) 1,853 (50.0) 13,369 (45.4) 4,052 (46.3) 481 (47.7) 3,571 (46.1)
 Female 3,362 (47.6) 2,589 (49.8) 773 (41.5) 17,920 (54.1) 1,850 (50.0) 16,070 (54.6) 4,704 (53.7) 528 (52.3) 4,176 (53.9)
Education level, n (%) < 0.001 < 0.001 < 0.001
 Less than high school education 6,271 (88.8) 4,762 (91.6) 1,509 (81.0) 11,950 (36.1) 2,476 (66.9) 9,474 (32.2) 7,279 (83.1) 957 (94.8) 6,322 (81.6)
 High school education or above 787 (11.2) 434 (8.4) 353 (19.0) 21,192 (63.9) 1,227 (33.1) 19,965 (67.8) 1,477 (16.9) 52 (5.2) 1,425 (18.4)
Marital status, n (%) 0.004 < 0.001 0.152
 Married/partnership 6,168 (87.4) 4,505 (86.7) 1,663 (89.3) 23,769 (71.7) 2,774 (74.9) 20,995 (71.3) 6,383 (72.9) 716 (71.0) 5,667 (73.2)
 Single/divorced/widowed 890 (12.6) 691 (13.3) 199 (10.7) 9,373 (28.3) 929 (25.1) 8,444 (28.7) 2,373 (27.1) 293 (29.0) 2,080 (26.8)
Smoking status, n (%) < 0.001 0.016 0.002
 Non-smoker 4,679 (66.3) 3,525 (67.8) 1,154 (62.0) 26,760 (80.7) 2,935 (79.3) 23,825 (80.9) 7,599 (86.8) 907 (89.9) 6,692 (86.4)
 Current smoker 2,379 (33.7) 1,671 (32.2) 708 (38.0) 6,382 (19.3) 768 (20.7) 5,614 (19.1) 1,157 (13.2) 102 (10.1) 1,055 (13.6)
Alcohol consumption, n (%) < 0.001 < 0.001 < 0.001
 No 4,580 (64.9) 3,474 (66.9) 1,106 (59.4) 3,308 (10.0) 925 (25.0) 2,383 (8.1) 6,361 (72.6) 794 (78.7) 5,567 (71.9)
 Yes 2,478 (35.1) 1,722 (33.1) 756 (40.6) 29,834 (90.0) 2,778 (75.0) 27,056 (91.9) 2,395 (27.4) 215 (21.3) 2,180 (28.1)
Physical activity, n (%) < 0.001 < 0.001 < 0.001
 Other 5,974 (84.6) 4,309 (82.9) 1,665 (89.4) 11,965 (36.1) 1,718 (46.4) 10,247 (34.8) 4,924 (56.2) 680 (67.4) 4,244 (54.8)
 Vigorous 1,084 (15.4) 887 (17.1) 197 (10.6) 21,177 (63.9) 1,985 (53.6) 19,192 (65.2) 3,832 (43.8) 329 (32.6) 3,503 (45.2)
Diabetes, n (%) 419 (5.9) 278 (5.4) 141 (7.6) < 0.001 3,482 (10.5) 495 (13.4) 2,987 (10.1) < 0.001 1,866 (21.3) 237 (23.5) 1,629 (21.0) 0.079
Hypertension, n (%) 1,818 (25.8) 1,295 (24.9) 523 (28.1) 0.008 13,109 (39.6) 1,497 (40.4) 11,612 (39.4) 0.257 4,163 (47.5) 494 (49.0) 3,669 (47.4) 0.356
BMI (kg/m2) 23.37 (3.83) 23.11 (3.78) 24.09 (3.89) < 0.001 26.76 (4.51) 27.28 (4.68) 26.69 (4.49) < 0.001 27.51 (4.67) 27.04 (4.93) 27.57 (4.64) 0.001
Open in a new tab

