Associations between cognitive activities and all-cause mortality among older adults with cognitive impairment: A prospective cohort study

Linjia Duan; Liming Zhao; Ziqiong Wang; Lu Liu; Ningying Song; Sen He

doi:10.1371/journal.pone.0319093

Abstract

Background

Evidence on the association between cognitive activities and mortality among older adults with cognitive impairment is limited. Therefore, the study aimed to assess the association and examine whether baseline cognitive function mediates the association.

Methods

A total of 10477 older participants with cognitive impairment (median age: 95.0 [IQR: 88.0–100.0], males: 27.9%, Mini-Mental State Examination score ≤24 points) from the Chinese Longitudinal Healthy Longevity Survey conducted between 1998 and 2014 were included, with follow-up until 2018. Exposures were three prevalent cognitive activities among older adults in China: reading books/newspapers, playing cards/mah-jong, and watching TV or listening to radio, and the outcome was all-cause mortality within a 10-year follow-up period. We evaluated the association between these activities and mortality using Cox regression models and also conducted a mediation analysis to examine the role of baseline cognitive function in this association.

Results

During a follow-up period of totaling 33632.1 person-years, there were 8763 recorded deaths (83.6%). For each cognitive activity, the risk of mortality decreased with an increased frequency of engagement in these activities. Moreover, the risk of mortality significantly decreased with a greater number of cognitive activities. With zero activities as reference, adjusted hazard ratios were 0.83 (95% CI: 0.79–0.87) for one activity, 0.76 (95% CI: 0.69–0.83) for two activities, and 0.67 (95% CI: 0.53–0.86) for three activities, respectively. Stratified and sensitivity analyses confirmed the robustness of these findings. Additionally, baseline cognitive function partially mediated the association between cognitive activities and mortality; compared to zero activities, the mediated proportions were 15.2% (95% CI: 10.9%–22.4%) for one activity, 13.4% (95% CI: 8.9%–21.3%) for two activities, and 9.3% (95% CI: 4.2%–23.4%) for three activities, respectively.

Conclusions

Among older adults with cognitive impairment in China, the risk of all-cause mortality significantly decreased as both the frequency and number of cognitive activities increased. Baseline cognitive function only mediated a small proportion of the benefits of cognitive activities in longevity.

Citation: Duan L, Zhao L, Wang Z, Liu L, Song N, He S (2025) Associations between cognitive activities and all-cause mortality among older adults with cognitive impairment: A prospective cohort study. PLoS ONE 20(2): e0319093. https://doi.org/10.1371/journal.pone.0319093

Editor: Kathleen Bennett, Royal College of Surgeons in Ireland, IRELAND

Received: October 3, 2024; Accepted: January 27, 2025; Published: February 20, 2025

Copyright: © 2025 Duan et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: 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 U.S. National Institute of Aging (NIA/NIH), Inter-university Consortium for Political and Social Research (ICPSR) at University of Michigan.

Funding: This study was supported by the Sichuan Science and Technology Program, China (Grant No. 2022YFS0186), the National Natural Science Foundation of China (Grant No. 81600299), the Science and Technology Major Project of Tibetan Autonomous Region of China (Grant No. XZ202201ZD0001G01), and the Key Research and Development Projects of Chengdu Science and Technology Bureau (Grant No. 2022-YF05-01335-SN). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Competing interests: The authors have declared no conflict of interest.

Introduction

Population aging represents one of the significant challenges confronting many countries worldwide, and cognitive impairment is notably prevalent among older adults, with its incidence increasing as age advances [1–3]. Moreover, numerous studies have shown that cognitive impairment is associated with an increased risk of death [4–6]. Inevitably, the increasing mortality rate among older adults with cognitive impairment will emerge as one of the major threats to future public health resources. On the other hand, cognitive activities have been a hot topic in the research field of older adults, and there is a large and growing body of evidence on the health benefits of engagement in cognitive activities [7–16]. Especially, emerging studies have presented that most cognitive activities potentially lead to a decreased risk of mortality [11,15,17–20].

Currently, important gaps between cognitive activities and mortality remain. First, previous studies about the association between cognitive activities and mortality mainly focused on the overall older adult population [11,15,17–20], without distinguishing those with abnormal cognitive levels from the general cohort. Second, cognitive function is commonly viewed as a mediator between cognitive activities and mortality [15,20]; furthermore, given the increased mortality risk with cognitive impairment [4–6], cognitive impairment might alleviate the benefits of cognitive activities in longevity. Therefore, evaluating the contribution of cognitive function is crucial for understanding the association between cognitive activities and mortality, particularly among older adults with cognitive impairment.

To address these knowledge gaps, the present study aimed to investigate the association between cognitive activities and all-cause mortality among older adults with cognitive impairment in China, using data from the Chinese Longitudinal Healthy Longevity Survey (CLHLS). In this study, we also investigated whether baseline cognitive function served as a mediator in the association between cognitive activities and all-cause mortality.

Methods

Study participants

The CLHLS is a nationwide, ongoing prospective cohort study focusing on community-dwelling older adults in China, and its general goal is to enhance our understanding of the determinants of healthy aging. Detailed information has been reported elsewhere [11,21]. Briefly, the baseline survey of the CLHLS commenced in 1998, with follow-up surveys conducted every two to three years. Data from eight surveys have been made available, specifically for the years 1998, 2000, 2002, 2005, 2008, 2011, 2014, and 2018. In the initial year of the study (1998), a multistage stratified cluster sampling method was employed to conduct the baseline survey across a randomly selected half of the counties and cities within 22 out of China’s 31 provinces, and the population residing in these surveyed areas represented approximately 85.0% of the total Chinese population. To mitigate attrition due to mortality and loss to follow-up, new participants have been enrolled since wave 2000, with finally covering 23 of 31 provinces in China. The surveys were conducted in the homes of participants by trained interviewers utilizing a structured questionnaire. In instances where participants were unable to respond, proxy respondents—typically a spouse or other close family members—were interviewed; however, questions pertaining to cognitive function and mood were answered directly by the participants themselves. The study was carried out in accordance with the principles outlined in the Declaration of Helsinki and received approval from the Research Ethics Committee of Peking University (IRB00001052–13074). All participants, or their proxy respondents, provided written informed consent.

