Article Text
PDF
Abstract
Background This study aimed to explore the impact of social activity frequency on mid- and long-term overall survival in older Chinese people.
Methods The association between social activity frequency and overall survival was analysed in 28 563 subjects from the Chinese Longitudinal Healthy Longevity Survey (CLHLS) cohorts.
Results A total of 21 161 (74.1%) subjects died during the follow-up of 132 558.6 person-years. Overall, more frequent social activity was associated with longer overall survival. From baseline to 5 years of follow-up, adjusted time ratios (TRs) for overall survival were 1.42 (95% CI 1.21 to 1.66, p<0.001) in the not monthly but sometimes group, 1.48 (95% CI 1.18 to 1.84, p=0.001) in the not weekly but at least once/month group, 2.10 (95% CI 1.63 to 2.69, p<0.001) in the not daily but at least once/week group, and 1.87 (95% CI 1.44 to 2.42, p<0.001) in the almost everyday group versus never group. From 5 years to the end of follow-up, adjusted TRs for overall survival were 1.05 (95% CI 0.74 to 1.50, p=0.766) in the not monthly but sometimes group, 1.64 (95% CI 1.01 to 2.65, p=0.046) in the not weekly but at least once/month group, 1.23 (95% CI 0.73 to 2.07, p=0.434) in the not daily but at least once/week group, and 3.04 (95% CI 1.69 to 5.47, p<0.001) in the almost everyday group versus the never group. Stratified and sensitivity analysis revealed similar results.
Conclusion Frequent participation in social activity was significantly associated with prolonged overall survival in older people. However, only participating in social activity almost every day could significantly prolong long-term survival.
- AGING
- MORTALITY
- PUBLIC HEALTH
- SOCIAL CAPITAL
Data availability statement
Data are available in a public, open access repository. Researchers can download the datasets free of charge from the following websites: (1) https://opendata.pku.edu.cn; Peking University Open Access Research Database; (2) https://www.icpsr.umich.edu/icpsrweb/NACDA/series/487; National Archive of Computerized Data on Aging (NACDA) sponsored by the US National Institute of Aging (NIA/NIH), Inter-university Consortium for Political and Social Research (ICPSR) at University of Michigan.
Statistics from Altmetric.com
WHAT IS ALREADY KNOWN ON THIS TOPIC
Population ageing is a global issue. The concept of active ageing has been given the priority for improving the long life and well-being for older individuals, and plenty of research has indicated that social activity can benefit their physical and mental health.
WHAT THIS STUDY ADDS
Few previous studies have investigated the dose–response relationship and determined the appropriate social activity frequency for better health outcomes in Asian older people. In addition, we do not know whether the impact of social activity frequency on health outcomes changes with time. In the present study, we found that social activity significantly prolonged the overall survival in older Chinese people—the greater the social activity frequency, the greater the likelihood of prolonged overall survival. However, we also observed a threshold effect and only participating in social activity almost every day could significantly prolong long-term overall survival.
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY
Frequent participation in social activity was significantly associated with prolonged overall survival in older Chinese people. However, in order to achieve a long-term beneficial effect of social activity, policymakers should focus on the appropriate social interventions to enhance the daily social activity participation in older people, and thus contribute to successful ageing in the era of global ageing.