Association between hobby engagement and functional disability

Table S2 and Fig. 1 present the hazard ratios for the association between hobby engagement and ADL and IADL disability across the three cohorts. As illustrated in Fig. 1, baseline hobby engagement was significantly associated with a reduced risk of ADL and IADL disability in all three cohorts. For ADL disability, fully adjusted models revealed a 23% reduction in CHARLS (HR = 0.77, 95% CI: 0.70–0.84), 27% in SHARE (HR = 0.73, 95% CI: 0.67–0.80), and 18% in MHAS (HR = 0.82, 95% CI: 0.73–0.92). The protective effects were more pronounced for IADL disability, with a risk reduction of 30% in CHARLS (HR = 0.70, 95% CI: 0.63–0.77), 35% in SHARE (HR = 0.65, 95% CI: 0.60–0.72), and 32% in MHAS (HR = 0.68, 95% CI: 0.60–0.78).

Fig. 1.

Fig. 1

Open in a new tab

Forest plot of hazard ratios for the association between baseline hobby engagement and Incident ADL and IADL disability

Patterns of hobby engagement change and functional disability

Table S3 and Fig. 2 show the association between changes in hobby engagement patterns and functional disability risk. As illustrated in Fig. 2, persistent hobby engagement demonstrated the strongest associations across all cohorts. Compared to persistent non-engagement, persistent engagement was associated with 25% decreased risk (HR: 0.75, 95% CI: 0.66–0.85) for ADL disability and 35% decreased risk (HR: 0.65, 95% CI: 0.57–0.74) for IADL disability in CHARLS. In SHARE, the corresponding HR was 0.76 (95% CI: 0.65–0.89) for ADL disability and 0.58 (95% CI: 0.50–0.69) for IADL disability, respectively. In MHAS, persistent engagement was associated with 22% decreased risk (HR: 0.78, 95% CI: 0.61–0.99) for ADL disability and 41% decreased risk (HR: 0.59, 95% CI: 0.46–0.75) for IADL disability, respectively.

Fig. 2.

Fig. 2

Open in a new tab

Forest plot of hazard ratios for the association between patterns of hobby engagement change and incident functional disability

Hobby engagement and changes in ADL and IADL scores over time

Table S4 presents the results examining hobby engagement and the annual rate of functional decline over time. In all cohorts, a significant negative Time × Hobby interaction was observed for both ADL and IADL scores, indicating a slower increase in functional disability scores per year for the participants with hobby engagement. Specifically, the annual change in ADL score was − 0.019 (95% CI, -0.026 to -0.011) in CHARLS, -0.029 (95% CI, -0.034 to -0.024) in SHARE, and − 0.025 (95% CI, -0.033 to -0.017) in MHAS for participants with hobby engagement compared to those without in the fully adjusted model. The annual change in IADL score was − 0.014 (95% CI, -0.021 to -0.008) in CHARLS, -0.032 (95% CI, -0.036 to -0.027) in SHARE, and − 0.027 (95% CI, -0.034 to -0.020) in MHAS for participants with hobby engagement compared to those without in the fully adjusted model. Figure 3 visualizes the rates of functional disability scores over time by hobby engagement estimated from Model 3. The protective effect of hobby engagement on the trajectory of functional decline was consistent across all three cohorts.

Fig. 3.

Fig. 3

Open in a new tab

Predicted trajectories of ADL and IADL scores by hobby engagement status across three cohorts

The mediating role of depressive symptoms

Table S5 and Fig. 4 present the results examining the role of depression in the association between hobby engagement and functional disability. The mediation analysis revealed that depressive symptoms were a significant pathway linking hobby engagement to functional disability across all cohorts. The protective effect of hobbies was partially explained by reducing depressive symptoms. The proportion of the total effect on ADL disability mediated by depressive symptoms was 25.20% in CHARLS, 21.05% in SHARE, and 32.25% in MHAS, respectively. For IADL disability, the mediated proportions were 14.52%, 13.80%, and 12.53% in CHARLS, SHARE, and MHAS, respectively.