The present study was based on seven waves (i.e., 1998, 2000, 2002, 2005, 2008, 2011, and 2014) within the CLHLS, and the final wave of interview was 2018. In the CLHLS study, cognitive function was assessed using the Chinese version of the Mini-Mental State Examination (MMSE) [6], which evaluates various domains of cognitive functioning, including calculation, language, orientation, and recall. The Chinese version of the MMSE takes into account the cultural and socioeconomic contexts specific to China, ensuring that all items in the test are understood and answered by participants with normal cognitive abilities. The validity and reliability of the Chinese version of the MMSE have been established by previous studies [21,22]. In the study, we classified responses of “unable to answer” as “incorrect” in the MMSE based on the literatures [23,24], and the total MMSE score ranged from 0 to 30 points, with higher scores indicating better cognitive function. Cognitive impairment was defined as a total MMSE score of ≤ 24 points [6,12]. A flow chart illustrating the participant enrollment process for this study is presented in Fig 1, and the final sample comprised 10477 older participants (aged ≥ 65 years) with cognitive impairment.

Download:

Fig 1. Flow chart of participants selection.

MMSE=mini-mental state examination.

https://doi.org/10.1371/journal.pone.0319093.g001

Assessment of cognitive activities

Participants were interviewed about their engagement in three most popular cognitive activities among older adults in China, and these activities included reading books/newspapers, playing cards/mah-jong and watching TV or listening to radio. Participants reported the frequency of their participation using categories such as “never,” “sometimes,” or “almost everyday” for the waves conducted in 1998 and 2000. For subsequent waves in 2002, 2005, 2008, 2011, and 2014, participants used a more detailed scale: “never,” “not monthly but sometimes,” “not weekly but at least once a month,” “not daily but once a week,” or “almost everyday.” We recoded these responses to create a uniform scale across the seven waves: (1) never: never; (2) sometimes: sometimes, not monthly/but sometimes, not weekly/but at least once for a month, not daily/but once for a week; (3) almost everyday: almost everyday.

To explore the effect of cumulative participation in cognitive activities on mortality, we then constructed a dichotomous variable for participation in cognitive activities by assigning a value of 0 for the scale (i.e., never) and 1 for the scales (i.e., sometimes and almost everyday), and we further summed the three dichotomised variables to create a cumulative measure, resulting in a cumulative cognitive activity variable with values ranging from zero to three.

Study outcome

The outcome of the study was all-cause mortality within a 10-year follow-up period. Information regarding survival status and date of death was collected through interviews with close family members during each survey and corroborated using death certificates, hospital admission records, and medical records when available. All participants were followed from the initial interview until either the occurrence of the outcome or their most recent interview.

Assessment of covariates

In the study, we also investigated a comprehensive range of factors that have been associated with cognitive activities and mortality. These factors include: sex, age, education level, marital status, place of residence, co-residence, regular consumption of fruits, regular consumption of vegetables, regular intake of meats, current smoking habits, patterns of alcohol consumption, exercise routines, and history of comorbidities (hypertension, diabetes, heart diseases, cerebrovascular diseases, respiratory diseases, and cancer), as well as self-rated health assessments. Detailed information regarding the scales utilized for these factors in the present study can be found in S1 Table.

Statistical analysis

S2 Table presents the distributions of baseline covariates with missing data. In the primary analysis, cases with missing data were excluded under the assumption of missing completely at random. Overall, the analyses consisted of four steps: (1) comparing baseline characteristics; (2) determinating the association between cognitive activities and all-cause mortality; (3) conducting stratified and sensitivity analyses; (4) assessing whether baseline cognitive function served as a mediator between cognitive activities and mortality.

Firstly, baseline characteristics were presented as median (interquartile range, IQR) for continuous variables and as number (percentage) for categorical variables. For continuous variables, the p-value for trend was calculated using the Pearson test when the row variable followed a normal distribution, and the Spearman test when it exhibited a non-normal distribution. In cases where the row variable was categorical, the p-value for trend was derived from the Mantel–Haenszel test of trend.

Secondly, the Kaplan-Meier method was used to estimate the cumulative incidence of all-cause mortality across various subgroups, with the log-rank test used for comparisons. Then, Cox proportional hazards models were used to estimate the association between cognitive activities and mortality, with follow-up time as the time scale, and we found no evidence indicating a potential violation of the proportional-hazards assumption regarding the exposure-outcome association (S1 Fig). We estimated the association between cognitive activities and mortality using two models: (1) model 1 included sex and age; and (2) model 2 incorporated sex, age, education, marital status, residence, co-residence, regular intake of fruits, regular intake of vegetables, regular intake of meats, current smoking, current drinking, current regular exercise, hypertension, diabetes, heart diseases, cerebrovascular diseases, respiratory diseases, cancer, and self-rated health.

Thirdly, to test the robustness of our primary findings, we repeated all analyses in various subgroups, along with several sensitivity analyses: (1) to address loss to follow-up, we conducted a sensitivity analysis by censoring losses at the end of follow-up time; (2) to exclude deaths that occurred within the first year or two years of follow-up, thereby reducing potential reverse causation; (3) to account for variations due to different waves, we adjusted for baseline interview timing; (4) to mitigate possible bias resulting from missing data, we implemented multiple imputation techniques in our analysis; and (5) to examine the possibility of unmeasured confounding between cognitive activities and mortality, we calculated E-values [25].

Finally, we performed a regression-based causal mediation analysis within the direct counterfactual framework to evaluate how baseline cognitive function, as measured by MMSE, mediate the association between cognitive activities and mortality, using the CMAverse package [26–28]. In brief, two regression models were fitted: one for the outcome (i.e., all-cause mortality; using a Cox proportional-hazards model) and another for the mediator (i.e., baseline cognitive function; using linear regression). In the two models, cognitive activities were treated as categorical variables, while the baseline cognitive function served as a linear term. Standard errors were estimated via bootstrapping with 1000 samples. All models were adjusted for potential confounders.

All analyses were conducted using R version 4.2.2 (available at http://www.R-project.org) with the implementation of essential packages, including “compareGroups”, “survival”, “mice”, “CMAverse”, and “stats”. All statistical tests employed were two-sided, and p-values less than 0.05 were considered statistically significant. In addition, the official organization (i.e., Center for Healthy Aging and Development, Peking University) of the CLHLS suggests that weights may not be necessary if researchers are only describing the sample’s status without comparing it to other populations. These suggestions are available on their website at http://chads.nsd.pku.edu.cn/yjxm/zglnjkyxysgztc/index.htm. Therefore, we did not incorporate the complex sampling design into our analyses.

