Mediating role of psychological resilience in the relationship between leisure-time activities and long-term mortality in older adults

Abstract

Background

Participation in leisure-time activities (LTAs) is associated with lower mortality; however, the related mechanisms have not been fully elucidated. The aim of this study was to explore the mediating role of psychological resilience in the relationship between LTAs and all-cause mortality in community-dwelling older adults.

Methods

A total of 33,679 participants who were aged ≥ 60 in the Chinese Longitudinal Healthy Longevity Study from 1998 to 2018 were recruited. The frequency of seven typical LTAs and psychological resilience levels were collected using standard questionnaires at baseline. The mediating role of psychological resilience between LTAs and all-cause mortality was explored using mediated analyses in Cox proportional hazard models.

Results

During 150,208.03 person-years of follow-up, 24,705 deaths were documented. Older adults who engaged in LTAs more frequently had higher levels of psychological resilience. After adjusting for socio-demographics and health-related factors, we observed linear negative associations between leisure activities (P-nonlinear = 0.192) and psychological resilience (P-nonlinear = 0.308) with all-cause mortality in the total sample, respectively. The relationship between LTAs and all-cause mortality was partially mediated by psychological resilience, accounting for 13.61% (95%CI: 11.91–15.31%) of the association. Higher mediated proportions were observed for social activities than for productive activities. The mediating effect remained stable in sensitivity analyses and was more pronounced in the oldest-old (80 + years) and women subgroups.

Conclusions

Psychological resilience mediated the association between LTAs and all-cause mortality. Frequent participation in leisure activities may play a beneficial role in preventing premature death by improving resilience levels in community-dwelling older people.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12877-025-06496-2.

Background

As one of the most populous countries in the world, China is experiencing an accelerating and significant aging problem, and already has the largest population of older people globally [1]. According to the National Bureau of Statistics of China, the number of individuals aged 60 or older in China reached 264 million in 2020, accounting for 18.7% of the total population. By 2050, an estimated one in three individuals in China is projected to surpass the age of 60 [2]. Global mortality rates and life expectancy differ significantly across countries and regions [3]. In particular, the mortality rates among older adults in low- and middle-income countries (LMICs) are markedly higher than those observed in the United States and Europe [4]. To address this escalating aging population in China, it is necessary to further explore the beneficial role of modifiable healthy lifestyles in preventing premature death among older individuals. Prioritizing the encouragement of participation in leisure-time activities (LTAs) among older adults is crucial in efforts aimed at supporting healthy aging, yet evidence from LMICs is relatively scarce.

Leisure activities have significant benefits for older adults, particularly after they retire. Engaging in leisure activities not only provides physical exercise but also promotes social interaction, cognitive stimulation, and emotional well-being [5, 6]. Previous longitudinal studies conducted across diverse countries such as the United States, Sweden, Israel, Japan, and China have shown that engagement in various LTAs, including physical, productive, and cognitive activities, was associated with a reduced mortality risk [7–12]. Evidence from the National Health and Nutrition Examination Survey indicated that the negative association between engagement in leisure activities and risk of mortality existed across all age groups and was strongest in very old individuals (80 + years) [9]. Despite numerous studies confirming the positive impact of LTAs on reducing mortality, few of them delved into the underlying mechanisms. Over the past two decades, various theories from distinct disciplines (psychological, biological, environmental, behavioral, etc.) have emerged to elucidate the influence of leisure activities on health across different levels, encompassing individual activities and social group interactions. Psychological processes represent one of the pathways through which engagement in leisure activities can potentially benefit health at the micro-level [13]. Hence, further investigation into whether psychological factors mediate the relationship between leisure activities and mortality risk in population-based studies is warranted.