Introduction
Population ageing has become a major challenge facing most countries in the world. The global population aged 60 years or over was 962 million in 2017, and the number of older people is expected to double by 2050.1 Similarly, the elderly population (aged ≥65 years) in China has also continued to increase and reached 158.31 million by the end of 2017.2 Therefore, as the world’s elderly population grows, the concept of ‘active ageing’ or ‘successful ageing’, which was defined as ‘the process of optimising opportunities for health, participation, and security in order to enhance quality of life as people age’ by the World Health Organization in 2013,3 has become an important topic, focusing on their long life and well-being. Some common risk factors, such as salt and fat intake, physical inactivity, smoking and alcohol use, have been prioritised for modification to reduce the overall risk of morbidities and mortalities; participation in social activity is another critical factor, which should be advocated to promote an active ageing society.4 5
There is a well-established literature investigating the physical and mental health benefits of participation in social activity. It has been reported that maintaining an active social involvement reduces the risk of dementia,4 coronary artery disease,5 and mental health problems,6 as well as reducing all-cause mortality risk.7 On the other hand, social isolation has been associated with a higher risk of cardiovascular diseases and all-cause mortality.8–10 However, some studies have also indicated that there is no significant association between participation in social activity and health status, especially after consideration of other conventional risk factors.11–13 The inconsistent results could be partially explained by different ethnicities, different genders, different types of social activities, various sociodemographic variables, and the lack of a universally accepted level of participation in social activity. However, the evidence about the health consequences of participation in social activity is largely derived from general populations of western countries; little is known about the association between participation in social activity and health outcomes in older people in non-western countries. Only a few studies have filled this gap. Chiao et al demonstrated that continuously participating or initiating participation in social activities reduced depressive symptoms among older Taiwanese adults.6 In another study, Yu et al revealed that social isolation was associated with increased risk of mortality in older subjects with cardiovascular disease in Taiwan.9 However, both studies had only a small sample size, and they focused on specific mental disease or sub-populations with confirmed disease. Few studies have investigated the dose–response relationship and determined the appropriate frequency of participation in social activity for better health outcomes; whether the impact of social activity frequency on health outcomes changes with time also needs further exploration.
Therefore, in this study, we aim to examine the association between social activity frequency and overall survival in a relatively large cohort of Chinese older people, and analyse the mid- and long-term impact of social activity frequency on overall survival.
Methods
Study subjects
The Chinese Longitudinal Healthy Longevity Survey (CLHLS) is an ongoing, prospective cohort study of community-dwelling Chinese older people.14 Details can be found elsewhere.15 Briefly, the CLHLS is a nationwide survey. The population in the survey areas constitutes about 85% of the whole population in China, which provides representative data to investigate determinants of longevity. It began in 1998, and examinations are carried out every 2–3 years. To reduce the attrition due to death and loss to follow-up, new participants are enrolled during the follow-up. The surveys are administered in participants’ homes by trained interviewers with a structured questionnaire. Proxy respondents, usually a spouse or other close family members, are interviewed when the participants are unable to answer questions, but questions regarding cognitive function and mood are answered by the participants themselves.
Because social activity frequency has been collected since wave 2002, the current study was based on five successive waves (2002, 2005, 2008, 2011, and 2014) within the CLHLS, with the final interview in 2018–2019. According to the inclusion and exclusion criteria, a total of 28 563 subjects were included as the study population (figure 1).
Flow chart. aAmong these subjects, 174 subjects who were all-cause mortality in the original data were redefined as lost to follow-up for lacking the years and months of deaths. Five hundred subjects lacked the days of deaths, and we reassigned the middle day, namely 15th, of a month to the date of a death. Seventy-six subjects were defined as lost to follow-up for the negative length of follow-up.
The CLHLS study was approved by the Research Ethics Committee of Peking University (IRB00001052-13074), and all participants or their proxy respondents provided written informed consent.
Assessment of social activity frequency and covariates
Social activity frequency was assessed by the question ‘Do you take part in some social activities?’ in the questionnaire, and the options for this question include: 1, almost everyday; 2, not daily, but at least once/week; 3, not weekly, but at least once/month; 4, not monthly, but sometimes; 5, never.
In the present study, we also adjusted some potential confounding variables associated with social activity frequency and mortality, including sex, age, education, marital status, residence, co-residence, household income, eating fresh fruit, eating fresh vegetables, current smoking, current drinking, current regular exercise, hypertension, diabetes, heart disease, cerebrovascular disease, respiratory disease, cancer, and self-rated health. For these variables, online supplemental table S1 shows the detailed information about the reclassifications used in the present study.