Results

Baseline characteristics

The study comprised a total of 10477 older participants with cognitive impairment (median age: 95.0 [IQR: 88.0–100.0]; males: 27.9%). Table 1 presents the baseline characteristics of participants categorized by the number of cognitive activities. Participants engaging in a greater number of cognitive activities were more likely to be male, younger, better educated, married, and residing in urban areas. Additionally, there was a higher prevalence of comorbidities among those participating in more cognitive activities, and their MMSE scores were also elevated compared to those with fewer cognitive activities. Further detailed information regarding baseline characteristics is provided in Table 1. Moreover, baseline characteristics stratified by the frequency of cognitive activities are presented in S3–S5 Tables.

Download:

Table 1. Baseline characteristics by number of cognitive activities.

https://doi.org/10.1371/journal.pone.0319093.t001

Associations between cognitive activities and all-cause mortality

During a follow-up period of 33632.1 person-years, a total of 8763 deaths (83.6%) were recorded. For the three individual cognitive activities, Kaplan-Meier curves indicated that a higher frequency of cognitive activities was associated with a more favorable clinical outcome (all log-rank p < 0.001, S2 Fig). Both univariate and multivariate analyses further demonstrated that an increased frequency of cognitive activities was associated with lower mortality risk (S6 Table).

Furthermore, we assessed the association between the number of cognitive activities and all-cause mortality. As the number of cognitive activities increased, there was a significant decrease in mortality rates (log-rank p < 0.001, Fig 2). Accordingly, when comparing participants without any cognitive activity (zero activities), crude mortality risks were lower for those engaging in one activity (hazard ratios [HR] = 0.70, 95% CI: 0.67–0.73), two activities (HR = 0.64, 95% CI: 0.59–0.70), and three activities (HR = 0.50, 95% CI: 0.39–0.63), respectively (Table 2). When using zero activities as the reference, adjusted HRs were 0.83 for one activity (95% CI: 0.79–0.87), 0.76 for two activities (95% CI: 0.69–0.83), and 0.67 for three activities (95% CI: 0.53–0.86), respectively (Table 2).

Download:

Fig 2. Kaplan-Meier analysis showing cumulative incidence of all-cause mortality by number of cognitive activities.

https://doi.org/10.1371/journal.pone.0319093.g002

Download:

Table 2. Associations of number of cognitive activities with all-cause mortality and medication analysis of number of cognitive activities by baseline cognitive function.

https://doi.org/10.1371/journal.pone.0319093.t002

Stratified and sensitivity analyses

We further performed stratified analyses to assess the association between the number of cognitive activities (zero activities vs. one activity vs. two activities vs. three activities) and the risk of all-cause mortality across various subgroups. Overall, we observed a decrease in mortality risk with an increasing number of cognitive activities in all subgroups (Fig 3, S3 Fig). None of the variables—including sex, age, education, marital status, residence, MMSE score, current smoking, current drinking, hypertension, and cerebrovascular diseases—significantly modified the association between the number of cognitive activities and mortality. Moreover, the sensitivity analyses did not find substantial changes in the results (Table 2).

Download:

Fig 3. Stratified analyses by potential modifiers of the association between the number of cognitive activities and all-cause mortality.

https://doi.org/10.1371/journal.pone.0319093.g003

Each stratification adjusted for all factors (sex, age, education, marital status, residence, co-residence, regular intake of fruits, regular intake of vegetables, regular intake of meats, current smoking, current drinking, current regular exercise, hypertension, diabetes, heart diseases, cerebrovascular diseases, respiratory diseases, cancer, and self-rated health) except the stratification factor itself. The grouping criteria for continuous variables were determined based on the median values. Due to the extensive size of the figure, certain information is included in S3 Fig.

Mediation analysis

Table 2 also presents the results of mediation analysis that assessed whether baseline cognitive function served as a mediator between the number of cognitive activities and all-cause mortality. The results indicated that both the natural direct effect and the natural indirect effect were statistically significant, suggesting a partial mediation effect of baseline cognitive function on the association between the number of cognitive activities and mortality. Specifically, when comparing with zero activities, the proportion mediated by baseline cognitive function was found to be 15.2% (95% CI: 10.9%–22.4%) for one activity, 13.4% (95% CI: 8.9%–21.3%) for two activities, and 9.3% (95% CI: 4.2%–23.4%) for three activities, respectively. Furthermore, we also evaluated whether baseline cognitive function acted as a mediator between individual cognitive activities and mortality; these results are detailed in S7 Table.

Additionally, another mediation analysis was performed to examine the directionality of effects by using baseline cognitive function as the predictor variable and cognitive activities as the mediator variable; however, all natural indirect effects were found to be non-significant (S8 Table).

Discussion

In analyses based on a large cohort of older adults with cognitive impairment, we found that frequent participation in cognitive activities—such as reading books/newspapers, playing cards/mah-jong, and watching TV or listening to radio—was associated with a lower risk of all-cause mortality. Furthermore, the risk of mortality also significantly decreased with an increased number of cognitive activities. Moreover, baseline cognitive function only mediated a small proportion of the benefits of cognitive activities in longevity.

Engagement in cognitive activities is associated with reduced mortality risk. Emerging evidence indicates reading books/newspapers may have a beneficial impact on survival [11,20,29]. Similarly, our findings are consonant with these studies, and the underlying mechanisms might be partially attributed to the inverse association of reading books/newspapers with cognitive impairment [20,30,31]. In contrast, some previous found reading had no effect on mortality [15,32], and the inconsistency might be the caused by the differences of study populations. In accordance with our results, some previous studies also showed playing cards/mah-jong reduced risk of all-cause mortality among older adults [11,15,18]. Playing cards/mah-jong is inherently social in nature, which has be associated with a reduced mortality risk [33]. For watching TV or listening to radio, some previous studies have suggested that watching TV for a long time might be associated with increased mortality risk [34,35]; on the other hand, a study reported positive effect of more participation in watching TV and listening to radio on decreased mortality risk among oldest-old adults [11]. Combined with these studies [11,34,35], the present findings might suggest increased frequency of watching TV may be associated with a decreased mortality risk, but excessively prolonged viewing time may increase mortality risk. Meanwhile, most of present study participants were illiterate and thus lacked many sources of information; therefore, watching TV or listening to the radio may maintain cognitive function, further reducing mortality risk [17]. Additionally, most studies have shown cognitive activities are also associated with other factors related to better health status. For example, cognitive activities are associated with a lower risk of activities of daily living disability [10], and those who participate more in cognitive activities may have a better consciousness of diseases and they may access the medical care system more actively [36]. Despite the diverse characteristics of the samples in these studies, a consistent positive association has been observed between cognitive activities and health. Furthermore, the present study indicated the risk of all-cause mortality progressively decreased with increased number of cognitive activities. Similarly, some previous studies [15,37], also displayed older adults were less possibly to die among those engaged in an increased number of activities than those who engaged in few activities. Although the p-values for interactions concerning age and education were less than 0.050 in the study, these findings may lack significant clinical implications due to multiple testing considerations and similar directional trends in associations. Thus, both previous studies and the current investigation suggest that frequent engagement in cognitive activities, as well as participation in a greater variety of such activities, may reduce the risk of mortality. Our findings further extend these results to older adults with cognitive impairment.

