Mental Activity and All-Cause Mortality in Older Adults: A 4-Year Community-Based Cohort

The mental activity, such as reading, playing mahjong or cards and computer use, is common among older adults in China. Previous researches suggest a protective role of mental activity against cognitive impairment. However, the relationship between mental activity and all-cause mortality has rarely been reported. the effect of mental activity on in


Page 3/19
Older population has been increasing around the world, presenting a major challenge to health and social care system. Chronic diseases are the leading causes of death and disability worldwide 1 . Plenty of research has implicated lifestyle risk behaviors, such as smoking 2 , alcohol use 3 and physical inactivity 4 et al, in adverse health outcomes, including cardiovascular disease, dementia, diabetes, some cancers, as well as mortality 5 . Hence, substantial disease, mortality, and economic burden could be prevented through modi cation of lifestyle behaviors [6][7][8] .
Mental activity, such as computer use, reading and playing mahjong or cards, is a type of modi able lifestyle behavior and popular in old adults especially in the retired people in China. In the past few years, the bene cial effect of mental activity on cognitive function has been reported. For example, computer use has been reported to improve cognitive function in older people 9 . Several studies have also identi ed that reading [10][11][12][13][14] , playing board games (mahjong, chess or poker) 15,16 , and playing cards 14,17 were associated with reduced risk of cognitive impairment. It is noteworthy that dementia is one of the most common cognitive-related disorders, ranks as the sixth leading cause of death in the United States and the fth leading cause of death in Americans age ≥ 65 years 18 . It is projected that, by 2050, 1.6 million or 43% of older adult deaths will be due to dementia and Alzheimer's disease 19 . In addition, accumulating evidence have indicated that leisure activity, including watching TV 20 , internet use 21 and reading 22,23 , can make a signi cant contribution to overall life satisfaction [24][25][26] , which have been identi ed as an important risk factor of mortality in older people [27][28][29] . All these studies compel us to examine whether mental activity is associated with all-cause mortality, which has rarely been reported.
Using a 4-year prospective cohort study, the present study explores a range of mental activity, including reading, playing mahjong or cards, and computer use. The objective of this study is to examine the association between mental activity and all-cause mortality.

Sampling and procedures
We recruited a random sample of 4050 participants, representative of the non-institutionalized population age ≥ 60 years in Songjiang District, Shanghai, China. Baseline data collection was conducted from June 2015 to March 2016. At the baseline, demographic and characteristic data was collected via a face-toface questionnaire survey by trained personnel, including birth date, gender, height, weight, education years, lifestyles, Physical activity(PA), mental activity (reading, playing mahjong or cards and computer use), medical histories of diabetes, hypertension, coronary heart disease (CHD) and stroke (classi ed as yes or no), et al. Participants joined the study by completing the questionnaire and the written informed consent form.

Measures
Mortality. All-cause mortality and the date of death were ascertained from the Death Surveillance System of Songjiang CDC for all participants after each follow-up, from July 2015 to November 2018. Research coordinators contacted all the participants and asked for the availability of a clinical interview. Those who could not be traced, refused to participate were de ned as "lost-to-follow up". Participants who missed any of three data of reading, playing mahjong or cards and computer use at baseline were excluded.
Mental activity index construction. Participants reported on a range of mental activities in the questionnaire. Reading status was derived from the question: "Do you read books or newspapers every day?" With "hardly reading" de ned as having no reading habits, while "occasionally reading" and "daily reading" de ned as having reading habits. Participants were asked, "Do you often play cards or mahjong?" Among them, "almost do not play " was de ned as having no habit of playing cards or mahjong, "several times a month" and "several times a week" are de ned as having the habit of playing cards or mahjong. A similar question was also asked, "Do you often use computers to access the internet?" With "Not at all" de ned as having no Internet habits, "not every day" and "every day (more than an hour at a time)" de ned as having internet habit.
Covariates. Sociodemographic characteristics were collected from participants' self-reports or physical examination. Age, sex, BMI (underweight, normal, overweight and obesity), education (illiteracy, primary school and ≥ junior school), marital status (married and single), and work status (retired, still working, no work). BMI was calculated as weight in kilograms divided by height in meters squared. Based on the BMI classi cation guidelines of the World Health Organization revised for the Asia-Paci c region, we classi ed the MCI participants into underweight (BMI < 18.5 kg/m 2 ), normal (BMI: 18.5 ~ 22.9 kg/m 2 ), overweight (BMI: 23.0 ~ 29.9 kg/m 2 ) and obesity (BMI ≥ 30.0 kg/m 2 ). Smoking status was categorized as current smokers, never smokers, and people who given up smoking. Drinking status was divided into drinking and never drinking. PA was assessed based on self-report of leisure-time activities, such as fasting walking, playing ball, running, or qigong. (Average physical-activity time must exceed 10 minutes per day.) Participants rated their PA levels as (1) inactive, (2) several times a month, (3) 3-4×/week, or (4) almost every day. In addition, we created a dichotomous variable for cardiovascular or metabolic disease, based on the self-report of CHD, stroke, hypertension, and diabetes. Based on the sample distribution, the index of cardiovascular or metabolic disease (CHD, stroke, hypertension, diabetes) was categorized as 0 and 1 (at least one disease). We created an additional dichotomous variable for cancer, based on selfreport for the prior to baseline data collection.

