Original Paper
1Department of Hospital Epidemiology and Infection Control, Sir Run Run Shaw Hospital, Zhejiang University School of Medicine, Hangzhou, China
2School of Public Health, Capital Medical University, Beijing, China
3Department of Clinical Nutrition, Beijing Hospital, Chinese Academy of Medical Sciences, Beijing, China
*these authors contributed equally
Corresponding Author:
Jin Zhao, MD
Department of Hospital Epidemiology and Infection Control
Sir Run Run Shaw Hospital
Zhejiang University School of Medicine
3 East Qingchun Road
Hangzhou, 310016
China
Phone: 86 571 86006435
Fax:86 571 86006435
Email: enqizhao@zju.edu.cn
Abstract
Background: Cognitive decline poses one of the greatest global challenges for health and social care, particularly in China, where the burden on the older adult population is most pronounced. Despite the rapid expansion of internet access, there is still limited understanding of the long-term cognitive impacts of internet use among middle-aged and older adults.
Objective: This study aims to explore the association between internet use and age-related cognitive decline among middle-aged and older Chinese adults. To gain a more comprehensive understanding of the effects of internet use, we also focused on assessing the impact of both the frequency of internet use and the types of internet devices on cognition. Moreover, we assessed the mediating role of internet use on cognitive function for characteristics significantly linked to cognition in stratified analysis.
Methods: We analyzed data based on 12,770 dementia-free participants aged ≥45 years from the China Health and Retirement Longitudinal Study. We used a fixed effects model to assess the relationship between internet use and cognitive decline and further validated it using multiple linear regression analysis, generalized estimating equations, propensity score matching, inverse probability of treatment weighting, and overlap weighting. We further examined the varying effects of internet device type and frequency on cognitive function using fixed effects models and Spearman rank correlations. The Karlson-Holm-Breen method was used to estimate the mediating role of internet use in the urban-rural cognitive gap.
Results: Participants using the internet (n=1005) were younger, more likely to be male, more educated, married, retired and living in an urban area and had higher cognitive assessment scores than nonusers (n=11,765). After adjusting for demographic and health-related risk factors, there was a positive correlation between internet use and cognitive function (β=0.551, 95% CI 0.391-0.710). Over the follow-up period, persistent internet users had a markedly lower 5-year incidence of neurodegenerative diseases, at 2.2% (15/671), compared with nonusers, at 5.3% (379/7099; P<.001). The negative impact of aging (>50 years) on cognitive function was consistently less pronounced among internet users than among nonusers. Furthermore, increased frequency of internet use was associated with greater cognitive benefits for middle-aged and older adults (rs=0.378, P<.001). Among digital devices used for internet access, cell phones (β=0.398, 95% CI 0.283-0.495) seemed to have a higher level of cognitive protection than computers (β=0.147, 95% CI 0.091-0.204). The urban-rural disparity in cognitive function was partially attributed to the disparity in internet use (34.2% of the total effect, P<.001).
Conclusions: This study revealed that the use of internet by individuals aged 45 years and older is associated with a reduced risk of cognitive decline. Internet use has the potential to be a viable, cost-effective, nonpharmacological intervention for cognitive decline among middle-aged and older adults.
doi:10.2196/57301
Keywords
Introduction
With the rapid increase in the aging population and the extension of life expectancy, age-related cognitive impairment has emerged as an influential global public health concern [Pais R, Ruano L, P Carvalho O, Barros H. Global cognitive impairment prevalence and incidence in community dwelling older adults-a systematic review. Geriatrics (Basel). Oct 27, 2020;5(4):84. [FREE Full text] [CrossRef] [Medline]1]. Cognitive decline is an irreversible cerebral pathological and physiological progression, serving as both a precursor to dementia and a potential catalyst for physical disability and mortality [Prince MJ, Wu F, Guo Y, Gutierrez Robledo LM, O'Donnell M, Sullivan R, et al. The burden of disease in older people and implications for health policy and practice. Lancet. Feb 07, 2015;385(9967):549-562. [CrossRef] [Medline]2]. A nationwide survey conducted in 2019 revealed that China has become the country with the highest number of individuals with dementia, as the population aged 60 years and older has soared to a staggering 253 million people, among which 6.0% have dementia [Liu J, Wang L, Tan J. Dementia in China: Current status. Neurology. Sep 17, 2013;81(12):1077-1078. [CrossRef]3,Jia L, Du Y, Chu L, Zhang Z, Li F, Lyu D, et al. Prevalence, risk factors, and management of dementia and mild cognitive impairment in adults aged 60 years or older in China: a cross-sectional study. The Lancet Public Health. Dec 2020;5(12):e661-e671. [CrossRef]4]. Cognitive impairment typically emerges in individuals aged 45 years to 60 years and not only imposes a significant burden on the patients’ families but also presents formidable challenges to the health care system and society [Frankish H, Horton R. Prevention and management of dementia: a priority for public health. The Lancet. Dec 2017;390(10113):2614-2615. [CrossRef]5]. Hence, the discovery of nonpharmacological strategies to intervene for middle-aged and older adults is crucial to reducing the incidence of cognitive decline, slowing the progression of cognitive impairment, and improving health and well-being.
Regarding the prevention of dementia, establishing a healthy lifestyle and implementing nonpharmacological interventions are critically important. Based on randomized clinical trials and meta-analyses, consistent research-based evidence indicates that social and cognitively stimulating activities can delay cognitive decline, particularly in middle-aged and older adults who are most susceptible to cognitive impairment [Satizabal CL, Beiser AS, Chouraki V, Chêne G, Dufouil C, Seshadri S. Incidence of dementia over three decades in the Framingham Heart Study. N Engl J Med. Feb 11, 2016;374(6):523-532. [CrossRef]6-García-Casal JA, Loizeau A, Csipke E, Franco-Martín M, Perea-Bartolomé MV, Orrell M. Computer-based cognitive interventions for people living with dementia: a systematic literature review and meta-analysis. Aging Ment Health. May 2017;21(5):454-467. [CrossRef] [Medline]8].