Fig. 4.

Fig. 4

Open in a new tab

Mediation analysis of depression scores in the association between hobby engagement and functional disability across three cohorts

Subgroup analyses

Subgroup analyses show the potential effect of hobby engagement with functional disability by sex and age (Table S6). The results were generally consistent across age groups in all cohorts, while no significant interactions were observed for ADL and IADL disability (all p for interaction > 0.05).

In contrast, sex-stratified analyses revealed significant heterogeneity for ADL disability in the SHARE and MHAS cohorts. Specifically, the protective effect of hobby engagement against ADL disability was significantly stronger in women (HR: 0.67, 95% CI: 0.60–0.75) than in men (HR: 0.80, 95% CI: 0.71–0.90; p for interaction: 0.005) in the SHARE cohort. Conversely, in the MHAS cohort, the protective effect against ADL disability was evident only in men (HR: 0.73, 95% CI: 0.61–0.86), while not statistically significant in women (HR: 0.90, 95% CI: 0.77–1.05; p for interaction: 0.025). No significant interaction by sex was observed in the CHARLS cohort.

Sensitivity analyses

Several sensitivity analyses were conducted to assess the robustness of the primary findings (Table S7-S9). First, when functional disability was redefined using the complete set of ADL and IADL items available in each cohort, hobby engagement remained significantly associated with lower risk of ADL and IADL. The HR (95% CI) for ADL disability was 0.78 (0.71–0.86) in CHARLS, 0.75 (0.68–0.82) in SHARE, and 0.81 (0.73–0.91) in MHAS. For IADL disability, the corresponding HR (95% CI) was 0.74 (0.67–0.81), 0.72 (0.67–0.78), and 0.69 (0.60–0.79), respectively. Second, after further excluding participants with a history of stroke or arthritis, the results did not change significantly, with HR of 0.77 (0.68–0.88) in CHARLS, 0.70 (0.62–0.78) in SHARE, and 0.85 (0.73–0.98) in MHAS for ADL disability, and 0.69 (0.61–0.78), 0.60 (0.54–0.68), and 0.72 (0.61–0.85) for IADL disability. Third, when inverse probability of treatment weighting (IPTW) was applied to balance baseline covariates, the associations remained consistent, with HR of 0.78 (0.71–0.86) in CHARLS, 0.68 (0.61–0.75) in SHARE, and 0.81 (0.71–0.93) in MHAS for ADL disability, and 0.71 (0.65–0.78), 0.62 (0.55–0.69), and 0.73 (0.63–0.85) for IADL disability.

Discussion

In this large-scale study of three international longitudinal cohorts, we found that hobby engagement was consistently associated with a reduced risk of functional disability and a slower trajectory of functional decline among adults aged 50 years and older. The protective effects were robust across culturally diverse populations from China, Europe, and Mexico, with risk reductions ranging from 18 to 27% for ADL disability and 30–35% for IADL disability. Our results revealed that persistent engagement in hobbies could confer substantial protection against both ADL and IADL disability. Furthermore, the beneficial effects of hobby engagement on functional status were partially explained by a reduction in depressive symptoms.