Psychological resilience is a concept which refers to the ability to adapt and cope effectively with significant adversity, trauma, or stress, and bounce back to a state of mental well-being [14]. Based on previous research, the dimensions of resilience in older adults can be divided into self-identity, maintaining optimism, adapting to the environment, and positively coping with aging [15–17]. Longitudinal studies in recent years have demonstrated that high resilience levels were associated with lower risk of all-cause mortality in community settings [18] and cancer patients [19, 20]. Recently, Daisy and colleagues have proposed a new multi-level theoretical framework, in which resilience was considered as one of the core mechanisms [13]. According to Daisy et al.‘s theory, leisure activities not only help individuals build psychological resilience but also enhance social support at the group level [13]. Previous studies have established the link between specific leisure activities and various dimensions of psychological resilience. For instance, a recent systematic review indicated that engagement in leisure hobbies correlated with higher levels of happiness and life satisfaction [21]. Similarly, a community-based study in Japan revealed that greater engagement in leisure activities, especially outdoor activities, was significantly associated with a lower likelihood of psychological distress [22].

Therefore, based on the theoretical association between leisure activities and psychological resilience, as well as previous research findings, we propose that psychological resilience potentially mediates the impact of leisure activities on all-cause mortality. Using a large-scale sample of community-dwelling older individuals from the Chinese Longitudinal Healthy Longevity Survey (CLHLS), this study aimed to test this hypothesis to better understand the significance of leisure activities for healthy aging.

Methods

Study population

We used data from the CLHLS, a national prospective cohort study that began in 1998 and was followed up every 2–3 years, with new participants enrolled at each follow-up [23]. Initially, 44,709 participants from 7 waves (1998, 2000, 2002, 2005, 2008, 2011, and 2014) were recruited across 23 provinces, municipalities, and autonomous regions in China, and all participants were followed up until 2018. We sequentially excluded participants who were younger than 60 years (n = 150), those with missing data on psychological resilience (n = 4,528), and those with missing data on leisure activities (n = 31). Besides, 6,321 participants who were lost in the first follow-up (n = 6,059) or had invalid death time (n = 262) were also excluded, and 33,679 older adults were included in the final analysis. The flowchart of participant recruitment criteria is shown in Supplementary Fig. 1. The CLHLS project was approved by the Biomedical Ethics Committee of Peking University (IRB00001052-13074), and all participants or their close relatives signed informed consent. Although participants were recruited from different waves of CLHLS, all variables used in the current study were assessed by the same measurement tool.

Mortality and survival data

In each subsequent follow-up visit, mortality data were obtained through death certificates. The time of death was also verified by the next-of-kin, local neighborhood committees, or community doctors to ensure accuracy [24, 25]. Participants who could not be contacted directly or via their families were defined as censored, and their survival time was the interval between the last follow-up survey and enrollment. For participants who died, survival time was defined as the interval between the times of death and enrollment.

Leisure activities

Consistent with previous CLHLS-based studies, seven LTAs related to physical or intellectual activities were included, such as doing housework, doing fieldwork, doing gardenwork, reading newspapers or books, keeping domestic animals or pets, playing cards or mah-jong (a traditional Chinese tile game played by multiple participants), and watching TV or listening to the radio [7, 26, 27]. Participants’ engagement in each activity was investigated during face-to-face interviews using a standardized questionnaire. The frequency of participation in each activity was classified as “almost every day = 3”, “sometimes = 2”, and “never = 1”. Thus, the total LTAs score was ranged from 7 to 21, with higher scores indicating more frequent engagement in leisure activities. The LTAs scale was culturally relevant for China and user-friendly for older individuals.

Psychological resilience

Levels of psychological resilience were evaluated using a standard questionnaire at baseline which comprised seven self-report items: look on the bright side of things; feel fearful or anxious; feel lonely and isolated; make own decisions; be happy as when younger; feel useless with age; and keep my belongings neat and clean. Participants rated the frequency of each item on a five-point Likert scale (always, often, sometimes, seldom, and never). The global resilience score ranged from 7 to 35, with higher scores indicating greater resilience. The scale showed good internal reliability, with a Cronbach’s alpha coefficient value of 0.73. Spearman correlation coefficients between the global psychological resilience score and each item ranged from 0.41 to 0.63 (P_S < 0.001). Details of the resilience scale are described in Supplementary Table 1.