The mediation analysis showed that only up to 27.9% of the association between cognitive activities and all-cause mortality was explained by baseline cognitive function, indicating that cognitive activities reducing mortality risk only depended on baseline cognitive function partly; in other words, even if the older adults had cognitive impairment, they may still get significant survival advantage by cognitive activities. Although the present findings showed baseline cognitive function only mediated a small proportion of the benefits of cognitive activities in mortality, another study [20], using the data from the Health and Retirement Study, suggested subsequent cognition was a complete mediator of the baseline book reading survival advantage after adjustment for baseline cognition in a total of 3635 older participants. Whether subsequent cognition was a complete mediator between cognitive activities and all-cause mortality in the older adults with cognitive impairment, further studies are needed and warranted. We also found that cognitive activities did not mediate the association between baseline cognitive function and survival, indicating that reverse causality mediation was not present, and that is higher baseline cognitive level could not cause increased cognitive activities, further reducing mortality risk.

As far as we know, this is the largest prospective cohort study to assess the association of cognitive activities with the risk of all-cause mortality among older adults with cognitive impairment. The main strengths of this study include the community-based design, the large sample, and the control as many potential confounding factors as possible, as well as the robust findings of sensitivity analyses. Nevertheless, our study still had some limitations. First, the current sample consisted of older Chinese adults, which might limit the applicability of the results to other populations. Furthermore, a majority of participants lacked formal education and resided in rural areas, potentially constraining the generalizability of the findings. Nevertheless, results from stratified analyses revealed similar patterns among participants who had received formal education and lived in urban settings. Second, because the study was a secondary analysis of data, we only investigated three kinds of cognitive activities, which did not represent the full scale of cognitive activities among older adults in China. Future study considering comprehensive cognitive activities needs to be performed to confirm the positive association between cognitive activities and reduced risk of mortality. Third, some potential factors, either unmeasured or unknown, might confound the association between cognitive activities and mortality. While, E-value analysis could suggest the robustness of the findings in some degree. Fourth, the factors assessed did not include information about the duration, possibly leading to imprecise measurements. Fifth, ‘reverse causality’ might interpretate the longitudinal association; however, we did not find substantial changes after excluding deaths within the first year or the first two years. Finally, since the activities of reading books/newspapers, as well as playing cards/mah-jong, were infrequently engaged in, these activities might play a minor role in the sample.

Conclusion

In conclusion, the study suggested that a greater frequency of participation in cognitive activities potentially resulted in lower risk of all-cause mortality among older adults with cognitive impairment. Furthermore, with an increased number of cognitive activities, mortality risk significantly decreased. Baseline cognitive function only mediated a small proportion of the benefits of cognitive activities in longevity among older adults with cognitive impairment; therefore, even if older adults experience cognitive decline, it is essential that they continue to engage positively in cognitive activities. Our findings may carry significant implications for public health policy and underscore the necessity of incorporating a diverse range of cognitive activities into the daily routines of older adults with cognitive impairment.

Supporting information

S1 Table. Definitions of baseline covariates in the present study.

https://doi.org/10.1371/journal.pone.0319093.s001

(PDF)

S2 Table. Distributions of baseline covariates with missing data.

https://doi.org/10.1371/journal.pone.0319093.s002

(PDF)

S3 Table. Baseline characteristics by reading books/newspapers.

https://doi.org/10.1371/journal.pone.0319093.s003

(PDF)

S4 Table. Baseline characteristics by playing cards/mah-jong.

https://doi.org/10.1371/journal.pone.0319093.s004

(PDF)

S5 Table. Baseline characteristics by watching TV or listening to radio.

https://doi.org/10.1371/journal.pone.0319093.s005

(PDF)

S6 Table. Associations of individual cognitive activities with all-cause mortality.

https://doi.org/10.1371/journal.pone.0319093.s006

(PDF)

S7 Table. Mediation analysis of individual cognitive activities by baseline cognitive function.

https://doi.org/10.1371/journal.pone.0319093.s007

(PDF)

S8 Table. Mediation analysis of baseline cognitive function by cognitive activities.

https://doi.org/10.1371/journal.pone.0319093.s008

(PDF)

S1 Fig. Testing proportional-hazards assumptions.

Note: The proportional-hazards assumption is an important premise for Cox proportional-hazards models; however, no hazards are perfectly proportional in nearly any clinical study _{(JAMA 2020;323:1401–1402)}. In small to moderate-sized samples, statistical tests may fail to reject such model assumptions due to insufficient power; however, even minor violations of the assumption can become evident with a sufficiently large sample size. Indeed, even when the model only approximately satisfies the proportional-hazards assumption, it can still yield reasonably accurate inferences _{(Am J Epidemiol 2024;193:926–927)}. With this consideration in mind, it’s best to use a combination of statistical tests and visual tests to determine the most serious violations. For statistical tests, we assessed the proportional-hazards assumption using the `cox.zph` function from the `survival` package, applying a p-value threshold of 0.001 ₍_①_{Am J Epidemiol 2024;193(6):926–927;} _②_{Nat Med 2024;30:85–97)}. For visual tests, we visually assessed this assumption by examining the relationship between β(t) of exposures and the timescale. We found no evidence indicating a potential violation of the proportional-hazards assumption for the exposure-outcome association, with a p-value from the Schoenfeld individual tests exceeding 0.001. In addition, the relationship between β(t) of exposures and the timescale could be characterized as approximately horizontal and constant. Overall, both statistical analyses and graphical representations supported that the proportional-hazards assumption was satisfied, or at least approximately met, allowing for data analysis using Cox proportional-hazards models.

https://doi.org/10.1371/journal.pone.0319093.s009

(PDF)

S2 Fig. Kaplan-Meier analysis showing cumulative incidence of all-cause mortality by the frequency of individual cognitive activities.

https://doi.org/10.1371/journal.pone.0319093.s010

(PDF)

S3 Fig. Stratified analyses by potential modifiers of the association between number of cognitive activities and all-cause mortality.