With the rapid development of the internet and the widespread use of smartphones, digital literacy is increasing in China, particularly among middle-aged and older adults. According to the 52nd Statistical Report on Internet Development in China, released by the China Internet Network Information Center [The 52nd Statistical Report on China’s Internet Development. China Internet Network Information Center (CNNIC). Aug 2023. URL: https://www.cnnic.com.cn/IDR/ReportDownloads/202311/P020231121355042476714.pdf [accessed 2024-11-21] 9], the number of internet users aged 60 years and older reached a staggering 140 million as of June 2023. It signifies a profound penetration of internet use within the older adult demographic, bringing about substantial transformations in daily life. Preliminary analysis of the English Longitudinal Study of Aging indicated that individuals who frequently use the internet or email exhibit a 3.1% improvement in delayed recall compared with those who do not use the internet or email [Xavier AJ, d'Orsi E, de Oliveira CM, Orrell M, Demakakos P, Biddulph JP, et al. English Longitudinal Study of Aging: can internet/e-mail use reduce cognitive decline? J Gerontol A Biol Sci Med Sci. Sep 12, 2014;69(9):1117-1121. [CrossRef] [Medline]10]. An Australian study revealed that men aged 70 years and older who use the internet had a 50% relative risk reduction for dementia, even after controlling for age, education, social relationships, and physical health status [Almeida OP, Yeap BB, Alfonso H, Hankey GJ, Flicker L, Norman PE. Older men who use computers have lower risk of dementia. PLoS One. Aug 28, 2012;7(8):e44239. [FREE Full text] [CrossRef] [Medline]11]. The study by Klimova [Klimova B. Use of the Internet as a prevention tool against cognitive decline in normal aging. Clin Interv Aging. Sep 2016;11:1231-1237. [FREE Full text] [CrossRef] [Medline]12] suggested that the use of the internet, specifically online cognitive training programs, may exert a beneficial effect on enhancing the cognitive function of healthy older adults.
The mechanisms through which internet use enhances cognitive abilities are highly complex, and current research has not fully elucidated them yet. A cross-sectional study showed that internet searches can enhance the activation of neural circuits in the frontal pole, anterior temporal region, anterior and posterior cingulate, and hippocampus areas among people older than 55 years [Small GW, Moody TD, Siddarth P, Bookheimer SY. Your brain on Google: patterns of cerebral activation during internet searching. The American Journal of Geriatric Psychiatry. Feb 2009;17(2):116-126. [CrossRef]13]. Numerous previous studies have indicated that the internet can provide people with more social contacts, social support, and positive attitudes toward life, enhancing social connections while reducing loneliness and social isolation, which can have a positive impact on cognitive function [Cresci MK, Jarosz PA, Templin TN. Are health answers online for older adults? Educational Gerontology. Jan 2012;38(1):10-19. [CrossRef]14-Ashby FG, Isen AM, Turken AU. A neuropsychological theory of positive affect and its influence on cognition. Psychol Rev. Jul 1999;106(3):529-550. [CrossRef] [Medline]16]. The study by Firth et al [Firth J, Torous J, Stubbs B, Firth JA, Steiner GZ, Smith L, et al. The "online brain": how the Internet may be changing our cognition. World Psychiatry. Jun 06, 2019;18(2):119-129. [FREE Full text] [CrossRef] [Medline]17] suggested that the use of the internet may influence cognitive function through 3 key factors: attentional capacity, memory processes, and social cognition.
However, not all research findings consistently demonstrate the cognitive benefits of internet use. A randomized controlled trial conducted in the Netherlands with a small sample of individuals aged 64 years to 75 years revealed that 12 months of computer and internet use did not have a significant impact on the cognitive function of older adults [Slegers K, van Boxtel M, Jolles J. Effects of computer training and internet usage on cognitive abilities in older adults: a randomized controlled study. Aging Clin Exp Res. Jul 26, 2013;21(1):43-54. [CrossRef]18]. Overall, despite the common perception that internet use can improve cognitive function in older adults, the current literature provides varying evidence, necessitating studies with larger sample sizes to elucidate the correlation.
The China Health and Retirement Longitudinal Study (CHARLS) is the first nationally representative survey of middle-aged and older adults in China, with a sample collected from 28 provinces and 150 regions across the country [Zhao Y, Hu Y, Smith JP, Strauss J, Yang G. Cohort profile: the China Health and Retirement Longitudinal Study (CHARLS). Int J Epidemiol. Feb 12, 2014;43(1):61-68. [CrossRef] [Medline]19]. This project aims to construct a high-quality public microdatabase collecting multidimensional information encompassing socioeconomic status and health conditions to meet the needs of gerontological research. Therefore, the CHARLS questionnaire includes several aspects such as demographic background, family structure, health status, health insurance, work, retirement, pensions, income, expenditures, assets, and house property. To ensure randomization and the representativeness of the sample, CHARLS adopts multistage stratified probability proportional to size sampling. Meanwhile, CHARLS conducts face-to-face interviews facilitated by well-trained surveyors and implements effective quality control through methods such as data verification, audio playback, and the use of a computer-assisted personal interview system.
Our primary aim was to assess the independent association between internet use and cognitive function in middle-aged and older adults by collecting survey data from the CHARLS study. To gain deeper insights into the cognitive effects of internet use, we also evaluated the age-related impacts on cognitive function in internet users and nonusers as one of our secondary objectives. Additionally, to explore the most effective means for enhancing cognitive function, we analyzed the correlations between cognitive function and both internet device use and frequency. We further intended to evaluate the mediating role of internet use on cognitive function for participant characteristics for which a significant association between internet use and cognitive function was observed in stratified analysis.
Methods
Participants
Participants aged 45 years and older were enrolled in the CHARLS study and included in subsequent secondary analyses, with their socioeconomic and health status being systematically recorded. The CHARLS survey was initiated in 2011 (Wave 1), followed by 4 consecutive waves of regular follow-up questionnaires conducted in 2013 (Wave 2), 2015 (Wave 3), 2018 (Wave 4), and 2020 (Wave 5). More comprehensive information on the study design and sampling strategies used in CHARLS has been previously described [Zhao Y, Hu Y, Smith JP, Strauss J, Yang G. Cohort profile: the China Health and Retirement Longitudinal Study (CHARLS). Int J Epidemiol. Feb 12, 2014;43(1):61-68. [CrossRef] [Medline]19]. Given the insufficient number of internet users in Waves 1 and 2, we conducted data collection for the subsequent waves, including Wave 3, Wave 4, and Wave 5.