Covariates

To mitigate potential confounding effects in this sizable sample, numerous covariates were integrated based on prior research for inclusion in the multiple models [7, 10, 15, 22, 28–30]. These covariates encompassed age, sex, ethnicity (Han or ethnic minorities), household registration (rural or urban), living arrangements (alone or with others), education level (illiterate or others), marital status (currently married or not), occupation (farmer or others), current smoking (yes or no), current alcohol consumption (yes or no), and body mass index (BMI). Additionally, activities of daily living (ADL) were assessed using the 6-item Katz Index, where a total score < 6 indicated ADL impairment [31]. Cognitive function was evaluated using the Mini-Mental State Examination (MMSE); subjects were categorized as having cognitive impairment if they were illiterate with MMSE scores < 18 or literate subjects with scores < 24 [32]. Moreover, a standard questionnaire was employed to gather information on history of chronic diseases, including hypertension, cerebrovascular disease, heart disease, respiratory disease, diabetes, and cancer. We utilized the multiple imputation by chained equations package to impute missing data for covariates using R software, and the median values from the 10 imputed datasets were used for pooled analyses [33].

Statistical analysis

Data were presented as mean ± standard deviation or number (percentage). Student’s t test, Wilcoxon rank sum test, or Chi-square test were used to compare the baseline characteristics between participants with different follow-up outcomes. Spearman correlations were used to analyze the association between LTAs and psychological resilience scores. We used analyses of variance to compare the psychological resilience levels between older adults with different frequencies of LTAs participation. Restricted cubic spline regressions were conducted to examine the linear or nonlinear relationships between LTAs, psychological resilience, and all-cause mortality risk. Adjusted Cox proportional hazard models, incorporating potential confounders, were used to calculate hazard ratio (HR) and 95% confidence intervals (CIs). We conducted a causal mediation analysis based on a counterfactual framework to explore the mediating effects of psychological resilience on the relationship between LTAs and all-cause mortality [34]. In the mediation analysis framework, the total effect (HRs of 1 unit increase in LTAs for mortality) was decomposed into pure natural direct effect and indirect effect. Proportion mediated (PM) was calculated to evaluate the effect of psychological resilience.

In addition, the following sensitivity analyses were conducted to test the stability of main results: (1) due to the inconsistent baseline times for all participants in this dynamic cohort, we further adjusted for recruitment time when analyzing the relationship between LTAs and mortality. (2) considering the potential reverse causality due to early decedents, we sequentially excluded 2772 death events in the first year; (3) since severe pathological conditions may confound the relationship, we sequentially excluded participants who had heart disease (n = 2574), cerebrovascular disease (n = 1442) or cancer (n = 436) at baseline; (4) we further adjusted for instrumental activities of daily living (IADLs) for the total sample and subgroups in the mediation analysis. All statistical analyses were performed using R software 4.2.0 version in which the mediation analysis was conducted using R package of “regmedint”, and a two-tailed P-value < 0.05 was considered significant.

Results

Sample characteristics

Demographic characteristics of 33,679 older adults are described in Table 1. The mean age of all participants was 87.21 ± 11.17 years, and 56.72% were female. Figure 1 (Fig. A) illustrates the frequency of older adults’ participation in seven leisure activities, with watching TV or listening to the radio, doing housework, and doing fieldwork ranking among the top three frequent activities. During 150,208.03 person years of follow-up, 24,705 older adults died. Participants who had lower LTAs and psychological resilience scores exhibited a higher likelihood of mortality (Ps < 0.001).

Table 1.