Our findings are consistent with previous research demonstrating protective effects of hobby engagement on multiple health outcomes. For instance, a prospective study of 469 elderly people showed that frequent participation in hobbies (reading, playing board games, playing musical instruments, dancing) was associated with a significantly lower risk of developing dementia [28]. Similarly, a cohort study of 56,000 Japanese adults found that hobby engagement was associated with a significantly decreased risk of cardiovascular disease during a 10-year follow up period [29]. Another cohort study of 3,583 individuals showed that hobby engagement was associated with a decreased incidence of frailty, a strong precursor to functional disability [30]. Although a previous study found that having hobbies was associated with a lower risk of ADL and IADL disability among Japanese older adults [15], this study relied on a three-year follow-up and relatively small sample sizes (about one thousand), which may be insufficient to fully capture the long-term development of functional disability. Our findings not only extend the existing literature by showing that hobby engagement was a strong predictor of functional ability maintenance, but also assess the associations using trajectory models over time, providing a more nuanced understanding of how functional capacity evolves dynamically over the aging process. Further, the consistency of our findings across CHARLS, SHARE and MHAS suggests that the benefits of hobbies are not limited to a particular cultural or socioeconomic context, but represent a potentially universal effect for promoting healthy aging. The subgroup analyses showed significant sex-specific effects of the hobby engagement that appeared to vary by cultural context. The benefits of hobby engagement in preventing ADL disability were more pronounced among women in the European cohort, whereas they were more pronounced among men in the Mexican cohort. This may reflect the differences in gender roles, access to hobbies, and societal attitudes toward leisure activities [31, 32]. Furthermore, the consistency of protective effects across age groups suggests that it is important to implement hobby promotion strategies across the entire spectrum of older adulthood rather than targeting specific age groups.

The mechanisms underlying the protective effect of hobby engagement on functional disability are likely multifactorial. First, many hobbies such as puzzles, reading or chess involve cognitive stimulation that can improve cognitive reserve and protect against cognitive decline, which is closely associated with functional independence [33–35]. Second, hobbies often incorporate elements of physical activity like gardening or sports. These physical components may directly contribute to maintaining muscle strength, coordination, and balance necessary for ADL and IADL performance [36–38]. Third, the social engagement inherent in many hobbies such as participation in clubs, group games, or community activities, may reduce social isolation and strengthen social support networks, which are significant associated with functional disability in older adults [39–41]. Such social connections can provide emotional support and practical assistance, thereby facilitating the maintenance of independence in daily life.

Beyond these pathways, the mediation analysis revealed that depressive symptoms might partially explain the relationship between hobby engagement and functional disability. Core features of depression, such as anhedonia, persistent fatigue, and decreased motivation, can lead to behavioral changes such as reduced physical activity, social withdrawal, and neglect of essential self-care behaviors (e.g., adequate nutrition and medication adherence). These behavioral changes can accelerate the loss of muscle mass and physiological reserve, thereby increasing the risk of frailty and limitations in ADL and IADL [42, 43]. Furthermore, depression is associated with cognitive deficits, particularly in executive function and attention, which can directly impair an older adult’s ability to perform complex IADLs, such as managing finances or medications [44]. In contrast, engaging in hobbies can protect against depression by providing structure, purpose, and opportunities for mastery, as well as strengthening social relationships [45, 46]. These psychosocial benefits can help maintain motivation, sustain daily functioning, and break the vicious cycle in which depression and disability reinforce each other. Thus, the mediating role of depression highlights an important psychological pathway through which hobby engagement promotes healthy aging. This suggests that interventions targeting involvement in hobbies may be a promising strategy for reducing disability in late life, in part by alleviating depressive symptoms.