Supplemental material
Study outcome
The study outcome was overall survival, defined as the time from baseline to any cause of death. Survival status and date of death were collected through interviews with close family members during each survey. All subjects were followed from the first interview up to the outcome or the most recent interview. A ‘lost to follow-up’ status was assigned to those who could not be contacted after baseline interview.
Statistical analysis
The missing values of baseline variables were mostly less than 0.70%. Due to such low missing rates, we deleted the cases with missing values in the statistical analyses, without imputing the missing values; online supplemental table S2 shows the distributions of baseline variables with missing data.
Supplemental material
Baseline characteristics of the study population were described as median with interquartile range (IQR) for continuous variables or percentages for categorical variables according to the social activity frequency. Survival proportions were estimated using the Kaplan-Meier method, and the log-rank test was used for comparison. Multivariable parametric accelerated failure time (AFT) models were used to evaluate the association of social activity frequency with survival, because Schoenfeld Residuals indicated a violation of proportional hazards assumption in the Cox proportional hazards models. The logistic distribution was selected for AFT models based on the minimum Akaike Information Criterion among different survival distributions (ie, Weibull, exponential, loglogistic, and gaussian). The AFT model estimates the time ratio (TR), which is interpreted as the expected time to events in one category relative to the reference group. Unlike the interpretation of proportional hazard model results where hazard ratios larger than 1 are equal to higher risk, a TR of greater than 1 is considered to have a longer time to events compared with the reference group. Additionally, we did landmark analyses to assess outcomes at 5 years and between 5 years and the end of follow-up. The hypothesised causal diagram is shown in online supplemental figure S1.
Supplemental material
Furthermore, stratified analysis assessed the consistency of association between social activity frequency and outcome in various subgroups, and interactions were examined by likelihood ratio testing. To assess robustness of the results, we performed two sensitivity analyses: (1) to address the issue of loss to follow-up, we conducted a sensitivity analysis by considering the losses censored at the end of the study; and (2) to exclude the deaths that occurred within the first year of follow-up, we also conducted a sensitivity analysis to reduce potential reverse causation.
The statistical analyses were performed with the use of R software, version 4.1.0 (R Project for Statistical Computing). For all statistical analyses, a two-sided p value of 0.050 was considered statistically significant.
Results
Baseline characteristics
Baseline characteristics of the 28 563 study subjects (median age 89.00 years, IQR 80.00 to 98.00 years, 11 855 males) by social activity frequency are shown in table 1 and online supplemental table S3). Among them, 25 406, 1379, 693, 553, and 532 subjects were, respectively, in the never group, the not monthly but sometimes group, the not weekly but at least once/month group, the not daily but at least once/week group, and the almost everyday group. Results revealed statistically significant differences in age, sex, education, marital status, residence, co-residence, household income, fresh fruit and fresh vegetables intake, lifestyle factors, and self-rated health, as well as several morbidities, including hypertension, diabetes, heart disease, and cancer across the five groups. Generally speaking, male subjects were more likely to attend social activity than female subjects. Subjects who were younger and received a longer period of education were more likely to engage in social activity as well. Subjects who stayed in marriage and subjects who lived in an urban community or co-residence with family members were more socially active than their counterparts. In addition, subjects with no morbidities or with good self-rated health were more likely to be involved in social activity.