Demographic characteristics of 33,679 older adults across follow-up outcomes

Characteristics	Overall (n = 33,679)	Survived (n = 8,974)	Deceased (n = 24,705)	P-value
Socio-demographics
Age, years	87.21 ± 11.17	78.98 ± 10.86	89.19 ± 9.69	< 0.001
Female, n (%)	19,103(56.72)	4,841(53.94)	14,262(57.73)	< 0.001
Han ethnic, n (%)	31,382(93.18)	8,497(94.68)	22,885(92.63)	< 0.001
Rural, n (%)	25,774(76.53)	6,431(71.66)	19,343(78.30)	< 0.001
Live alone, n (%)	4,514(13.41)	1,294(14.44)	3,220(13.04)	0.001
< 1-year schooling, n (%)	21,976(65.25)	4,932(54.96)	17,044(68.99)	< 0.001
Current married, n (%)	9,573(28.44)	4,360(48.64)	5,213(21.11)	< 0.001
Farmer, n (%)	21,069(62.56)	5,259(58.60)	15,810(64.00)	< 0.001
Health status/behaviors
Body mass index, kg/m2	19.72 ± 4.03	20.99 ± 4.04	19.25 ± 3.93	< 0.001
Smoking, n (%)	6,451(19.16)	1,916(21.36)	4,535(18.36)	< 0.001
Alcohol consumption, n (%)	7,169(21.30)	1,942(21.66)	5,227(21.17)	0.326
ADL impairment, n (%)	8,198(24.34)	904(10.07)	7,294(29.52)	< 0.001
Cognitive impairment, n (%)	6,596(19.58)	794(8.85)	5,802(23.48)	< 0.001
Leisure activities score	10.98 ± 2.93	12.65 ± 2.81	10.37 ± 2.73	< 0.001
Psychological resilience score	24.51 ± 3.92	25.52 ± 3.84	24.13 ± 3.88	< 0.001
Disease history
Hypertension, n (%)	5,403(16.17)	1,800(20.24)	3,603(14.70)	< 0.001
Diabetes, n (%)	580(1.74)	227(2.55)	353(1.44)	< 0.001
Heart disease, n (%)	2,574(7.70)	847(9.52)	1,727(7.04)	< 0.001
Cerebrovascular disease, n (%)	1,442(4.31)	418(4.70)	1,024(4.18)	0.048
Respiratory disease, n (%)	3,819(11.42)	892(10.01)	2,927(11.93)	< 0.001
Cancer, n (%)	436(1.29)	134(1.49)	302(1.22)	0.056

P-values were generated using Student’s t test, Wilcoxon rank sum test, or Chi-square test

Fig. 1.

Frequency of leisure activities participation and its correlation with psychological resilience (Panel A: Frequency of leisure activities participation in the total sample; Panel B: Differences in psychological resilience scores across total LTAs score groups; Panel C: Differences in psychological resilience scores by frequency of specific leisure activities participation)

Associations of LTAs and psychological resilience

We found LTAs scores were positively correlated with psychological resilience (r = 0.332, P < 0.001). As shown in Fig. 1, older adults who participated more frequently in leisure activities have higher resilience scores (Fig. B). Besides, participants’ psychological resilience levels were statistically different between the frequency of participation in any of the leisure activities (Fig C: Ps < 0.001).

Independent associations of LTAs and psychological resilience with all-cause mortality

Figure 2 presents the dose-response relationship between LTAs, psychological resilience, and all-cause mortality risk. After adjusting for age and sex, ethnicity, residence, living status, education, marital status, occupation, BMI, smoking, alcohol consumption, ADLs, cognitive function, and history of diseases, both LTAs scores (P-nonlinear = 0.192) and psychological resilience levels (P-nonlinear = 0.308) were linearly negatively associated with all-cause mortality in the total sample, separately. Per 1 SD increase in LTAs and psychological resilience was associated with 19% (adj-HR: 0.81, 95%CI: 0.79–0.93) and 7% (adj-HR: 0.93, 95%CI: 0.91–0.95) decrease of mortality risk, respectively. In addition, we found nonlinear associations of LTAs and psychological resilience with all-cause mortality risk in the < 80 years subgroups (all P-nonlinear < 0.05). HRs of the total LTAs scores (as an ordered categorical variable) and activity-specific frequencies for all-cause mortality are summarized in the Supplementary Table 2. The risk of death was negatively correlated with the total LTAs scores, as well as the frequency of either type of activity (all P-trend < 0.001).

Fig. 2.