Note: Each stratification with adjustment for all factors (sex, age, education, marital status, residence, co-residence, regular intake of fruits, regular intake of vegetables, regular intake of meats, current smoking, current drinking, current regular exercise, hypertension, diabetes, heart diseases, cerebrovascular diseases, respiratory diseases, cancer, and self-rated health) except the stratification factor itself.

https://doi.org/10.1371/journal.pone.0319093.s011

(PDF)

References

1. Lipnicki DM, Crawford JD, Dutta R, Thalamuthu A, Kochan NA, Andrews G, et al; Cohort Studies of Memory in an International Consortium (COSMIC). Age-related cognitive decline and associations with sex, education and apolipoprotein E genotype across ethnocultural groups and geographic regions: a collaborative cohort study. PLoS Med. 2017;14(3):e1002261. pmid:28323832
- View Article
- PubMed/NCBI
- Google Scholar
2. Petersen RC, Lopez O, Armstrong MJ, Getchius TSD, Ganguli M, Gloss D, et al. Practice guideline update summary: Mild cognitive impairment: Report of the Guideline Development, Dissemination, and Implementation Subcommittee of the American Academy of Neurology. Neurology. 2018;90(3):126–35. pmid:29282327
- View Article
- PubMed/NCBI
- Google Scholar
3. Cao Q, Tan CC, Xu W, Hu H, Cao XP, Dong Q, et al. The prevalence of dementia: a systematic review and meta-analysis. J Alzheimers Dis. 2020;73(3):1157–66. pmid:31884487
- View Article
- PubMed/NCBI
- Google Scholar
4. Bassuk SS, Wypij D, Berkman LF. Cognitive impairment and mortality in the community-dwelling elderly. Am J Epidemiol. 2000;151(7):676–88. pmid:10752795
- View Article
- PubMed/NCBI
- Google Scholar
5. Perna L, Wahl HW, Mons U, Saum KU, Holleczek B, Brenner H. Cognitive impairment, all-cause and cause-specific mortality among non-demented older adults. Age Ageing. 2015;44(3):445–51. pmid:25468013
- View Article
- PubMed/NCBI
- Google Scholar
6. An R, Liu GG. Cognitive impairment and mortality among the oldest-old Chinese. Int J Geriatr Psychiatry. 2016;31(12):1345–53. pmid:26891153
- View Article
- PubMed/NCBI
- Google Scholar
7. Yates LA, Ziser S, Spector A, Orrell M. Cognitive leisure activities and future risk of cognitive impairment and dementia: systematic review and meta-analysis. Int Psychogeriatr. 2016;28(11):1791–806. pmid:27502691
- View Article
- PubMed/NCBI
- Google Scholar
8. Nilsen C, Agahi N, Shaw BA. Does the association between leisure activities and survival in old age differ by living arrangement? J Epidemiol Community Health. 2018;72(1):1–6. pmid:29079586
- View Article
- PubMed/NCBI
- Google Scholar
9. Goodman-Casanova JM, Guzman-Parra J, Guerrero G, Vera E, Barnestein-Fonseca P, Cortellessa G, et al. TV-based assistive integrated service to support European adults living with mild dementia or mild cognitive impairment (TV-AssistDem): study protocol for a multicentre randomized controlled trial. BMC Geriatr. 2019;19(1):247. pmid:31492113
- View Article
- PubMed/NCBI
- Google Scholar
10. Li ZH, Chen Q, Byers Kraus V, Shen D, Zhang XR, Zhang PD, et al. Leisure activities and disability in activities of daily living among the oldest-old Chinese population: evidence from the Chinese Longitudinal Healthy Longevity Study. Aging (Milano). 2020;12(11):10687–703. pmid:32532929
- View Article
- PubMed/NCBI
- Google Scholar
11. Li ZH, Zhang XR, Lv YB, Shen D, Li FR, Zhong WF, et al. Leisure activities and all-cause mortality among the chinese oldest-old population: a prospective community-based cohort study. J Am Med Dir Assoc. 2020;21(6):713–9.e2. pmid:31588027
- View Article
- PubMed/NCBI
- Google Scholar
12. McCollum L, Karlawish J. Cognitive impairment evaluation and management. Med Clin North Am. 2020;104(5):807–25. pmid:32773047
- View Article
- PubMed/NCBI
- Google Scholar
13. Dieser RB. How leisure activities affect health: the serious leisure perspective. Lancet Psychiat. 2021;8(7):562. pmid:34147170
- View Article
- PubMed/NCBI
- Google Scholar
14. 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 Psychiat. 2021;8(4):329–39. pmid:33581775
- View Article
- PubMed/NCBI
- Google Scholar
15. Liu X, Ruan Y, Huang L, Guo Y, Sun S, Chen H, et al. Cognitive leisure activity and all-cause mortality in older adults: a 4-year community-based cohort. BMC Geriatr. 2021;21(1):
- View Article
- Google Scholar
16. Silagi ML, Romero VU, de Oliveira MO, Tres ES, Brucki SMD, Radanovic M, et al. Inference comprehension from reading in individuals with mild cognitive impairment. Acta Neurol Belg. 2021;121(4):879–87. pmid:31970703
- View Article
- PubMed/NCBI
- Google Scholar
17. Bygren LO, Konlaan BB, Johansson SE. Attendance at cultural events, reading books or periodicals, and making music or singing in a choir as determinants for survival: Swedish interview survey of living conditions. BMJ. 1996;313(7072):1577–80. pmid:8990990
- View Article
- PubMed/NCBI
- Google Scholar
18. Sun R, Liu Y. Mortality of the oldest old in China. J Aging Health. 2006;18(1):37–55. pmid:16470963
- View Article
- PubMed/NCBI
- Google Scholar
19. Vaananen A, Murray M, Koskinen A, Vahtera J, Kouvonen A, Kivimaki M. Engagement in cultural activities and cause-specific mortality: prospective cohort study. Prev Med. 2009;49(2-3):142–7. pmid:19589351
- View Article
- PubMed/NCBI
- Google Scholar
20. Bavishi A, Slade MD, Levy BR. A chapter a day: Association of book reading with longevity. Soc Sci Med. 2016;164:44–8. pmid:27471129
- View Article
- PubMed/NCBI
- Google Scholar
21. Zeng Y, Feng Q, Hesketh T, Christensen K, Vaupel JW. Survival, disabilities in activities of daily living, and physical and cognitive functioning among the oldest-old in China: a cohort study. Lancet. 2017;389(10079):1619–29. pmid:28285816
- View Article
- PubMed/NCBI
- Google Scholar
22. Lv YB, Gao X, Yin ZX, Chen HS, Luo JS, Brasher MS, et al. Revisiting the association of blood pressure with mortality in oldest old people in China: community based, longitudinal prospective study. BMJ. 2018;361:k2158. pmid:29871897
- View Article
- PubMed/NCBI
- Google Scholar
23. Zhang Z, Gu D, Hayward MD. Childhood nutritional deprivation and cognitive impairment among older Chinese people. Soc Sci Med. 2010;71(5):941–9. pmid:20591545
- View Article
- PubMed/NCBI
- Google Scholar
24. Lv X, Li W, Ma Y, Chen H, Zeng Y, Yu X, et al. Cognitive decline and mortality among community-dwelling Chinese older people. BMC Med. 2019;17(1):63. pmid:30871536
- View Article
- PubMed/NCBI
- Google Scholar
25. Haneuse S, VanderWeele TJ, Arterburn D. Using the E-value to assess the potential effect of unmeasured confounding in observational studies. JAMA. 2019;321(6):602–3. pmid:30676631
- View Article
- PubMed/NCBI
- Google Scholar
26. VanderWeele TJ. Causal mediation analysis with survival data. Epidemiology. 2011;22(4):582–5. pmid:21642779
- View Article
- PubMed/NCBI
- Google Scholar
27. 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–e2. pmid:34028370
- View Article
- PubMed/NCBI
- Google Scholar
28. Li Y, Yoshida K, Kaufman JS, Mathur MB. A brief primer on conducting regression-based causal mediation analysis. Psychol Trauma. 2023;15(6):930–8. pmid:36701540
- View Article
- PubMed/NCBI
- Google Scholar
29. Agahi N, Parker MG. Leisure activities and mortality: does gender matter? J Aging Health. 2008;20(7):855–71. pmid:18815413
- View Article
- PubMed/NCBI
- Google Scholar
30. Verghese J, Lipton RB, Katz MJ, Hall CB, Derby CA, Kuslansky G, et al. Leisure activities and the risk of dementia in the elderly. N Engl J Med. 2003;348(25):2508–16. pmid:12815136
- View Article
- PubMed/NCBI
- Google Scholar
31. Lindstrom HA, Fritsch T, Petot G, Smyth KA, Chen CH, Debanne SM, et al. The relationships between television viewing in midlife and the development of Alzheimer’s disease in a case-control study. Brain Cogn. 2005;58(2):157–65. pmid:15919546
- View Article
- PubMed/NCBI
- Google Scholar
32. Menec VH. The relation between everyday activities and successful aging: a 6-year longitudinal study. J Gerontol Ser B Psychol Sci Soc Sci. 2003;58(2):S74–82. pmid:12646596
- View Article
- PubMed/NCBI
- Google Scholar
33. Holt-Lunstad J, Smith TB, Layton JB. Social relationships and mortality risk: a meta-analytic review. PLoS Med. 2010;7(7):e1000316. pmid:20668659
- View Article
- PubMed/NCBI
- Google Scholar
34. Grøntved A, Hu FB. Television viewing and risk of type 2 diabetes, cardiovascular disease, and all-cause mortality: a meta-analysis. JAMA. 2011;305(23):2448–55. pmid:21673296
- View Article
- PubMed/NCBI
- Google Scholar
35. Wijndaele K, Brage S, Besson H, Khaw KT, Sharp SJ, Luben R, et al. Television viewing time independently predicts all-cause and cardiovascular mortality: the EPIC Norfolk study. Int J Epidemiol. 2011;40(1):150–9. pmid:20576628
- View Article
- PubMed/NCBI
- Google Scholar
36. Wang HX, Mittleman MA, Orth-Gomer K. Influence of social support on progression of coronary artery disease in women. Soc Sci Med. 2005;60(3):599–607. pmid:15550307
- View Article
- PubMed/NCBI
- Google Scholar
37. Glass TA, de Leon CM, Marottoli RA, Berkman LF. Population based study of social and productive activities as predictors of survival among elderly Americans. BMJ. 1999;319(7208):478–83. pmid:10454399
- View Article
- PubMed/NCBI
- Google Scholar