Statistical Analysis
All statistical analyses were performed using SPSS version 22.0 (SPSS, Chicago, IL, USA). Hazard ratios (HR) and 95% con dence intervals (CI) were estimated using Cox proportional hazard models for the analysis of association between MAI and mortality. Kaplan-Meier curves was also used to estimate the relationship between MAI and mortality. The outcome variable was survival time, which was measured as the time interval from the date of baseline data collection to death or censoring. All Cox proportional hazards regression models were adjusted for sex, age (continuous variable), BMI (continuous variable), educational attainment, marital status, work status, smoking status, drinking status, and PA, with covariates classi ed categorically as per Table 1. We also examined the independent association of each mental activity and all-cause mortality.  26.7% had one mental activity, and 6.1% and 0.35% had a MAI score of 2 and 3, respectively. Higher MAI scores were more prevalent among males, those aged 60-70y, those who were married, those had a junior school degree or higher, and those who were retired(P 0.05, Table 1).

Individual Mental Activity and All-Cause Mortality
When all three dichotomized individual mental activities were entered in the model with all covariates, playing mahjong or cards showed independent associations with all-cause mortality (P = 0.007, supplementary Figure S1). Reading and computer use also displayed potential bene cial role in all-cause mortality, but with no signi cant association with all-cause mortality (supplementary Figure S1).

Mental Activity Index and All-Cause Mortality
Kaplan-Meier survival analysis showed that participants with higher MAI score have signi cantly decreased risk of death (P = 0.043, Fig. 1). Cox proportional hazards regression analyses also showed the inverse association between the MAI scores and all-cause mortality (HR = 0.72, 95%CI 0.54-0.96, P = 0.025), adjusted for age, sex, BMI, educational attainment, marital status, work status, smoking status, alcohol use status, and physical activity (Fig. 2). All-cause mortality HRs compared to individuals without mental activity were 0.71 (P = 0.045) and 0.47 (P = 0.049) for those with 1 and 2 mental activities in univariate analysis, while these signi cances were not found in multivariate analysis (Fig. 2).