Dose-response relationship between leisure activities and psychological resilience with all-cause mortality risk. Adjusted for age and sex, ethnicity, residence, living status, education, marital status, occupation, BMI, smoking, alcohol consumption, ADLs, cognitive function, and history of diseases

Mediation analysis

The total LTAs were included as a continuous variable in the mediation analysis, and all HRs in the mediation model were statistically significant after adjusting for all covariates. As Fig. 3 showed, the total effect and pure natural direct effect of 1 point increase in LTAs on mortality was 0.906 (95%CI: 0.901–0.911) and 0.919 (95%CI: 0.914–0.924), respectively. The indirect effect mediated by psychological resilience was 0.986 (95%CI: 0.985–0.988). Thus, the PM for psychological resilience in the relationship between LTAs and all-cause mortality were 13.61% (95CI: 11.91–15.31%). In age and sex stratified analyses, we found the mediating effect were more pronounced in the 80 + years (PM: 17.26%, 95%CI: 15.15–19.37%) and women (PM: 14.68%, 95%CI: 12.45–16.91%) subgroups. Table 2 further demonstrates the robustness of the PM of psychological resilience even with a 5-point increase or 1 SD increase in LTAs. In addition, we also tested the mediating effect of psychological resilience for seven specific types of leisure activities (as dichotomous variables) on mortality. We found that higher PMs were observed for social activities than for productive activities. The three types of activities with the largest proportions of mediation were playing cards or mah-jong (PM: 27.75%, 95%CI: 17.36–38.14%), reading newspapers or books (PM: 23.97%, 95%CI: 18.33–29.62%), and watching TV or listening to the radio (PM: 20.16%, 95%CI: 17.10–23.23%).

Fig. 3.

Mediating roles of psychological resilience in the relationship between leisure activities and all-cause mortality. Adjusted for age, sex, ethnicity, residence, living status, education, marital status, occupation, BMI, smoking, alcohol consumption, ADLs, cognitive function, and history of diseases

Table 2.

Mediating effects of psychological resilience on the association between leisure activities and all-cause mortality

Exposure variables	Total effect	P-value	Direct effect	P-value	Indirect effect	P-value	Proportion mediated, %	P-value
Total LTAs scores
Per 1 point increase	0.906 (0.901,0.911)	< 0.001	0.919 (0.914,0.924)	< 0.001	0.986 (0.985,0.988)	< 0.001	13.61 (11.91,15.31)	< 0.001
Per 5 points increase	0.611 (0.594,0.627)	< 0.001	0.654 (0.636,0.673)	< 0.001	0.932 (0.925,0.939)	< 0.001	11.38 (10.16,12.89)	< 0.001
Per 1 SD increase	0.748 (0.736,0.761)	< 0.001	0.780 (0.767,0.793)	< 0.001	0.959 (0.955,0.964)	< 0.001	12.50 (10.89,14.11)	< 0.001
Specific types of LTAs*
Doing housework	0.685 (0.665,0.706)	< 0.001	0.721 (0.699,0.743)	< 0.001	0.950 (0.945,0.956)	< 0.001	11.37 (9.799,12.94)	< 0.001
Doing fieldwork	0.780 (0.758,0.802)	< 0.001	0.813 (0.790,0.837)	< 0.001	0.958 (0.954,0.963)	< 0.001	15.31 (12.77,17.84)	< 0.001
Doing gardenwork	0.677 (0.639,0.717)	< 0.001	0.711 (0.671,0.753)	< 0.001	0.952 (0.944,0.959)	< 0.001	10.60 (8.19,13.01)	< 0.001
Reading newspapers or books	0.770 (0.733,0.809)	< 0.001	0.825 (0.785,0.867)	< 0.001	0.933 (0.926,0.940)	< 0.001	23.97 (18.33,29.62)	< 0.001
Keeping domestic animals or pets	0.797 (0.767,0.829)	< 0.001	0.820 (0.789,0.852)	< 0.001	0.973 (0.967,0.978)	< 0.001	11.07 (8.23,13.91)	< 0.001
Playing cards or mah-jong	0.821 (0.770,0.875)	< 0.001	0.871 (0.817,0.928)	< 0.001	0.943 (0.934,0.952)	< 0.001	27.75 (17.36,38.14)	< 0.001
Watching TV or listening to the radio	0.766 (0.744,0.788)	< 0.001	0.813 (0.790,0.836)	< 0.001	0.942 (0.936,0.947)	< 0.001	20.16 (17.10,23.23)	< 0.001

Adjusted for age, sex, ethnicity, residence, living status, education, marital status, occupation, BMI, smoking, alcohol consumption, ADLs, cognitive function, and history of diseases