[ref1] 1. Lipnicki DM, Crawford JD, Dutta R, Thalamuthu A, Kochan NA, Andrews G, et al; Cohort Studies of Memory in an International Consortium (COSMIC). Age-related cognitive decline and associations with sex, education and apolipoprotein E genotype across ethnocultural groups and geographic regions: a collaborative cohort study. PLoS Med. 2017;14(3):e1002261. pmid:28323832
View Article
PubMed/NCBI
Google Scholar

[2] View Article

[3] PubMed/NCBI

[4] Google Scholar

[ref2] 2. Petersen RC, Lopez O, Armstrong MJ, Getchius TSD, Ganguli M, Gloss D, et al. Practice guideline update summary: Mild cognitive impairment: Report of the Guideline Development, Dissemination, and Implementation Subcommittee of the American Academy of Neurology. Neurology. 2018;90(3):126–35. pmid:29282327
View Article
PubMed/NCBI
Google Scholar

[6] View Article

[7] PubMed/NCBI

[8] Google Scholar

[ref3] 3. Cao Q, Tan CC, Xu W, Hu H, Cao XP, Dong Q, et al. The prevalence of dementia: a systematic review and meta-analysis. J Alzheimers Dis. 2020;73(3):1157–66. pmid:31884487
View Article
PubMed/NCBI
Google Scholar

[10] View Article

[11] PubMed/NCBI

[12] Google Scholar

[ref4] 4. Bassuk SS, Wypij D, Berkman LF. Cognitive impairment and mortality in the community-dwelling elderly. Am J Epidemiol. 2000;151(7):676–88. pmid:10752795
View Article
PubMed/NCBI
Google Scholar