The prevalence of hobby engagement differed substantially across the three cohorts, ranging from 26.4% in CHARLS to approximately 89% in SHARE and MHAS. Beyond genuine cultural differences in leisure participation, this disparity likely reflects how hobby engagement was assessed, including the number and types of activities queried and the length of the recall period. CHARLS asked about five activities during the past month, whereas SHARE and MHAS used broader inventories of seven and eight activities over the past 12 months, respectively. The absolute prevalence of hobby engagement is therefore not directly comparable across cohorts and should be interpreted with caution. Nevertheless, the protective association between hobby engagement and functional disability was consistent in direction and magnitude across all three cohorts despite these markedly different baseline prevalences. This consistency suggests that the observed associations were not driven by differences in how hobby engagement was captured. This study has several strengths. The multi-cohort design incorporating three large, internationally recognized studies provides exceptional external validity and enables examination of associations across diverse populations with varying cultural backgrounds, healthcare systems, and socioeconomic contexts. The longitudinal follow-up periods spanning multiple waves of data collection allow for robust examination of temporal relationships and trajectory analyses. Nevertheless, several limitations should be acknowledged. First, hobby engagement and functional disability were assessed through self-reported questionnaires that may be susceptible to recall bias. Second, the measurement of hobby engagement was based on a dichotomous variable. This approach lacks granularity regarding the type, frequency, intensity, or duration of the activities. Third, the observational design cannot definitively establish causality. Although participants with functional disability at baseline were excluded and the results were consistent in sensitivity analyses that further excluded participants with stroke or arthritis, healthier individuals may be both more likely to engage in hobbies and less likely to develop functional disability. Consequently, the observed association may be partly explained by better underlying health status rather than by hobby engagement alone. Fourth, because both hobby engagement and depressive symptoms were measured at baseline, the temporal ordering required for a causal mediation analysis was not fully established. The mediation results should therefore be interpreted as suggesting that depressive symptoms may partly explain the observed association rather than as evidence of a causal mediating effect. Fifth, residual confounding also remains possible because the models did not include cognitive function, socioeconomic indicators beyond education, chronic pain, cardiovascular or pulmonary disease, or baseline frailty. Finally, the inclusion of adults aged 50 years and older from three cohorts may limit the generalizability of our findings to younger populations and other national contexts.

Conclusion

In summary, this multi-cohort study provides compelling evidence that hobby engagement was a robust and independent protective factor against the development of functional disability in aging populations across different cultural settings. The benefits of hobbies were partly mediated by an improvement in depressive symptoms. The findings from this study have significant implications for public health policy. Hobby engagement represents a low-cost, accessible, and highly modifiable lifestyle behavior that would preserve functional independence in later life. Interventions to promote hobby engagement may be an effective strategy for healthy aging and reducing the societal burden of functional disability.

Supplementary Information

12889_2026_29185_MOESM1_ESM.docx (6.6MB, docx)

Supplementary Material 1: Figure S1 Flowchart of participant selection in CHARLS. Figure S2 Flowchart of participant selection in SHARE. Figure S3 Flowchart of participant selection in MHAS.

Acknowledgements

We are grateful to the research teams of CHARLS, SHARE and MHAS for providing the public-use data employed in this study. We also thank the Gateway to Global Aging Data team for providing the harmonized datasets that facilitated cross-national analyses.

Authors' contributions

Jia-min Yan: Writing - original draft, Writing - review & editing, Methodology, Formal analysis; Qi-qiang He: Writing-review & editing, Supervision.

Funding

Not applicable.

Data availability

The datasets analyzed in this study are publicly available through the following sources: CHARLS data are available at http://charls.pku.edu.cn/en; SHARE data are available at http://www.share-project.org/; and MHAS data are available at https://www.mhasweb.org/.

Declarations

Ethics approval and consent to participate

This study utilized secondary data from three established longitudinal cohort studies: the China Health and Retirement Longitudinal Study (CHARLS), the Survey of Health, Ageing and Retirement in Europe (SHARE), and the Mexican Health and Aging Study (MHAS). All three studies were conducted in accordance with the Declaration of Helsinki and obtained ethical approval from their respective institutional review boards and ethics committees prior to data collection (Table S10). Written informed consent was obtained from all participants in the original studies.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Footnotes