LTAs leisure-time activities

^* Specific types of LTAs were included as dichotomous variables

Sensitivity analyses

When we additionally adjusted for the time of recruitment in the multiple Cox models, the HRs of LTAs for all-cause mortality were materially unchanged (Supplementary Table 3). We found that exclusion of participants who were aged > 100 years or died in the first follow-up year did not appreciably attenuate the mediating effect of psychological resilience on the relationship between LTAs and mortality (Supplementary Table 4). Additionally, when we excluded participants who had heart disease, cerebrovascular disease, or cancer at baseline, the main results remained stable (Supplementary Table 4). After further adjusting for IADLs in the mediation analysis, we found that the direct effect of LTAs on mortality, as well as the proportion mediated by psychological resilience, were slightly attenuated but remained statistically significant in sex- and age-stratified subgroups (Supplementary Fig. 2).

Discussion

In this longitudinal study with 20-years follow-up, we demonstrated that both leisure activities and psychological resilience were negatively associated with the all-cause mortality in community-dwelling older adults. Moreover, we identified that the protective effect of engagement in leisure activities against mortality was mediated by psychological resilience.

We found that both the overall leisure activity scores and specific types of activities were significantly associated with the risk of mortality. The relationship remained significant even after we further adjusting for psychological resilience (Supplementary Table 2). In the total sample, every 1 SD increase in LTAs corresponded to a 9% decrease in mortality risk. A linear dose-response relationship between LTAs and all-cause mortality was evident in both male and female older adults. Our findings were consistent with previous large-scale studies. For example, a multicenter cohort study in the United States revealed that various leisure activities (running, cycling, swimming, aerobic exercise, racquet sports, golf, and walking for exercise) were linked to reduced all-cause mortality risk but exhibited no effects on cardiovascular or cancer mortality [29]. Similarly, a study in Japan indicated a linear negative correlation between global leisure activity scores and mortality risk among 48,216 older individuals aged 65 and above [10]. In addition to the general population, leisure-time physical activity has been shown to be associated with lower mortality risk among individuals across various conditions such as diabetes [11, 35], cardiovascular disease [36], and cancer [37].

The mechanisms explaining the health benefits of leisure activities can be broadly categorized into micro (individual), meso (community and group), and macro (social and cultural) levels. First, doing housework, fieldwork, and gardenwork contributes to the physical health by enhancing physical fitness, muscle strength, and bone health among older adults [28, 38]. Secondly, social activities such as playing cards or mahjong is beneficial for cognitive function and social interaction [39]. Thirdly, keeping domestic animals or pets can help older adults alleviate loneliness, particularly for those living alone or widowed [40]. Finally, leisure activities provide older adults with an opportunity to express emotions and alleviate stress. Our research has found that even solitary activities such as reading newspapers or listening to the radio/watching television were also associated with lower mortality risk. Similarly, results from the Jerusalem Longitudinal Study showed that daily reading can reduce mortality rates among community-dwelling male adults aged 70 and above [12].

In terms of specific types of LTAs, engaging in housework almost daily exhibited the highest reduction in mortality risk compared to non-participants (HR: 0.72, 95% CI: 0.69–0.74). With the introduction of China’s two-child policy, the demand for older adults to engage in household labor is rising. Activities like laundry, cooking, and caring for grandchildren benefit both individual health and family involvement, thereby fostering a positive attitude toward aging. Compared to previous studies, the seven activities included in this study align more with Chinese culture and are simpler and more feasible. Conversely, certain outdoor exercises (such as golf, swimming, and ball sports, etc.) promoted in Western countries often require specialized equipment and facilities. Therefore, encouraging older people to participate in appropriate leisure activities could be an effective way to extend life expectancy among community-dwelling older individuals. Interestingly, as supported by previous studies [7, 9], leisure activities still had a linear protective effect on reducing mortality risk even among the oldest-old adults, but this effect diminished among those under 80. These findings suggest that engaging in leisure activities is beneficial for survival even in later-life, and further validation is needed to understand the age-differences relationship.