[14] View Article

[15] PubMed/NCBI

[16] Google Scholar

[ref5] 5. Perna L, Wahl HW, Mons U, Saum KU, Holleczek B, Brenner H. Cognitive impairment, all-cause and cause-specific mortality among non-demented older adults. Age Ageing. 2015;44(3):445–51. pmid:25468013
View Article
PubMed/NCBI
Google Scholar

[18] View Article

[19] PubMed/NCBI

[20] Google Scholar

[ref6] 6. An R, Liu GG. Cognitive impairment and mortality among the oldest-old Chinese. Int J Geriatr Psychiatry. 2016;31(12):1345–53. pmid:26891153
View Article
PubMed/NCBI
Google Scholar

[22] View Article

[23] PubMed/NCBI

[24] Google Scholar

[ref7] 7. Yates LA, Ziser S, Spector A, Orrell M. Cognitive leisure activities and future risk of cognitive impairment and dementia: systematic review and meta-analysis. Int Psychogeriatr. 2016;28(11):1791–806. pmid:27502691
View Article
PubMed/NCBI
Google Scholar

[26] View Article

[27] PubMed/NCBI

[28] Google Scholar

[ref8] 8. Nilsen C, Agahi N, Shaw BA. Does the association between leisure activities and survival in old age differ by living arrangement? J Epidemiol Community Health. 2018;72(1):1–6. pmid:29079586
View Article
PubMed/NCBI
Google Scholar

[30] View Article

[31] PubMed/NCBI

[32] Google Scholar

[ref9] 9. Goodman-Casanova JM, Guzman-Parra J, Guerrero G, Vera E, Barnestein-Fonseca P, Cortellessa G, et al. TV-based assistive integrated service to support European adults living with mild dementia or mild cognitive impairment (TV-AssistDem): study protocol for a multicentre randomized controlled trial. BMC Geriatr. 2019;19(1):247. pmid:31492113
View Article
PubMed/NCBI
Google Scholar

[34] View Article

[35] PubMed/NCBI

[36] Google Scholar

[ref10] 10. Li ZH, Chen Q, Byers Kraus V, Shen D, Zhang XR, Zhang PD, et al. Leisure activities and disability in activities of daily living among the oldest-old Chinese population: evidence from the Chinese Longitudinal Healthy Longevity Study. Aging (Milano). 2020;12(11):10687–703. pmid:32532929
View Article
PubMed/NCBI
Google Scholar

[38] View Article

[39] PubMed/NCBI

[40] Google Scholar

[ref11] 11. Li ZH, Zhang XR, Lv YB, Shen D, Li FR, Zhong WF, et al. Leisure activities and all-cause mortality among the chinese oldest-old population: a prospective community-based cohort study. J Am Med Dir Assoc. 2020;21(6):713–9.e2. pmid:31588027
View Article
PubMed/NCBI
Google Scholar

[42] View Article

[43] PubMed/NCBI

[44] Google Scholar

[ref12] 12. McCollum L, Karlawish J. Cognitive impairment evaluation and management. Med Clin North Am. 2020;104(5):807–25. pmid:32773047
View Article
PubMed/NCBI
Google Scholar

[46] View Article

[47] PubMed/NCBI

[48] Google Scholar

[ref13] 13. Dieser RB. How leisure activities affect health: the serious leisure perspective. Lancet Psychiat. 2021;8(7):562. pmid:34147170
View Article
PubMed/NCBI
Google Scholar

[50] View Article

[51] PubMed/NCBI

[52] Google Scholar

[ref14] 14. 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 Psychiat. 2021;8(4):329–39. pmid:33581775
View Article
PubMed/NCBI
Google Scholar

[54] View Article

[55] PubMed/NCBI

[56] Google Scholar

[ref15] 15. Liu X, Ruan Y, Huang L, Guo Y, Sun S, Chen H, et al. Cognitive leisure activity and all-cause mortality in older adults: a 4-year community-based cohort. BMC Geriatr. 2021;21(1):
View Article
Google Scholar

[58] View Article

[59] Google Scholar

[ref16] 16. Silagi ML, Romero VU, de Oliveira MO, Tres ES, Brucki SMD, Radanovic M, et al. Inference comprehension from reading in individuals with mild cognitive impairment. Acta Neurol Belg. 2021;121(4):879–87. pmid:31970703
View Article
PubMed/NCBI
Google Scholar

[61] View Article

[62] PubMed/NCBI

[63] Google Scholar

[ref17] 17. Bygren LO, Konlaan BB, Johansson SE. Attendance at cultural events, reading books or periodicals, and making music or singing in a choir as determinants for survival: Swedish interview survey of living conditions. BMJ. 1996;313(7072):1577–80. pmid:8990990
View Article
PubMed/NCBI
Google Scholar

[65] View Article

[66] PubMed/NCBI

[67] Google Scholar

[ref18] 18. Sun R, Liu Y. Mortality of the oldest old in China. J Aging Health. 2006;18(1):37–55. pmid:16470963
View Article
PubMed/NCBI
Google Scholar

[69] View Article

[70] PubMed/NCBI

[71] Google Scholar

[ref19] 19. Vaananen A, Murray M, Koskinen A, Vahtera J, Kouvonen A, Kivimaki M. Engagement in cultural activities and cause-specific mortality: prospective cohort study. Prev Med. 2009;49(2-3):142–7. pmid:19589351
View Article
PubMed/NCBI
Google Scholar

[73] View Article

[74] PubMed/NCBI

[75] Google Scholar

[ref20] 20. Bavishi A, Slade MD, Levy BR. A chapter a day: Association of book reading with longevity. Soc Sci Med. 2016;164:44–8. pmid:27471129
View Article
PubMed/NCBI
Google Scholar

[77] View Article

[78] PubMed/NCBI

[79] Google Scholar

[ref21] 21. Zeng Y, Feng Q, Hesketh T, Christensen K, Vaupel JW. Survival, disabilities in activities of daily living, and physical and cognitive functioning among the oldest-old in China: a cohort study. Lancet. 2017;389(10079):1619–29. pmid:28285816
View Article
PubMed/NCBI
Google Scholar

[81] View Article

[82] PubMed/NCBI

[83] Google Scholar

[ref22] 22. Lv YB, Gao X, Yin ZX, Chen HS, Luo JS, Brasher MS, et al. Revisiting the association of blood pressure with mortality in oldest old people in China: community based, longitudinal prospective study. BMJ. 2018;361:k2158. pmid:29871897
View Article
PubMed/NCBI
Google Scholar