Supplemental material
Baseline characteristics
Association between social activity frequency and overall survival from baseline to the end of follow-up
Overall, 21 161 (74.1%) deaths occurred during the follow-up of 132 558.6 person-years. The all-cause mortality rate gradually decreased from never to almost everyday groups, and the rates in each group were 17.2 (95% CI 17 to 17.4), 9.3 (95% CI 8.7 to 9.9), 9.1 (95% CI 8.2 to 9.9), 9.0 (95% CI 8 to 9.9), and 7.4 (95% CI 6.5 to 8.3) per 100 person-years, respectively (table 2). Kaplan-Meier analysis also demonstrated that the survival probability was significantly higher in the groups with more frequent social activity (log-rank p<0.001) (figure 2A). In the AFT analysis, when comparing to the never group, the unadjusted TRs were 18.10 (95% CI 13.53 to 24.22, p<0.001) in the not monthly but sometimes group, 22.04 (95% CI 14.65 to 33.17, p<0.001) in the not weekly but at least once/month group, 28.60 (95% CI 18.24 to 44.87, p<0.001) in the not daily but at least once/week group, and 54.84 (95% CI 33.41 to 90.04, p<0.001) in the almost everyday group (table 2). After adjusting for potential confounding factors, results showed that the overall survival time of subjects with more frequent social activity was longer compared to subjects who never participated in social activity. Adjusted TRs were 1.71 (95% CI 1.35 to 2.15, p<0.001) in the not monthly but sometimes group, 1.85 (95% CI 1.32 to 2.59, p<0.001) in the not weekly but at least once/month group, 2.58 (95% CI 1.79 to 3.71, p<0.001) in the not daily but at least once/week group, and 3.48 (95% CI 2.36 to 5.12, p<0.001) in the almost everyday group (figure 3). The associations between other variables and survival are shown in online supplemental table S4. Stratified analyses indicated that the survival benefit of social activity was more prominent in female subjects (p for interaction=0.022), in the oldest-old subjects over 80 (p for interaction=0.041), and in subjects who lived with family members (p for interaction=0.003); no significant interaction was detected in other subgroups (online supplemental figure S2).
Supplemental material
Supplemental material
Kaplan-Meier survival curves of baseline social activity frequency for overall survival. (A) Survival probability in different groups. (B) Landmark analysis discriminating between survival probability before and after 5 years of follow-up.
Adjusted associations of social activity frequency with overall survival.aPer 100 person-years. bAdjusted for sex, age, education, marital status, residence, co-residence, household income, fresh fruit, fresh vegetables, current smoking, current drinking, current regular exercise, hypertension, diabetes, heart disease, cerebrovascular disease, respiratory disease, cancer, and self-rated health. TR, time ratio.
Unadjusted associations of social activity frequency with all-cause mortality, and adjusted sensitivity analyses
Association between social activity frequency and overall survival from baseline to 5 years of follow-up
From baseline to 5 years of follow-up, the results remained similar with the above findings. The all-cause mortality rate gradually decreased from 18.4 (95% CI 18.2 to 18.7) per 100 person-years in the never group to 7.3 (95% CI 6.2 to 8.5) per 100 person-years in the almost everyday group (table 2). Landmark analysis also showed that subjects with more frequent social activity tended to have significantly higher survival probability (log-rank p<0.001) (figure 2B). Comparing to the never group, the overall survival time was significantly longer in the not monthly but sometimes group (TR 1.42, 95% CI 1.21 to 1.66, p<0.001), in the not weekly but at least once/month group (TR 1.48, 95% CI 1.18 to 1.84, p=0.001), in the not daily but at least once/week group (TR 2.10, 95% CI 1.63 to 2.69, p<0.001), and in the almost everyday group (TR 1.87, 95% CI 1.44 to 2.42, p<0.001) (figure 3). The associations between other variables and survival are shown in online supplemental table S4. Stratified analyses also revealed that sex (p for interaction=0.005), age (p for interaction <0.001), marital status (p for interaction=0.046), co-residence (p for interaction=0.047), and self-rated health (p for interaction=0.032) had interactive effects, and there was no significant interaction between other variables and social activity for survival (online supplemental figure S3).