While previous evidence has extensively explored the association between leisure activities and multiple health outcomes, our study delved into the role of resilience for the first time in relation to leisure activities and mortality. Similar to previous studies, we found an independent negative correlation between psychological resilience and mortality risk among the community population, even after adjusting for multiple covariates. Maintaining and enhancing psychological resilience is essential for healthy aging. Longitudinal studies from China have found that maintaining psychological resilience could reduce mortality risk by 15.5%, with longevous population exhibiting higher resilience levels compared to the general older population [15, 30]. In an Italian cohort study with a follow-up period of 11.2 years, resilience levels were negatively associated with overall mortality risk, although no link was found with cardiovascular disease mortality risk [18]. Our study’s strength lies in further analyzing the dose-response relationship between levels of psychological resilience and mortality risk across different age and gender groups using RCS.

The beneficial impact of psychological resilience on daily activities and cognitive function in older individuals has been well established [15, 16, 30], which helps explain the direct effect of psychological resilience in reducing the risk of mortality. The related mechanism is manifold and can be explained from both psychological and biological perspectives. Firstly, the stress process model proposed by Pearlin et al. provides a useful theoretical framework to understand the psychological pathway [41]. Resilience plays a crucial role in buffering the negative effects of stressful life events and enhancing individuals’ ability to cope effectively. Older adults with higher resilience are more likely to maintain a sense of control and optimism in the face of adversity, reducing the risk of developing functional decline and mental health problems. Secondly, basic research suggest that psychological resilience can improve autonomic nervous system regulation and contribute to activating stress responses, thereby improving immune function, and reducing the occurrence of various chronic diseases [19, 42]. The Swedish military recruitment cohort demonstrated that higher resilience levels were related to lower rates of stroke, coronary heart disease, hypertension, diabetes, and heart failure [43–45].

Through mediation analyses, we found that 13.61% of the effect of leisure activities on reducing mortality risk was mediated by psychological resilience. This finding indicated that resilience was an important factor in explaining the mechanism of leisure activities, and was supported by multiple theories. Daisy’s multi-level framework theory has made significant progress in understanding how leisure activities can contribute to higher psychological resilience levels. Participating in leisure activities offers individuals the opportunity to experience positive emotions, enhance self-esteem, foster social connections, and cultivate a sense of purpose and mastery. These factors collectively contribute to the development of psychological resilience, making them crucial for personal health. Our study further demonstrated a positive correlation between levels of psychological resilience and engagement in leisure activities, with older adults engaging daily in any leisure activity demonstrating the highest overall psychological resilience scores. These correlations serve as the basis for our mediation analysis and was consistence with previous research findings. For instance, Elena et al. found that psychological distress significantly influenced leisure activities participation among older adults in Sweden [46]. In the Health and Retirement Study, participation in community arts groups was associated with higher levels of positive affect, life satisfaction, and purpose in life, but was not associated with negative affect or personal control constraints [47]. Hei et al.‘s systematic review also demonstrated that the correlation between hobby engagement and mental health was universal across various countries regardless of socioeconomic factors [21]. However, the predominant evidence hails from Western countries, prompting our study to hold substantial public health implications, particularly for middle- and low-income developing countries. While several studies have explored the mediating role of psychological resilience, they were limited by their cross-sectional designs and sample representativeness. A cross-sectional study in Italian residents demonstrated that resilience mediated the relationship between leisure activities and subjective well-being [48]. Another CLHLS-based study confirmed the mediating role of psychological resilience between social participation and life satisfaction among community-dwelling older [17]. In the current study, we focused on long-term mortality over a 20-year follow-up period, which may provide new insights into the effect of healthy lifestyles on survival.