[85] View Article

[86] PubMed/NCBI

[87] Google Scholar

[ref23] 23. Zhang Z, Gu D, Hayward MD. Childhood nutritional deprivation and cognitive impairment among older Chinese people. Soc Sci Med. 2010;71(5):941–9. pmid:20591545
View Article
PubMed/NCBI
Google Scholar

[89] View Article

[90] PubMed/NCBI

[91] Google Scholar

[ref24] 24. Lv X, Li W, Ma Y, Chen H, Zeng Y, Yu X, et al. Cognitive decline and mortality among community-dwelling Chinese older people. BMC Med. 2019;17(1):63. pmid:30871536
View Article
PubMed/NCBI
Google Scholar

[93] View Article

[94] PubMed/NCBI

[95] Google Scholar

[ref25] 25. Haneuse S, VanderWeele TJ, Arterburn D. Using the E-value to assess the potential effect of unmeasured confounding in observational studies. JAMA. 2019;321(6):602–3. pmid:30676631
View Article
PubMed/NCBI
Google Scholar

[97] View Article

[98] PubMed/NCBI

[99] Google Scholar

[ref26] 26. VanderWeele TJ. Causal mediation analysis with survival data. Epidemiology. 2011;22(4):582–5. pmid:21642779
View Article
PubMed/NCBI
Google Scholar

[101] View Article

[102] PubMed/NCBI

[103] Google Scholar

[ref27] 27. 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–e2. pmid:34028370
View Article
PubMed/NCBI
Google Scholar

[105] View Article

[106] PubMed/NCBI

[107] Google Scholar

[ref28] 28. Li Y, Yoshida K, Kaufman JS, Mathur MB. A brief primer on conducting regression-based causal mediation analysis. Psychol Trauma. 2023;15(6):930–8. pmid:36701540
View Article
PubMed/NCBI
Google Scholar

[109] View Article

[110] PubMed/NCBI

[111] Google Scholar

[ref29] 29. Agahi N, Parker MG. Leisure activities and mortality: does gender matter? J Aging Health. 2008;20(7):855–71. pmid:18815413
View Article
PubMed/NCBI
Google Scholar

[113] View Article

[114] PubMed/NCBI

[115] Google Scholar

[ref30] 30. Verghese J, Lipton RB, Katz MJ, Hall CB, Derby CA, Kuslansky G, et al. Leisure activities and the risk of dementia in the elderly. N Engl J Med. 2003;348(25):2508–16. pmid:12815136
View Article
PubMed/NCBI
Google Scholar

[117] View Article

[118] PubMed/NCBI

[119] Google Scholar

[ref31] 31. Lindstrom HA, Fritsch T, Petot G, Smyth KA, Chen CH, Debanne SM, et al. The relationships between television viewing in midlife and the development of Alzheimer’s disease in a case-control study. Brain Cogn. 2005;58(2):157–65. pmid:15919546
View Article
PubMed/NCBI
Google Scholar

[121] View Article

[122] PubMed/NCBI

[123] Google Scholar

[ref32] 32. Menec VH. The relation between everyday activities and successful aging: a 6-year longitudinal study. J Gerontol Ser B Psychol Sci Soc Sci. 2003;58(2):S74–82. pmid:12646596
View Article
PubMed/NCBI
Google Scholar

[125] View Article

[126] PubMed/NCBI

[127] Google Scholar

[ref33] 33. Holt-Lunstad J, Smith TB, Layton JB. Social relationships and mortality risk: a meta-analytic review. PLoS Med. 2010;7(7):e1000316. pmid:20668659
View Article
PubMed/NCBI
Google Scholar

[129] View Article

[130] PubMed/NCBI

[131] Google Scholar

[ref34] 34. Grøntved A, Hu FB. Television viewing and risk of type 2 diabetes, cardiovascular disease, and all-cause mortality: a meta-analysis. JAMA. 2011;305(23):2448–55. pmid:21673296
View Article
PubMed/NCBI
Google Scholar

[133] View Article

[134] PubMed/NCBI

[135] Google Scholar

[ref35] 35. Wijndaele K, Brage S, Besson H, Khaw KT, Sharp SJ, Luben R, et al. Television viewing time independently predicts all-cause and cardiovascular mortality: the EPIC Norfolk study. Int J Epidemiol. 2011;40(1):150–9. pmid:20576628
View Article
PubMed/NCBI
Google Scholar

[137] View Article

[138] PubMed/NCBI

[139] Google Scholar

[ref36] 36. Wang HX, Mittleman MA, Orth-Gomer K. Influence of social support on progression of coronary artery disease in women. Soc Sci Med. 2005;60(3):599–607. pmid:15550307
View Article
PubMed/NCBI
Google Scholar

[141] View Article

[142] PubMed/NCBI

[143] Google Scholar

[ref37] 37. Glass TA, de Leon CM, Marottoli RA, Berkman LF. Population based study of social and productive activities as predictors of survival among elderly Americans. BMJ. 1999;319(7208):478–83. pmid:10454399
View Article
PubMed/NCBI
Google Scholar

[145] View Article

[146] PubMed/NCBI

[147] Google Scholar

Figures

Abstract

Background

Methods

Results

Conclusions

Introduction

Methods

Study participants

Assessment of cognitive activities

Study outcome

Assessment of covariates

Statistical analysis

Results

Baseline characteristics

Associations between cognitive activities and all-cause mortality

Stratified and sensitivity analyses

Mediation analysis

Discussion

Conclusion

Supporting information

S1 Table. Definitions of baseline covariates in the present study.

S2 Table. Distributions of baseline covariates with missing data.

S3 Table. Baseline characteristics by reading books/newspapers.

S4 Table. Baseline characteristics by playing cards/mah-jong.

S5 Table. Baseline characteristics by watching TV or listening to radio.

S6 Table. Associations of individual cognitive activities with all-cause mortality.

S7 Table. Mediation analysis of individual cognitive activities by baseline cognitive function.

S8 Table. Mediation analysis of baseline cognitive function by cognitive activities.

S1 Fig. Testing proportional-hazards assumptions.

S2 Fig. Kaplan-Meier analysis showing cumulative incidence of all-cause mortality by the frequency of individual cognitive activities.

S3 Fig. Stratified analyses by potential modifiers of the association between number of cognitive activities and all-cause mortality.

References