Supplemental material
Association between social activity frequency and overall survival from 5 years to the end of follow-up
From 5 years to the end of follow-up, although landmark and unadjusted AFT analyses indicated similar results to the previous findings (table 2 and figure 2B), the results of adjusted AFT analysis changed materially. Adjusted TRs were 1.05 (95% CI 0.74 to 1.50, p=0.766) in the not monthly but sometimes group, 1.64 (95% CI 1.01 to 2.65, p=0.046) in the not weekly but at least once/month group, 1.23 (95% CI 0.73 to 2.07, p=0.434) in the not daily but at least once/week group, and 3.04 (95% CI 1.69 to 5.47, p<0.001) in the almost everyday group versus never group (figure 3). That is to say, a threshold effect was observed, and only subjects participating in social activity almost everyday could have significantly longer overall survival time after 5 years of follow-up. The association between other variables and survival are shown in online supplemental table S4). No significant interaction was found in all subgroups (online supplemental figure S4).
Supplemental material
Sensitivity analysis
After considering the losses censored at the end of the study or excluding deaths within the first year, the results remained unchanged (table 2). From baseline to 5 years of follow-up, more frequent social activity could exert more survival benefits. After 5 years of follow-up, there was a threshold effect regarding the survival benefits of social activity, and only those who participated in social activity almost every day could have significantly longer overall survival time.
Discussion
This study found that frequent participation in social activity was associated with prolonged overall survival time. From baseline to 5 years of follow-up, the more frequent the social activity, the more prolonged the survival time. However, after 5 years of follow-up, there was a threshold effect regarding the association between social activity frequency and overall survival time, and only participating in social activity almost every day could significantly extend the overall survival time.
Previous studies have investigated the frequency of social participation in health maintenance. A population-based study in Chile, involving individuals aged >60 years, observed that subjects’ participation in social activities had a 22% lower risk of death than those who did not participate during the 5-year study period.16 The Nord-Trøndelag Health Study (HUNT), a longitudinal study with a mean follow-up of 8.15 years in Norway, demonstrated that the frequency of social participation of 0.5 to less than 1, 1 to less than 2, and 2 or more times per week significantly reduced the mortality risk by 18%, 31%, and 39%, respectively.17 Shimatani et al also explored the association between the change of the frequency in social participation and all-cause mortality for individuals aged ≥60 years in Japan; they found that continued or decreased frequency of social participation was associated with a decreased risk of all-cause mortality. Initiation of social participation after the age of 60 years failed to reduce the mortality risk.18 In the present study, we mainly focused on the impact of the social activity frequency on mid- and long-term overall survival, and confirmed that social participation was a strong protective factor of health and longevity for older people. In the stratified analysis, for the oldest-old subjects, social activity showed an even more profoundly protective effect on extending overall survival time within the first 5 years. However, a previous study reported that the social participation rate decreases significantly over time in subjects aged over 85. Thus, interventions that contribute to the maintenance of participation in social activity in very old subjects should be encouraged.
Mechanisms behind the association between social participation and health outcomes are not completely understood, but there are some possible explanations. In one aspect, social participation could affect individuals’ health behaviour. Previous studies have indicated that baseline daily smokers who had remained as daily smokers had higher rates of non-participation,19 while high social participation contributed to the maintenance of smoking cessation.20 Additionally, it has been reported that social participation among Japanese older people was associated with more physical activity and less sedentary time,21 which had potential benefits for some types of chronic diseases. Social participation may also encourage healthier dietary behaviours, such as increased intake of fresh fruit and vegetables,22 and sufficient fruit and vegetable consumption was significantly correlated with better quality of life in older people.23 Therefore, subjects involved in social participation are exposed to peer and social influences, which may ultimately have an impact on their normative views and subsequent behaviours, such as dropping risky health behaviours (eg, smoking) and engaging in beneficial health behaviours (eg, physical activity, healthy dietary habitats), and these healthy behaviours may partially mediate the association between social participation and health outcomes. In another aspect, social participation may also buffer the deleterious influence of acute or chronic stressors on health, contributing to the subjects’ psychological well-being.24 In our study, although the association between social activity frequency and overall survival attenuated after adjusting for sociodemographic factors, socioeconomic status, healthy behaviours and several morbidities, it still remained statistically significant, which indicated that social activity participation per se was an independent predictor for overall survival in older people.