Subsequent sex-and age-stratified analyses unveiled a more pronounced mediating effect of resilience among the oldest individuals and women, with proportions of 17.26% and 14.68% respectively. This might be attributed to the increased prevalence of chronic diseases and functional decline as individuals age. Moreover, in line with the male-female survival paradox, despite women having longer lifespans, their overall quality of life tends to be comparatively lower than that of men [49]. We further compare the mediating role of psychological resilience on different types of LTAs. Specifically, psychological resilience exhibited a greater mediating effect on social activities compared to productive/physical activities, with the most substantial proportion observed in “playing cards or mah-jong”. Cards games and mah-jong are common group recreational activities among older adults in China, known to foster social interaction and cognitive function [50, 51]. Our finding further suggested that bolstering psychological adaptation could be another mechanism through which they positively impact health. Additionally, resilience also showed a relatively high mediation effect on “reading newspapers or books” and “watching TV or listening to the radio”. While reading, watching TV, and listening to the radio are often perceived as solitary activities, these activities provide older individuals with a means to stay informed about the world and possess certain social attributes. Particularly, during the baseline data collection period from 1998 to 2014, an era when smartphones and mobile computers were not yet widespread, newspapers, TV, and radio served as significant sources of entertainment for older adults. Future studies should delve deeper into the significance of internet use in LTAs among older populations. While our findings suggested that engagement in leisure activities conferred benefits for life expectancy through psychological resilience, a disconcerting trend emerged over the past 20 years. Both the young-old and oldest-old in China have shown a declining frequency of participation in social leisure activities [26, 52]. This worrisome trend necessitates policy attention and underscores the urgent need for public health interventions aimed at promoting healthy aging.

Limitations

This study contributed new insights into the mechanisms linking LTAs to survival. Despite the advantages of a nationally representative sample, long-term follow-up, and sufficient statistical power, some limitations should be acknowledged. First, although we considered a variety of demographic and health-related factors in the causal mediation analysis, there were still some unmeasured residual confounders that were not included. Second, we only obtained the frequency of leisure activities through self-reported questionnaires, which might introduce recall bias, warranting further exploration of activity timing and intensity in future studies. Third, the relatively advanced age of participants in the CLHLS cohort might limit generalizability to middle-aged older adults. Considering the potential confound of severe conditions, we conducted a series of sensitivity analyses. The main results remained stable even after excluding centenarians, participants who died early or suffered from cardiovascular diseases and cancer. Additionally, inclusion of over 20,000 oldest-old adults in our sample was significant for understanding healthy aging. Fourth, during the entire follow-up period, approximately 15% of the participants were lost to follow-up and could not provide accurate information on survival, which may introduce potential bias into the associations. Fifth, we only assessed leisure activities and psychological resilience at a single time point. Future research should longitudinally measure these variables to elucidate the effects of long-term lifestyle changes on survival outcomes.

Conclusions and policy implications

In conclusion, psychological resilience mediated the association between LTAs and all-cause mortality. Frequent participation in leisure activities might play a beneficial role in preventing premature death by enhancing resilience levels in community-dwelling older individuals. Further research is needed to gain a better understanding of the specific mechanisms underlying this relationship and develop targeted interventions aimed at fostering resilience through leisure activities. Our findings highlight that integrating LTAs into public health initiatives can be an effective strategy for promoting healthy aging. Psychological interventions aimed at building resilience should be considered as part of comprehensive health promotion efforts.

Abbreviations

LTAs

Leisure-time activities

CLHLS

Chinese Longitudinal Healthy Longevity Survey

BMI

Body mass index

ADL

Activities of daily living

MMSE

Mini-Mental State Examination

HR

Hazard ratio

CI

Confidence interval

PM

Proportion mediated

Authors’ contributions

Author’s contributionsCZ and SS analyzed the data and wrote the manuscript; XW, AB, and XZ contributed to critical revision of the article; ZL and CM collected and interpreted the data; JS and PZ participated in the design of the study and contributed to supervision; GF participated in the study design and contributed to critical revision of the article. All authors have read and agreed to the published version of the article.

Funding

This work was supported by and National High Level Hospital Clinical Research Funding (grant numbers: BJ-2023-072, BJ-2025-253, BJ-2025-240, BJ-2024-219) and National Key R&D Program of China (grant numbers: 2022YFF0607000).

Data availability

The raw data used in the current study can be found here: https://opendata.pku.edu.cn/dataverse/pku.

Declarations

Ethics approval and consent to participate

The study design was approved by the Ethics Committee of Beijing Hospital (2022BJYYEC-172-01). The CLHLS was approved by the Ethics Committee of Peking University (No. IRB00001052-13074). Written informed consent was obtained from all participants or their guardians (wherever applicable) prior to the study.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.