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

References

  • 1.Jane Osareme O, Muonde M, Maduka CP, Olorunsogo TO, Omotayo O. Demographic shifts and healthcare: A review of aging populations and systemic challenges. Int J Sci Res Arch. 2024;11:383–95. [Google Scholar]
  • 2.Fuster V. Changing demographics: a new approach to global health care due to the aging population. J Am Coll Cardiol. 2017;69(24):3002–5. [DOI] [PubMed]
  • 3.Reuter A, Bärnighausen T, Kohler S. Global Ageing and Health. In: Bloom DE, Sousa-Poza A, Sunde U, editors. The Routledge Handbook of the Economics of Ageing. London and New York: Routledge; 2024. Chapter 37. eBook ISBN: 978-1-003-15039-8. 10.4324/9781003150398-44. [DOI]
  • 4.World Health Organization. World report on ageing and health. Geneva: World Health Organization; 2015. ISBN: 978-92-4-156504-2.
  • 5.Carmona-Torres JM, Rodríguez-Borrego MA, Laredo-Aguilera JA, López-Soto PJ, Santacruz-Salas E, Cobo-Cuenca AI. Disability for basic and instrumental activities of daily living in older individuals. PLoS ONE. 2019;14(7):e0220157. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Millán-Calenti JC, Tubío J, Pita-Fernández S, González-Abraldes I, Lorenzo T, Fernández-Arruty T, Maseda A. Prevalence of functional disability in activities of daily living (ADL), instrumental activities of daily living (IADL) and associated factors, as predictors of morbidity and mortality. Arch Gerontol Geriatr. 2010;50(3):306–10. [DOI] [PubMed] [Google Scholar]
  • 7.Amlak BT, Getinet M, Getie A, Kebede WM, Tarekegn TT, Belay DG. Functional disability in basic and instrumental activities of daily living among older adults globally: a systematic review and meta-analysis. BMC Geriatr. 2025;25(1):413. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Covinsky KE, Palmer RM, Fortinsky RH, Counsell SR, Stewart AL, Kresevic D, Burant CJ, Landefeld CS. Loss of independence in activities of daily living in older adults hospitalized with medical illnesses: increased vulnerability with age. J Am Geriatr Soc. 2003;51(4):451–8. [DOI] [PubMed] [Google Scholar]
  • 9.Alanazi HMN. The role of leisure activities in enhancing well-being in Saudi’s retired community: A mixed methods study. Humanit Social Sci Commun. 2024;11(1):1–17. [Google Scholar]
  • 10.Iso-Ahola SE, Baumeister RF. Leisure and meaning in life. Front Psychol. 2023;14:1074649. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Fancourt D, Aughterson H, Finn S, Walker E, Steptoe A. How leisure activities affect health: a narrative review and multi-level theoretical framework of mechanisms of action. Lancet Psychiatry. 2021;8(4):329–39. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.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):2233–40. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Huang W, Xiao Q, Li Z, Chen Y, Wang X, Liu Q. Positive association between hobby participation and objective and subjective cognition among adults aged 50 years and over in 24 countries. Soc Sci Med. 2025;383:118487. [DOI] [PubMed] [Google Scholar]
  • 14.Guo Y, Yang F. Hobby engagement and all-cause and cause-specific mortality risk among people aged 50 years and older in 19 countries. J Glob Health. 2025;15:04181. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Tomioka K, Kurumatani N, Hosoi H. Relationship of Having Hobbies and a Purpose in Life With Mortality, Activities of Daily Living, and Instrumental Activities of Daily Living Among Community-Dwelling Elderly Adults. J Epidemiol. 2016;26(7):361–70. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Zhao Y, Hu Y, Smith JP, Strauss J, Yang G. Cohort profile: the China Health and Retirement Longitudinal Study (CHARLS). Int J Epidemiol. 2014;43(1):61–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Borsch-Supan A, Brandt M, Hunkler C, Kneip T, Korbmacher J, Malter F, Schaan B, Stuck S, Zuber S, Team SCC. Data Resource Profile: the Survey of Health, Ageing and Retirement in Europe (SHARE). Int J Epidemiol. 2013;42(4):992–1001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Wong R, Michaels-Obregon A, Palloni A. Cohort Profile: The Mexican Health and Aging Study (MHAS). Int J Epidemiol. 2017;46(2):e2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Katz S, Ford AB, Moskowitz RW, Jackson BA, Jaffe MW. Studies of illness in the aged: the index of ADL: a standardized measure of biological and psychosocial function. JAMA. 1963;185(12):914–9. [DOI] [PubMed] [Google Scholar]