One of the interesting findings is that we observed a threshold effect regarding the association between social activity frequency and long-term overall survival. In order to prolong long-term overall survival, daily social activity participation is suggested for older people. Whereas we cannot confirm the causality, a word of caution is that we only considered the baseline social activity frequency. The long-term association could be much more complicated due to the possible change of social activity frequency, as well as other health-related variables such as socioeconomic status, lifestyle behaviour, comorbidities, psychological distress and so on during the follow-up time. Since no other historical studies reported similar results, whether this is a chance finding or not needs further exploration.
Some limitations need to be acknowledged. First, the present study used information from the baseline survey; we did not account for the changes in social participation over the follow-up period. Other health related variables could also change over time, which can bias our estimates. Second, although we have included a wide range of control variables for adjustment, there might be other factors linking social participation to mortality risk. Third, we only focused on the objective social activity frequency. The association between different types of social participation which the respondents participated in and the likelihood of survival were not analysed. However, a previous study based on the five waves of CLHLS have reported that both secular social participation and religious participation contributed to deceased mortality risk.25 Fourth, the present study only included Chinese people, and over 90% of them were Han Chinese. We did not differentiate between Han Chinese and minorities in our study. Extending the present findings to other ethnicities should be approached with caution. Fifth, several variables were collected by questionnaire and based on subjects’ self-report. Thus, recall bias may exist. However, it would be hard to address this issue considering the nature of the epidemiological study. These limitations suggest the need for more in-depth analysis regarding the impact of social activity frequency on mid- and long-term mortality risk for older people.
Conclusion
Frequent participation in social activity was significantly associated with prolonged overall survival in older Chinese people. However, in order to achieve long-term survival benefits of social activity, daily participation in social activity should be urged.
Data availability statement
Data are available in a public, open access repository. Researchers can download the datasets free of charge from the following websites: (1) https://opendata.pku.edu.cn; Peking University Open Access Research Database; (2) https://www.icpsr.umich.edu/icpsrweb/NACDA/series/487; National Archive of Computerized Data on Aging (NACDA) sponsored by the US National Institute of Aging (NIA/NIH), Inter-university Consortium for Political and Social Research (ICPSR) at University of Michigan.
Ethics statements
Patient consent for publication
Ethics approval
This study involves human participants and was approved by the Research Ethics Committee of Peking University (IRB00001052-13074). All participants or their proxy respondents provided written informed consent to participate in the study before taking part.
Acknowledgments
We are grateful to the CLHLS study, which provided the data in this research. The CLHLS was supported by funds from the US National Institute on Aging (NIA), the China Natural Science Foundation, the China Social Science Foundation and the United Nations Fund for Population Activities (UNFPA) and was managed by the Center for Healthy Aging and Development Studies, Peking University.
References
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Footnotes
ZW and YZ contributed equally.
Contributors ZW and YZ designed the concept, analysed the data, interpreted the data and prepared the manuscript. HR designed the concept, interpreted the outcome and reviewed the manuscript. LL designed the concept and reviewed the manuscript. LD designed the concept and reviewed the manuscript. SH designed the concept, analysed the data, interpreted the data, and reviewed the manuscript. All authors have read and approved the manuscript, and ensure that this is the case. SHnis responsible for the overall content as guarantor.
Funding This study was supported by Sichuan Science and Technology Program, China (grant Number: 2022YFS0186), the National Key R&D Program of China (grant number: 2017YFC0910004), and the National Natural Science Foundation of China (grant number: 81600299).
Competing interests None declared.
Provenance and peer review Not commissioned; externally peer reviewed.
© Author(s) (or their employer(s)) 2023. No commercial re-use. See rights and permissions. Published by BMJ.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.