Measurements
Internet Use
The primary exposure and mediating variable was internet use, which was assessed based on a binary response (yes/no) to the question regarding internet access. With regard to the frequency of internet use, participants’ questionnaire responses were classified into 4 categories: never, infrequently, weekly, and daily. Additionally, the devices used for internet access were categorized into 2 groups: computers (including desktop computers, laptop computers, and tablet computers) and cell phones. The specific wording of the questionnaire survey on internet use is detailed in The questionnaire section on internet use and cognitive function in China Health and Retirement Longitudinal Study (CHARLS).Multimedia Appendix 1
Cognitive Function
We used 2 established composite measures, mental intactness and episodic memory, to evaluate global cognitive function. These measures are similar to the assessment indicators used by the Health and Retirement Study in the United States [Manly JJ, Jones RN, Langa KM, Ryan LH, Levine DA, McCammon R, et al. Estimating the prevalence of dementia and mild cognitive impairment in the US: the 2016 Health and Retirement Study Harmonized Cognitive Assessment protocol project. JAMA Neurol. Dec 01, 2022;79(12):1242-1249. [FREE Full text] [CrossRef] [Medline]20] and have been validated among the Chinese population [Meng Q, Wang H, Strauss J, Langa KM, Chen X, Wang M, et al. Validation of neuropsychological tests for the China Health and Retirement Longitudinal Study Harmonized Cognitive Assessment Protocol. Int Psychogeriatr. Dec 2019;31(12):1709-1719. [FREE Full text] [CrossRef] [Medline]21]. Mental intactness is determined based on 3 indicators of cognitive status, which comprise time orientation (naming month, day, year, week, and season), numerical ability (subtracting 7 consecutively from 100, 5 times), and visual and spatial abilities (the task of redrawing 2 overlapping pentagons). The mental intactness score for the 3 elements ranges from 0 to 11. Episodic memory can be evaluated through the recall of commonly used phrases, including immediate word memory (score range 0-10) and delayed word memory (score range 0-10, with a total score of 20). Cognitive function is constructed from these 2 components ( The questionnaire section on internet use and cognitive function in China Health and Retirement Longitudinal Study (CHARLS).Multimedia Appendix 1
Covariates
This study accounted for multiple potential confounding factors that are known risk factors for cognitive function [Kivipelto M, Mangialasche F, Ngandu T. Lifestyle interventions to prevent cognitive impairment, dementia and Alzheimer disease. Nat Rev Neurol. Nov 5, 2018;14(11):653-666. [CrossRef] [Medline]22], in order to provide more compelling evidence. In our study, we took into account the following covariates: baseline or prior cognitive score (continuous variable), gender (male or female), age (continuous variable), marital status (yes [married or partnered] or no [unmarried, separated, divorced, widowed]), residency (urban or rural), retirement status (yes or no), education level (illiterate, low education level [primary or junior high school], mid education level [senior high school], or high education level [college or above]), smoking status (current smoker, former smoker, never smoked), alcohol consumption (never drank, drink less than once a month, drink more than once a month), diabetes (yes or no), hypertension (yes or no), and per capita household expenditure (quartiles).
Statistical Analysis
To comprehensively assess the relationship between internet use and cognitive function, we used both longitudinal and cross-sectional analytical approaches, thereby aiming to complement the limitations inherent in each research method. We used fixed effects analysis to analyze longitudinal data with time-varying variables and performed multiple linear regression analyses on cross-sectional data from each wave. Furthermore, we used the restricted cubic spline regression method to investigate the differences in cognitive performance between internet users and nonusers across various ages. We used the Spearman rank correlation test to analyze the correlation between internet use frequency and cognitive function. We regarded baseline characteristics as covariates and conducted stratified and interaction analyses to assess the potential differences in the impact of internet use among different subgroups. After stratifying by baseline characteristics, we constructed a fixed effects model and incorporated the cross-product terms (internet use × baseline characteristics) within the same model to evaluate the interaction between internet use and baseline characteristics.
The mediation analysis was performed using the Karlson-Holm-Breen (KHB) method [Karlson KB, Holm A, Breen R. Comparing regression coefficients between same-sample nested models using logit and probit. Sociological Methodology. Nov 21, 2012;42(1):286-313. [CrossRef]23], which dissects the influence of urban-rural areas on cognitive functionality into its distinct direct and indirect constituents. In particular, it measures the extent to which the intermediate variable (internet use) mediates the association between the independent variable (urban-rural status) and the dependent variable (cognitive function).
In sensitivity analysis, we conducted propensity score matching (PSM) [Rosenbaum PR, Rubin DB. The central role of the propensity score in observational studies for causal effects. Biometrika. Apr 1983;70(1):41. [CrossRef]24] with a 1:1 ratio to adjust for covariates and analyze the impact of internet use on cognitive function. Second, we also used inverse probability of treatment weighting (IPTW) [Robins J. A new approach to causal inference in mortality studies with a sustained exposure period—application to control of the healthy worker survivor effect. Mathematical Modelling. Jan 1986;7(9-12):1393-1512. [FREE Full text] [CrossRef]25] and overlap weighting (OW) [Thomas LE, Li F, Pencina MJ. Overlap weighting: a propensity score method that mimics attributes of a randomized clinical trial. JAMA. Jun 16, 2020;323(23):2417-2418. [CrossRef] [Medline]26] methods to weight the balance based on covariates, aiming to validate our main results. To obtain robust results and further enhance the sensitivity analysis, we additionally selected individuals who had been continuous internet users and continuous nonusers during a 5-year follow-up period to investigate their impact on cognitive function using generalized estimating equations (GEE) and multiple linear regression analysis. Continuous variables are shown as means and SDs, while categorical variables are represented as numbers (proportions). All analyses were performed using R (version 4.0.2) and Stata (version 14.1).
Ethical Considerations
The Biomedical Ethics Review Committee of Peking University approved the ethical application (IRB00001052-11014 and IRB00001052-11015) for collecting data from human participants in all waves, as well as for secondary analysis of CHARLS data. Written informed consent was obtained from all participants prior to the questionnaire survey, and the original informed consent allows secondary analysis without additional consent. All respondents who completed the questionnaire received cash as compensation according to the unified standard amount set by the CHARLS in advance. In order to protect the privacy of the respondents, all the data in this study are anonymous.