  • 20.Lawton MP, Brody EM. Assessment of older people: self-maintaining and instrumental activities of daily living. Gerontologist. 1969;9(3):179–86. [PubMed]
  • 21.Courtin E, Knapp M, Grundy E, Avendano-Pabon M. Are different measures of depressive symptoms in old age comparable? An analysis of the CES-D and Euro-D scales in 13 countries. Int J Methods Psychiatr Res. 2015;24(4):287–304. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Wang Y, Liu M, Yang F, Chen H, Wang Y, Liu J. The associations of socioeconomic status, social activities, and loneliness with depressive symptoms in adults aged 50 years and older across 24 countries: findings from five prospective cohort studies. Lancet Healthy Longev. 2024;5(9):100618. [DOI] [PubMed] [Google Scholar]
  • 23.Yan R, Liu X, Xue R, Duan X, Li L, He X, Cui F, Zhao J. Association between internet exclusion and depressive symptoms among older adults: panel data analysis of five longitudinal cohort studies. EClinicalMedicine. 2024;75:102767. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Guerra M, Ferri C, Llibre J, Prina AM, Prince M. Psychometric properties of EURO-D, a geriatric depression scale: a cross-cultural validation study. BMC Psychiatry. 2015;15:12. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Zhang Z, Xu H, Zhang R, Yan Y, Ling X, Meng Y, Zhang X, Wang Y. Frailty and depressive symptoms in relation to cardiovascular disease risk in middle-aged and older adults. Nat Commun. 2025;16(1):6008. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Shi B, Choirat C, Coull BA, VanderWeele TJ, Valeri L. CMAverse: A Suite of Functions for Reproducible Causal Mediation Analyses. Epidemiology. 2021;32(5):e20–2. [DOI] [PubMed] [Google Scholar]
  • 27.Austin PC. An Introduction to Propensity Score Methods for Reducing the Effects of Confounding in Observational Studies. Multivar Behav Res. 2011;46(3):399–424. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Verghese J, Lipton RB, Katz MJ, Hall CB, Derby CA, Kuslansky G, Ambrose AF, Sliwinski M, Buschke H. Leisure activities and the risk of dementia in the elderly. N Engl J Med. 2003;348(25):2508–16. [DOI] [PubMed] [Google Scholar]
  • 29.Wang X, Dong JY, Shirai K, Yamagishi K, Kokubo Y, Saito I, Yatsuya H, Iso H, Tsugane S, Sawada N, et al. Having hobbies and the risk of cardiovascular disease incidence: A Japan public health center-based study. Atherosclerosis. 2021;335:1–7. [DOI] [PubMed] [Google Scholar]
  • 30.Fushiki Y, Ohnishi H, Sakauchi F, Oura A, Mori M. Relationship of hobby activities with mortality and frailty among community-dwelling elderly adults: results of a follow-up study in Japan. J Epidemiol. 2012;22(4):340–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Avital D. Gender differences in leisure patterns at age 50 and above: Micro and macro aspects. Ageing Soc. 2017;37(1):139–66. [Google Scholar]
  • 32.Ong CH, Pham BL, Levasseur M, Tan GR, Seah B. Sex and gender differences in social participation among community-dwelling older adults: a systematic review. Front Public Health. 2024;12:1335692. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Schultz SA, Larson J, Oh J, Koscik R, Dowling MN, Gallagher CL, Carlsson CM, Rowley HA, Bendlin BB, Asthana S, et al. Participation in cognitively-stimulating activities is associated with brain structure and cognitive function in preclinical Alzheimer’s disease. Brain Imaging Behav. 2015;9(4):729–36. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Lindbergh CA, Dishman RK, Miller LS. Functional Disability in Mild Cognitive Impairment: A Systematic Review and Meta-Analysis. Neuropsychol Rev. 2016;26(2):129–59. [DOI] [PubMed] [Google Scholar]