Discussion
This is the rst study to our knowledge to investigate a MAI incorporating reading, playing mahjong or cards and computer use in relation to all-cause mortality. We found that multiple mental activities among older Chinese adults were associated with a decreased risk for all-cause mortality over 4 y of follow-up. There was a clear association between the number of mental activities, as indicated by MAI score, and all-cause mortality.
Previous evidence is accumulating on the cognitive health of the mental activities. Li and colleagues indicated that reading and computer use were associated with lower risk of mild cognitive impairment in a population-based study 34 . Lindstrom et al. found an inverse relationship between intellectual activities (reading, playing cards, playing a musical instrument, and letter writing) and Alzheimer's disease or other forms of dementia in a US-based population 17 . Verghese et al reported that cognitive activities (reading, writing, doing crossword puzzles, playing board games or cards, and playing musical instruments) were associated with a reduced risk of dementia 14 . Despite the heterogeneous measures, risk classi cation, sample characteristics, and follow-up time of these studies, the association between mental activities and cognitive health has been consistent, suggesting the generalizability of these ndings. Such bene cial role of mental activities is furthered here by implicating its protective role against all-cause mortality in older people in our study. Cognitive impairment has a signi cant impact on mortality and disability of order population 35 . According to data from the Centers for Disease Control and Prevention (CDC), 121,404 people died from Alzheimer's disease in 2017 and the rate of death from Alzheimer's disease dramatically with age, especially after age 65 18 . Therefore, a potential mechanism by which mental activities in uence mortality is through protecting the cognitive impairment, at least in part.
It is worthy to note that, among the three dichotomized individual mental activities, playing mahjong or cards showed independent association with all-cause mortality. One explanation is that playing mahjong or cards incorporates social engagement. Social engagement, de ned as the maintenance of many social connections and a high level of participation in social activities, has been indicated to prevent cognitive decline in older persons [36][37][38] . Additionally, social activities predominantly affect the immune system and in uence in ammatory processes in the brain 39,40 . All these results support our ndings that playing mahjong or cards has an independent protective role in all-cased mortality.
Strati ed analysis showed a signi cant relation between mental activity and all-cause mortality among participants with physical inactivity in late life, indicating the supplemented role of mental activity in healthy living, especially for the older people who is unable to perform effective physical activity due to severe chronic disease. Physical activity is a pivotal lifestyle behavior. Regular physical activity has been irrefutably identi ed as protective factor for all-cause mortality [41][42][43] , and the bene t of physical activity was independent of the type of physical activity 44 . Here our study showed the consistent effect of mental activity with PA on all-cause mortality in older population. In addition, some studies have indicated that the cognitive function and physical function in uenced each other in a feedback loop. 45,46 . A protective effect of physical activity against cognitive impairment has been reported in many studies [47][48][49][50] and the bene ts of physical activity on cognitive function can be attributed to an ameliorated overall health condition 51 . Conversely, mental activity has been reported to be associated with enhanced memory, executive function, language, and cognitive skill 52 , which may in uence the practice of regular physical activity. For example, the execution functions, including of volition, planning, purposive action, performance monitoring and inhibition 53 , may enable people to consistently engage in physical activity in older to achieve long-term health bene ts 54 .
We also found that MAI score was associated with lower all-cause mortality in participants without cancer, but not in cancer patients. In fact, many studies showed the bene cial effect of mental activity or social activity on the quality of life which was reported to decreased the risk of breast cancer mortality and recurrence 55 , enhanced the colorectal cancer overall survival 56 , and in uenced the cancer patient outcomes, including physical burden, psychosocial burden, and nancial burden 57 . Hence, we believe that whether from improving the mental health of cancer patients or improving the survival rate of non-cancer patients, mental activity should be concerned in older people health. In addition, we also found the potential protective role of MAI in all-cause mortality among participants who were diagnosed with cardiovascular or metabolic disease (HR = 0.67, P = 0.025 in univariate analysis, HR = 0.70, P = 0.067 in multivariate analysis, Fig. 3). Accumulating evidence have indicated that leisure activity, including watching TV 20 , internet use 21 and reading 22,23 , can make a signi cant contribution to overall life satisfaction and psychological well-being 24,25 , which in turn is associated with lower risk of cardiovascular disease 58,59 . Thus, a potential pathway by which mental activity in uence all-cause mortality may be through reducing the risk of cardiovascular disease or reducing the effect of cardiovascular disease on mortality.
Limitations in the current study should be acknowledged. Firstly, it is important to acknowledge that not all three mental activities contribute to mortality similarly and that their combined effects may not be additive. However, because of the short follow-up period and small sample size, we didn't get the enough prevalence of speci c combination pattern of mental activities to analysis their associations with allcause mortality (e.g., prevalence of combination of reading and computer use, combination of playing mahjong or cards and computer use, and combination of both three mental activities were 1.0% (n = 41), 0.2% (n = 9) and 0.4% (n = 14), respectively). Secondly, time spent in each activity was not measured, this may modify the effect of mental activity on mortality. Thirdly, the effect of mental activity on mortality was not adjusted for cognition status because there was an absence of measure of cognition at baseline. Baseline participation in mental activities may have been in uenced by cognition and future studies incorporating the cognitive data are needed to illustrate the modifying effect of cognition status on mortality caused by mental activity. Fourthly, this study could be further strengthened by including causespeci c mortality outcomes, but these data are not yet available for the time period studied. Finally, this study was composed of older Chinese adults living in a large city, Shanghai, thus potentially limiting the generalizability of our results.

Conclusion
This study demonstrates the importance of mental activity in health lifestyle, here evidenced for adults aged 60 y and older. This analysis investigated three mental activities, namely, reading, playing mahjong or cards and using the computers, which may be added to behavioral indices or risk combinations to quantify health risk of the older people in China. In addition, our ndings advance current knowledge on the older people health and provide a new prevention strategy in older populations. Availability of data and materials:

Abbreviations
Datasets used during the current study are available from the corresponding author on reasonable request and with permission of The National Center for Chronic and Noncommunicable Disease Control and Prevention.  Crude cumulative death rates and hazard ratios for all-cause mortality by mental activity index score among a community-based Chinese elderly samples (2015-2018, n = 4003). HR = hazard ratio; CI = con dence interval. a HR adjusted for age, sex, BMI, smoking status, alcohol use status, marital status, education level, work status, physical activity.