Results
Descriptive Statistics
A total of 12,770 Chinese adults aged 45 years and older were ultimately included in this study, with the process of selection and exclusion detailed in Figure 1. In addition, Table 1 presents the baseline (ie, Wave 3) demographic characteristics of the study participants. At baseline, internet users accounted for 7.9% (1005/12,770) of the participants, and nonusers accounted for 92.1% (11,765/12,770) of the participants. Participants engaging in internet use had a significantly lower mean age of 53.04 (SD 7.33) years in contrast to the 59.32 (SD 9.16) years among nonusers (P<.001). More men were internet users (583/1005, 58%) than nonusers (5507/11,765, 46.8%). In terms of educational attainment, the internet-using population had a markedly lower illiteracy rate of 0.6% (6/1005) than the nonuser population (2269/11,765, 19.3%). This trend was consistent across educational strata, with 11.2% (113/1005), 71.9% (722/1005), and 16.3% (164/1005) of internet users classified as having low, mid, and high levels of education, respectively, as opposed to 44.1% (5189/11,765), 35.3% (4159/11,765), and 1.3% (148/11,765), respectively, of the nonuser population. Marital status among internet users was predominantly characterized by marriage, with 93.9% (944/1005) of this group being married, which was a higher proportion than the 90.3% (10,623/11,765) of nonusers who were married (P<.001). The proportion of retired individuals was also higher within the internet-using group, at 21% (211/1005) compared with 9.1% (1072/11,765) among nonusers (P<.001). Residence in rural areas was less common among internet users, with 46.9% (471/1005) living in such regions, a significantly lower proportion than the 83.5% (9825/11,765) of nonusers (P<.001). Notably, in terms of cognitive function, the cognitive assessment scores of internet users across different age groups (45-54 years, 55-64 years, 65-74 years, and 75 years) were significantly higher than those of nonusers (P<.001).
Characteristic | Nonusers (n=11,765) | Internet users (n=1005) | P value | ||
Gender, n (%) | <.001 | ||||
Male | 5507 (46.8) | 583 (58) | |||
Female | 6258 (53.2) | 422 (42) | |||
Age (years), mean (SD) | 59.32 (9.16) | 53.04 (7.33) | <.001 | ||
Age group (years), n (%) | <.001 | ||||
45-54 | 4384 (37.3) | 692 (68.8) | |||
55-64 | 3917 (33.3) | 219 (21.8) | |||
65-74 | 2754 (23.4) | 84 (8.4) | |||
≥75 | 710 (6) | 10 (1) | |||
Household income per capitaa, n (%) | <.001 | ||||
Quartile 1 | 2779 (23.6) | 137 (13.6) | |||
Quartile 2 | 2877 (24.4) | 146 (14.5) | |||
Quartile 3 | 3091 (26.3) | 250 (24.9) | |||
Quartile 4 | 3018 (25.7) | 472 (47) | |||
Maritalstatus, n (%) | <.001 | ||||
Yes | 10,623 (90.3) | 944 (93.9) | |||
No | 1142 (9.7) | 61 (6.1) | |||
Residency, n (%) | <.001 | ||||
Rural | 9825 (83.5) | 471 (46.9) | |||
Urban | 1940 (16.5) | 534 (53.1) | |||
Educational attainment, n (%) | .001 | ||||
Illiterate | 2269 (19.3) | 6 (0.6) | |||
Low education | 5189 (44.1) | 113 (11.2) | |||
Mid education | 4159 (35.3) | 722 (71.9) | |||
High education | 148 (1.3) | 164 (16.3) | |||
Retirement status, n (%) | <.001 | ||||
No | 10,693 (90.9) | 794 (79) | |||
Yes | 1072 (9.1) | 211 (21) | |||
Smoking status, n (%) | <.001 | ||||
Never smoked | 4533 (38.5) | 294 (29.3) | |||
Quit | 2234 (19) | 195 (19.4) | |||
Still smoke | 4998 (42.5) | 516 (51.3) | |||
Drinking status, n (%) | <.001 | ||||
Never drank | 7579 (64.4) | 432 (43) | |||
Drink less than once a month | 996 (8.5) | 195 (19.4) | |||
Drink more than once a month | 3190 (27.1) | 378 (37.6) | |||
Hypertension, n (%) | .045 | ||||
No | 8353 (71) | 744 (74) | |||
Yes | 3412 (29) | 261 (26) | |||
Dyslipidemia, n (%) | .12 | ||||
No | 8526 (72.5) | 752 (74.8) | |||
Yes | 3239 (27.5) | 253 (25.2) | |||
Diabetes, n (%) | .004 | ||||
No | 8387 (71.3) | 760 (75.6) | |||
Yes | 3378 (28.7) | 245 (24.4) | |||
Cognitive assessment score by age (years), mean (SD) | |||||
45-54 | 16.00 (4.85) | 19.99 (3.61) | <.001 | ||
55-64 | 14.50 (4.98) | 19.69 (3.30) | <.001 | ||
65-74 | 13.49 (5.20) | 19.24 (3.67) | <.001 | ||
≥75 | 11.46 (5.25) | 18.00 (4.71) | <.001 |
aDivided into 4 quartiles based on the 25th percentile (¥1643.60 [US $226.89]), 50th percentile (¥4497.50 [US $620.86]), and 75th percentile (¥10,959.80 [US $1512.95]) of the overall data.
Association Between Internet Use and Cognitive Function
This study demonstrated a positive association between internet use and cognitive function among middle-aged and older adults (Figure 2) based on the multiple linear regression analysis to evaluate cross-sectional data from Wave 3, Wave 4, and Wave 5 and the fixed effects model to analyze the longitudinal data across the entire population. Model 1 was adjusted for baseline or prior cognitive score; Model 2 was additionally adjusted for age, gender, marital status, educational attainment, residency, and retirement status; Model 3 was additionally adjusted for smoking, drinking, hypertension, and diabetes; and Model 4 was additionally adjusted for household income per capita. There were significant differences between internet users and nonusers in both mental intactness score and episodic memory score (P<.001; Cognitive scores of internet users and non-users at different waves.Multimedia Appendix 2
To further investigate the association between internet use and specific clinical manifestations of cognitive function, we analyzed the incidence of neurodegenerative diseases (Alzheimer disease or Parkinson disease) among internet users and nonusers ( The prevalence and incidence of neurodegenerative diseases (Alzheimer disease or Parkinson disease) in internet users and non-users during the 5-year period. The prevalence and incidence of neurodegenerative diseases (Alzheimer disease or Parkinson disease) in internet users and non-users during the 5-year period.Multimedia Appendix 3
Multimedia Appendix 3
Impact of Internet Use on Cognitive Function Across Age
In order to further investigate the effects of internet use on cognitive function across different ages, we used restricted cubic splines to fit the curves that influence cognitive function (Figure 3). Our results indicated that age becomes a risk factor for cognitive function in internet nonusers after the age of 50.02 years, whereas for internet users, it occurs after the age of 49.02 years. Interestingly, we observed that the impact of age on cognitive decline was consistently minor among internet users compared with among nonusers after the age of 50 years. In other words, internet use has the potential to slow age-related cognitive decline. The results also showed that the older the age, the larger the gap in the effect of age on cognitive function between internet users and nonusers. This implies that the protective effect of internet use on cognitive decline is increasingly evident with age.