  • 35.McGuire LC, Ford ES, Ajani UA. Cognitive functioning as a predictor of functional disability in later life. Am J Geriatr Psychiatry. 2006;14(1):36–42. [DOI] [PubMed] [Google Scholar]
  • 36.Liu C-j, Shiroy DM, Jones LY, Clark DO. Systematic review of functional training on muscle strength, physical functioning, and activities of daily living in older adults. Eur Rev aging Phys activity. 2014;11(2):95–106. [Google Scholar]
  • 37.Wang DXM, Yao J, Zirek Y, Reijnierse EM, Maier AB. Muscle mass, strength, and physical performance predicting activities of daily living: a meta-analysis. J Cachexia Sarcopenia Muscle. 2020;11(1):3–25. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Roberts CE, Phillips LH, Cooper CL, Gray S, Allan JL. Effect of Different Types of Physical Activity on Activities of Daily Living in Older Adults: Systematic Review and Meta-Analysis. J Aging Phys Act. 2017;25(4):653–70. [DOI] [PubMed] [Google Scholar]
  • 39.Lai ET, Chau AKC, Ho IY, Hashimoto H, Kim CY, Chiang TL, Chen YM, Marmot M, Woo J. The impact of social isolation on functional disability in older people: A multi-cohort study. Arch Gerontol Geriatr. 2024;125:105502. [DOI] [PubMed] [Google Scholar]
  • 40.Guo L, An L, Luo F, Yu B. Social isolation, loneliness and functional disability in Chinese older women and men: a longitudinal study. Age Ageing. 2021;50(4):1222–8. [DOI] [PubMed] [Google Scholar]
  • 41.Jiang M, Li X, Lu Y. Social isolation, loneliness, and functional disability in Chinese older women and men: a longitudinal cross-lag analysis. BMC Psychol. 2024;12(1):674. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Lenze EJ, Rogers JC, Martire LM, Mulsant BH, Rollman BL, Dew MA, Schulz R, Reynolds CF 3. The association of late-life depression and anxiety with physical disability: a review of the literature and prospectus for future research. Am J Geriatr Psychiatry. 2001;9(2):113–35. [PubMed] [Google Scholar]
  • 43.Ormel J, Rijsdijk FV, Sullivan M, van Sonderen E, Kempen GI. Temporal and reciprocal relationship between IADL/ADL disability and depressive symptoms in late life. J Gerontol B Psychol Sci Soc Sci. 2002;57(4):P338–347. [DOI] [PubMed] [Google Scholar]
  • 44.Yen YC, Rebok GW, Gallo JJ, Jones RN, Tennstedt SL. Depressive symptoms impair everyday problem-solving ability through cognitive abilities in late life. Am J Geriatr Psychiatry. 2011;19(2):142–50. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45.Cleary M, Le Lagadec D, Thapa DK, Kornhaber R. Exploring the Impact of Hobbies on Mental Health and Well-Being: A Scoping Review. Issues Ment Health Nurs. 2025;46(8):804–14. [DOI] [PubMed] [Google Scholar]
  • 46.Fancourt D, Opher S, de Oliveira C. Fixed-Effects Analyses of Time-Varying Associations between Hobbies and Depression in a Longitudinal Cohort Study: Support for Social Prescribing? Psychother Psychosom. 2020;89(2):111–3. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

12889_2026_29185_MOESM1_ESM.docx (6.6MB, docx)

Supplementary Material 1: Figure S1 Flowchart of participant selection in CHARLS. Figure S2 Flowchart of participant selection in SHARE. Figure S3 Flowchart of participant selection in MHAS.

Data Availability Statement

The datasets analyzed in this study are publicly available through the following sources: CHARLS data are available at http://charls.pku.edu.cn/en; SHARE data are available at http://www.share-project.org/; and MHAS data are available at https://www.mhasweb.org/.

Articles from BMC Public Health are provided here courtesy of BMC

RESOURCES