Impact of Internet Device and Frequency on Cognitive Function
The fixed effects model (Figure 4A) suggests that, among middle-aged and older adults, internet users who use cell phones experience cognitive benefits (model 1: β=0.406, 95% CI 0.300 to 0.512, P<.001; model 2: β=0.412, 95% CI 0.306 to 0.519, P<.001; model 3: β=0.392, 95% CI 0.287 to 0.499, P<.001; model 4: β=0.398, 95% CI 0.283 to 0.495, P<.001). Furthermore, internet use on computer devices also has a positive influence on cognitive function (model 1: β=0.153, 95% CI 0.097 to 0.210, P<.001; model 2: β=0.158, 95% CI 0.101 to 0.214, P<.001; model 3: β=0.149, 95% CI 0.092 to 0.205, P<.001; model 4: β=0.147, 95% CI 0.091 to 0.204, P<.001), though to a lesser extent than cell phone–based internet use. This indicates that using cell phones as internet devices results in greater cognitive benefits than using computers. The proportion of mobile phone users among internet users increased dramatically over the past 5 years, from 64.5% (648/1005) to 99.3% (5463/5503), as detailed in Changes in the proportion of internet users with different internet access devices and frequencies. Changes in the proportion of internet users with different internet access devices and frequencies.Multimedia Appendix 4
Multimedia Appendix 4
Stratified and Interaction Analyses
Stratified analysis revealed that the positive correlation between internet use and cognitive function was similar across subgroups differentiated by age, gender, residency, and retirement status (Figure 5). In addition, internet use significantly affected individuals who were married (P<.001), while among those who were unmarried, there was only a trend present without a significant difference (P=.18). Interestingly, among middle-aged and older adults, there seemed to be a positive association between lower levels of educational attainment and higher levels of cognitive benefits gained from internet use (illiterate: β=0.807, 95% CI 0.209 to 1.407, P=.008; low education level: β=0.545, 95% CI 0.280 to 0.810, P<.001; mid education level: β=0.389, 95% CI 0.160 to 0.617, P=.001; high education level: β=0.160, 95% CI –0.215 to 0.537, P=.40). As the majority of middle-aged and older adults in China currently have low levels of education [Fang Y, Wang H, Ouyang Z. An overview report on the situation of elderly people in urban and rural China: 2000–2015. China popul. dev. stud. Oct 9, 2018;2(3):323-345. [CrossRef]27], there may be tremendous potential for cognitive benefits from internet use.
Mediation Analyses
Different areas of residence (urban or rural) among middle-aged and older adults resulted in notable differences in cognitive function (P<.001; Cognitive scores of rural-urban participants at different waves.Multimedia Appendix 5
Sensitivity Analyses
Sensitivity analyses proved that the main findings of our study were robust. Following the implementation of PSM to achieve a balanced distribution of covariates (standardized mean difference<0.1) between internet users and nonusers, we obtained similar results in line with the fixed effects model (Figure 7). Moreover, we used IPTW to adjust for covariates, and the results were broadly consistent with the main analysis ( The relationship between internet use and cognitive function in middle-aged and older adults was analyzed using inverse probability of treatment weighting (IPTW) and overlap weighting (OW) methods. The relationship between internet use and cognitive function in middle-aged and older adults was analyzed using inverse probability of treatment weighting (IPTW) and overlap weighting (OW) methods. Differences in cognitive function between persistent internet users (n=671) and persistent non-users (n=7099) using generalized estimating equations (GEE) and multiple linear regression models.Multimedia Appendix 6
Multimedia Appendix 6
Multimedia Appendix 7
Discussion
Principal Findings
In this nationwide cohort study focusing on individuals aged 45 years and older, we found a positive relationship between internet use and cognitive function. Moreover, a higher frequency of internet use was associated with a lower risk of cognitive decline. Previous studies have suggested that using the internet can reduce social isolation, increase social interactions (both face to face and virtual or online), stimulate the acquisition of new knowledge, enhance cognitive demand for mastering novel skills using digital tools, and facilitate access to health and cultural information [d'Orsi E, Xavier AJ, Rafnsson SB, Steptoe A, Hogervorst E, Orrell M. Is use of the internet in midlife associated with lower dementia incidence? Results from the English Longitudinal Study of Ageing. Aging Ment Health. Nov 10, 2018;22(11):1525-1533. [FREE Full text] [CrossRef] [Medline]28]. Individuals who use the internet may improve attention and psychomotor skills and enhance cognitive reserve through stimulation and reduction of depressive symptoms [Shapira N, Barak A, Gal I. Promoting older adults' well-being through Internet training and use. Aging Ment Health. Sep 2007;11(5):477-484. [CrossRef] [Medline]29]. For middle-aged and older adults who use the internet, going online may help their brains use more efficient cognitive networks to counteract dementia [La Rue A. Healthy brain aging: role of cognitive reserve, cognitive stimulation, and cognitive exercises. Clin Geriatr Med. Feb 2010;26(1):99-111. [CrossRef] [Medline]30]. Another explanation is that internet use is not only a form of cognitive stimulation but also a mode of social engagement. The sense of belonging created by online activities and the formation of social networks facilitate the brain's evolution and functional development [Fratiglioni L, Paillard-Borg S, Winblad B. An active and socially integrated lifestyle in late life might protect against dementia. The Lancet Neurology. Jun 2004;3(6):343-353. [CrossRef]31]. These mechanisms partially account for the beneficial cognitive effects of internet use observed in our study.
Our study further explores the trends in the impact of aging on cognitive function among internet users and nonusers. In contrast to internet nonusers, we found that internet users showed a delay in age-related cognitive decline. Although we cannot ascertain a causal relationship from this observational study, we have reasons to speculate that internet use, as a cognitive-stimulating activity, may lead to beneficial physiological changes in the brain. With the progression of age, cognitive abilities such as attention, memory, executive function, language, and visuospatial skills prominently decline [Murman D. The impact of age on cognition. Semin Hear. Aug 9, 2015;36(3):111-121. [CrossRef] [Medline]32]. Additionally, the physiological changes in brain structure that accompany aging encompass alterations in neuronal structure, synaptic loss, and dysfunction of neuronal networks [Masliah E, Mallory M, Hansen L, DeTeresa R, Terry RD. Quantitative synaptic alterations in the human neocortex during normal aging. Neurology. Jan 1993;43(1):192-197. [CrossRef] [Medline]33-Morrison JH, Baxter MG. The ageing cortical synapse: hallmarks and implications for cognitive decline. Nat Rev Neurosci. Mar 07, 2012;13(4):240-250. [CrossRef] [Medline]35]. Research on brain plasticity provides strong support for the social modulation of cognition, as activity-related stimuli can induce angiogenesis, synaptogenesis, and neurogenesis in the brain [Churchill JD, Galvez R, Colcombe S, Swain RA, Kramer AF, Greenough WT. Exercise, experience and the aging brain. Neurobiol Aging. Sep 2002;23(5):941-955. [CrossRef] [Medline]36,Fillit HM, Butler RN, O'Connell AW, Albert MS, Birren JE, Cotman CW, et al. Achieving and maintaining cognitive vitality with aging. Mayo Clin Proc. Jul 2002;77(7):681-696. [CrossRef] [Medline]37]. Previous studies have not fully elucidated the cognitive benefits of internet use. Future studies could incorporate more detailed assessments of cognitive domains, neuroimaging data, or biomarkers to better understand the pathways through which internet use may influence cognitive function.
Our findings indicate that both computer-based and mobile-based internet use provide cognitive-enhancing effects, which is consistent with previous studies [Jin Y, Jing M, Ma X. Effects of digital device ownership on cognitive decline in a middle-aged and elderly population: longitudinal observational study. J Med Internet Res. Jul 29, 2019;21(7):e14210. [FREE Full text] [CrossRef] [Medline]38-Lin L, Jing X, Lv S, Liang J, Tian L, Li H, et al. Mobile device use and the cognitive function and depressive symptoms of older adults living in residential care homes. BMC Geriatr. Feb 03, 2020;20(1):41. [FREE Full text] [CrossRef] [Medline]40]. The research findings by Jin et al [Jin Y, Jing M, Ma X. Effects of digital device ownership on cognitive decline in a middle-aged and elderly population: longitudinal observational study. J Med Internet Res. Jul 29, 2019;21(7):e14210. [FREE Full text] [CrossRef] [Medline]38] revealed that middle-aged and older adults who own a cell phone or desktop computer at home experience a delay in cognitive decline. However, this study was based on the assessment of cell phone and desktop computer ownership, which is not as straightforward as our study in which we inquired about the specific devices used for internet access. Interestingly, our research findings suggest that middle-aged and older adults who use cell phones to access the internet have higher cognitive scores than those who use computers. This outcome might be explained by the fact that mobile phones have more advantages regarding operation and accessibility and are more user-friendly for older adults, so mobile phone users used the internet at a higher frequency and for longer durations. With the rapid development of social information technology, some mobile apps (such as WeChat and TikTok) have become increasingly vital to the daily lives of middle-aged and older adults, as they not only facilitate social communications and connections but also provide leisure and entertainment [Wang G, Zhang W, Zeng R. WeChat use intensity and social support: The moderating effect of motivators for WeChat use. Computers in Human Behavior. Feb 2019;91:244-251. [FREE Full text] [CrossRef]41]. Meanwhile, these apps have lower operational barriers, making them more user-friendly. A previous study [Joe J, Demiris G. Older adults and mobile phones for health: a review. J Biomed Inform. Oct 2013;46(5):947-954. [FREE Full text] [CrossRef] [Medline]42] showed that, as of 2012, older adults aged 65 years and older were more likely to own a mobile phone (69%) rather than a desktop computer (48%) or a laptop (32%). The study conducted by Liu et al [Liu X, Wang X, Zhang H, Pei M, Li N. Relationship between digital exclusion and cognitive impairment in Chinese adults. Front Aging Neurosci. Jul 3, 2023;15:1194348. [FREE Full text] [CrossRef] [Medline]43] also suggested that the absence of using mobile social media platforms, such as WeChat and mobile payment systems, may be positively associated with cognitive impairments. Therefore, the impact of cell phone use on age-related cognitive decline among middle-aged and older adults deserves more attention.
The cognitive benefits of internet use among middle-aged and older adults with different educational levels seem to vary. The results of the stratified analysis suggest that internet use has a greater extent of cognitive benefits for those with lower educational attainment. Based on previous research, there continues to be controversy regarding the impact of education on age-related cognitive decline. Some research reports that education may slow the rate of age-related cognitive decline [Bennett DA, Wilson RS, Schneider JA, Evans DA, Mendes de Leon CF, Arnold SE, et al. Education modifies the relation of AD pathology to level of cognitive function in older persons. Neurology. Jun 24, 2003;60(12):1909-1915. [CrossRef] [Medline]44-Lyketsos CG, Chen L, Anthony JC. Cognitive decline in adulthood: an 11.5-year follow-up of the Baltimore Epidemiologic Catchment Area study. Am J Psychiatry. Jan 1999;156(1):58-65. [CrossRef] [Medline]46], while others [Carmelli D, Swan GE, LaRue A, Eslinger PJ. Correlates of change in cognitive function in survivors from the Western Collaborative Group Study. Neuroepidemiology. Dec 3, 1997;16(6):285-295. [CrossRef] [Medline]47,Hultsch DF, Hertzog C, Dixon RA, Small BJ. Memory change in the aged. New York, NY. Cambridge University Press; 1998. 48] suggest that education has no significant effect on the rate of cognitive decline. Moreover, there have been studies indicating a relationship between higher levels of education and an accelerated cognitive decline, particularly in the domains of verbal memory, processing speed, and verbal fluency [Alley D, Suthers K, Crimmins E. Education and cognitive decline in older Americans: results from the AHEAD sample. Res Aging. Jan 01, 2007;29(1):73-94. [CrossRef] [Medline]49,Proust-Lima C, Amieva H, Letenneur L, Orgogozo J, Jacqmin-Gadda H, Dartigues J. Gender and education impact on brain aging: a general cognitive factor approach. Psychol Aging. Sep 2008;23(3):608-620. [CrossRef] [Medline]50]. Although the impact of education on cognitive decline remains uncertain, internet use still holds enormous potential for improving cognition in individuals with low educational levels, particularly in China where 71.1% of older adults either have no formal education or have only received primary schooling [Fang Y, Wang H, Ouyang Z. An overview report on the situation of elderly people in urban and rural China: 2000–2015. China popul. dev. stud. Oct 9, 2018;2(3):323-345. [CrossRef]27]. On the other hand, although middle-aged and older adults with lower educational attainment receive great cognitive benefits from surfing the internet, they may face greater potential risks or unintended consequences (such as exposure to misinformation, online scams, and privacy concerns). Hence, ensuring internet safety for middle-aged and older adults, especially those with lower levels of education, constitutes a major challenge for developing a digital society in the future.
The Spearman rank correlation results in this study indicate that, among middle-aged and older adults, higher frequencies of internet use are associated with better cognitive performance. Although higher use frequency means better cognitive benefits, the impact of overuse of the internet cannot be ignored. Since specific durations of internet use for each individual could not be observed in this study, analyzing the cognitive impacts resulting from internet overuse was not feasible. More research is needed to understand the relationship between cognitive function and both the frequency and duration of internet use, especially in the case of overuse.
The phenomenon of urban-rural disparity in cognitive performance among middle-aged and older adults is prevalent in both lower-income and higher-income countries [Glauber R. Rural depopulation and the rural-urban gap in cognitive functioning among older adults. J Rural Health. Sep 07, 2022;38(4):696-704. [CrossRef] [Medline]51,Zhang T, Lu B, Wang X. Urban-rural disparity in cognitive performance among older Chinese adults: explaining the changes from 2008 to 2018. Front Public Health. Mar 23, 2022;10:843608. [FREE Full text] [CrossRef] [Medline]52]. Previous research has indicated that cognitive functional disparities between urban and rural areas may be caused by various factors, including health care service resources, public education resources, early-life cognitive reserves, race, social class, social activities, and social engagement [Wight R, Aneshensel C, Miller-Martinez D, Botticello A, Cummings J, Karlamangla A, et al. Urban neighborhood context, educational attainment, and cognitive function among older adults. Am J Epidemiol. Jun 15, 2006;163(12):1071-1078. [CrossRef] [Medline]53,Liu D, Li L, An L, Cheng G, Chen C, Zou M, et al. Urban-rural disparities in mild cognitive impairment and its functional subtypes among community-dwelling older residents in central China. Gen Psychiatr. Oct 27, 2021;34(5):e100564. [FREE Full text] [CrossRef] [Medline]54]. Our research revealed that internet use, as a tool for social participation, exerts a noteworthy mediating role in the cognitive function disparities between urban and rural areas. This indicates that promoting the use of the internet in rural regions could help to reduce the urban-rural gap in cognitive performance.
Limitations
This study is not without limitations, and caution should be exercised when interpreting the findings. First, the study participants may not fully represent the middle-aged and older adult population in China, as the questionnaire survey was conducted only within the community and did not include public institutions such as nursing homes and hospitals. Therefore, the generalizability of our study findings primarily applies to healthy middle-aged and older adults, rather than frail older adults. Second, our analysis was based on a follow-up interval of approximately 5 years, and we cannot ascertain whether this correlation will endure over a more extended period. Third, the age of data and the changes in internet use over the past decade are factors that must be considered when interpreting the results of this study, as the pace of internet development and the expansion of internet access among older adults over the past 10 years may differ from what is expected in the next decade. Fourth, our approach to assessing cognitive function was not entirely based on specialized screening scales for cognitive impairment, such as the Mini-Mental State Examination or the Montreal Cognitive Assessment Scale. Therefore, caution must be exercised when interpreting the effects of internet use on mild cognitive impairment or Alzheimer disease. Fifth, due to the presence of loss to follow-up and unmeasured confounders, there may be bias in the association estimation. However, to ensure the robustness of the research data, we used various sensitivity analyses, including multiple linear regression analysis, GEE, PSM, IPTW, and OW, to validate the cognitive benefits of internet use. Sixth, our study did not directly establish a causal relationship between internet use and cognitive decline. However, the lower incidence of neurodegenerative diseases in persistent internet users indicated that internet use may prevent neurodegenerative diseases. Therefore, in order to prove the causal relationship between internet use and cognitive function, more research needs to be carried out, such as physiological mechanism research, instrumental variable analysis, and a Mendelian randomization study. Last, although the positive association between internet use and cognitive function was statistically significant after considering various potential confounders, there may still be some relevant variables, such as the apolipoprotein E4 genotype, that cannot be obtained from the survey questionnaire but could potentially influence the results.
Conclusions
In conclusion, this study presents valuable evidence supporting a longitudinal protective association between internet use and cognitive function in middle-aged and older adults, with a higher frequency of internet use having a greater extent of cognitive benefit. As digital devices for internet access, cell phones may offer a potentially superior level of cognitive protection than computers. The cognitive gap among middle-aged and older adults in urban and rural areas can be partially attributed to internet use. This study suggests that internet use may be a promising cognitive stimulation activity that mitigates age-related cognitive decline and prevents neurodegenerative diseases. Although further research is needed to confirm the causal relationship, the findings highlight the potential benefits of promoting internet use among middle-aged and older adults, especially in rural areas, as a strategy to support cognitive health and potentially reduce the risk of dementia.
Acknowledgments
The authors thank the China Health and Retirement Longitudinal Study (CHARLS) research team and all respondents for their contribution. This study was supported by grants from the National Natural Science Foundation of China (number 81703216).
Data Availability
The data sets used and analyzed during the study are available on the China Health and Retirement Longitudinal Study (CHARLS) website [2020 National Tracking Survey. CHARLS. URL: https://charls.charlsdata.com/pages/data/111/zh-cn.html [accessed 2024-11-21] 55].
Authors' Contributions
BC and JZ designed the research. BC conducted the statistical analysis and wrote the manuscript drafts. PL supervised the use of statistical methods, validated the data, and replicated the results. SR and CY conducted the data visualization. PL and JZ revised the manuscript. All authors read and approved the submitted version of the manuscript.
Conflicts of Interest
None declared.
Multimedia Appendix 1
The questionnaire section on internet use and cognitive function in China Health and Retirement Longitudinal Study (CHARLS).
DOCX File , 16 KBMultimedia Appendix 2
Cognitive scores of internet users and non-users at different waves.
DOCX File , 13 KBMultimedia Appendix 3
The prevalence and incidence of neurodegenerative diseases (Alzheimer disease or Parkinson disease) in internet users and non-users during the 5-year period.
DOCX File , 17 KBMultimedia Appendix 4
Changes in the proportion of internet users with different internet access devices and frequencies.
DOCX File , 50 KBMultimedia Appendix 5
Cognitive scores of rural-urban participants at different waves.
DOCX File , 13 KBMultimedia Appendix 6
The relationship between internet use and cognitive function in middle-aged and older adults was analyzed using inverse probability of treatment weighting (IPTW) and overlap weighting (OW) methods.
DOCX File , 13 KBMultimedia Appendix 7
Differences in cognitive function between persistent internet users (n=671) and persistent non-users (n=7099) using generalized estimating equations (GEE) and multiple linear regression models.
DOCX File , 14 KBReferences
- Pais R, Ruano L, P Carvalho O, Barros H. Global cognitive impairment prevalence and incidence in community dwelling older adults-a systematic review. Geriatrics (Basel). Oct 27, 2020;5(4):84. [FREE Full text] [CrossRef] [Medline]
- Prince MJ, Wu F, Guo Y, Gutierrez Robledo LM, O'Donnell M, Sullivan R, et al. The burden of disease in older people and implications for health policy and practice. Lancet. Feb 07, 2015;385(9967):549-562. [CrossRef] [Medline]
- Liu J, Wang L, Tan J. Dementia in China: Current status. Neurology. Sep 17, 2013;81(12):1077-1078. [CrossRef]
- Jia L, Du Y, Chu L, Zhang Z, Li F, Lyu D, et al. Prevalence, risk factors, and management of dementia and mild cognitive impairment in adults aged 60 years or older in China: a cross-sectional study. The Lancet Public Health. Dec 2020;5(12):e661-e671. [CrossRef]
- Frankish H, Horton R. Prevention and management of dementia: a priority for public health. The Lancet. Dec 2017;390(10113):2614-2615. [CrossRef]
- Satizabal CL, Beiser AS, Chouraki V, Chêne G, Dufouil C, Seshadri S. Incidence of dementia over three decades in the Framingham Heart Study. N Engl J Med. Feb 11, 2016;374(6):523-532. [CrossRef]
- Anstey KJ, Bahar-Fuchs A, Herath P, Kim S, Burns R, Rebok GW, et al. Body brain life: A randomized controlled trial of an online dementia risk reduction intervention in middle-aged adults at risk of Alzheimer's disease. Alzheimers Dement (N Y). Jun 10, 2015;1(1):72-80. [FREE Full text] [CrossRef] [Medline]
- García-Casal JA, Loizeau A, Csipke E, Franco-Martín M, Perea-Bartolomé MV, Orrell M. Computer-based cognitive interventions for people living with dementia: a systematic literature review and meta-analysis. Aging Ment Health. May 2017;21(5):454-467. [CrossRef] [Medline]
- The 52nd Statistical Report on China’s Internet Development. China Internet Network Information Center (CNNIC). Aug 2023. URL: https://www.cnnic.com.cn/IDR/ReportDownloads/202311/P020231121355042476714.pdf [accessed 2024-11-21]
- Xavier AJ, d'Orsi E, de Oliveira CM, Orrell M, Demakakos P, Biddulph JP, et al. English Longitudinal Study of Aging: can internet/e-mail use reduce cognitive decline? J Gerontol A Biol Sci Med Sci. Sep 12, 2014;69(9):1117-1121. [CrossRef] [Medline]
- Almeida OP, Yeap BB, Alfonso H, Hankey GJ, Flicker L, Norman PE. Older men who use computers have lower risk of dementia. PLoS One. Aug 28, 2012;7(8):e44239. [FREE Full text] [CrossRef] [Medline]
- Klimova B. Use of the Internet as a prevention tool against cognitive decline in normal aging. Clin Interv Aging. Sep 2016;11:1231-1237. [FREE Full text] [CrossRef] [Medline]
- Small GW, Moody TD, Siddarth P, Bookheimer SY. Your brain on Google: patterns of cerebral activation during internet searching. The American Journal of Geriatric Psychiatry. Feb 2009;17(2):116-126. [CrossRef]
- Cresci MK, Jarosz PA, Templin TN. Are health answers online for older adults? Educational Gerontology. Jan 2012;38(1):10-19. [CrossRef]
- Hakkarainen P. ‘No good for shovelling snow and carrying firewood’: Social representations of computers and the internet by elderly Finnish non-users. New Media & Society. May 07, 2012;14(7):1198-1215. [CrossRef]
- Ashby FG, Isen AM, Turken AU. A neuropsychological theory of positive affect and its influence on cognition. Psychol Rev. Jul 1999;106(3):529-550. [CrossRef] [Medline]
- Firth J, Torous J, Stubbs B, Firth JA, Steiner GZ, Smith L, et al. The "online brain": how the Internet may be changing our cognition. World Psychiatry. Jun 06, 2019;18(2):119-129. [FREE Full text] [CrossRef] [Medline]
- Slegers K, van Boxtel M, Jolles J. Effects of computer training and internet usage on cognitive abilities in older adults: a randomized controlled study. Aging Clin Exp Res. Jul 26, 2013;21(1):43-54. [CrossRef]
- Zhao Y, Hu Y, Smith JP, Strauss J, Yang G. Cohort profile: the China Health and Retirement Longitudinal Study (CHARLS). Int J Epidemiol. Feb 12, 2014;43(1):61-68. [CrossRef] [Medline]
- Manly JJ, Jones RN, Langa KM, Ryan LH, Levine DA, McCammon R, et al. Estimating the prevalence of dementia and mild cognitive impairment in the US: the 2016 Health and Retirement Study Harmonized Cognitive Assessment protocol project. JAMA Neurol. Dec 01, 2022;79(12):1242-1249. [FREE Full text] [CrossRef] [Medline]
- Meng Q, Wang H, Strauss J, Langa KM, Chen X, Wang M, et al. Validation of neuropsychological tests for the China Health and Retirement Longitudinal Study Harmonized Cognitive Assessment Protocol. Int Psychogeriatr. Dec 2019;31(12):1709-1719. [FREE Full text] [CrossRef] [Medline]
- Kivipelto M, Mangialasche F, Ngandu T. Lifestyle interventions to prevent cognitive impairment, dementia and Alzheimer disease. Nat Rev Neurol. Nov 5, 2018;14(11):653-666. [CrossRef] [Medline]
- Karlson KB, Holm A, Breen R. Comparing regression coefficients between same-sample nested models using logit and probit. Sociological Methodology. Nov 21, 2012;42(1):286-313. [CrossRef]
- Rosenbaum PR, Rubin DB. The central role of the propensity score in observational studies for causal effects. Biometrika. Apr 1983;70(1):41. [CrossRef]
- Robins J. A new approach to causal inference in mortality studies with a sustained exposure period—application to control of the healthy worker survivor effect. Mathematical Modelling. Jan 1986;7(9-12):1393-1512. [FREE Full text] [CrossRef]
- Thomas LE, Li F, Pencina MJ. Overlap weighting: a propensity score method that mimics attributes of a randomized clinical trial. JAMA. Jun 16, 2020;323(23):2417-2418. [CrossRef] [Medline]
- Fang Y, Wang H, Ouyang Z. An overview report on the situation of elderly people in urban and rural China: 2000–2015. China popul. dev. stud. Oct 9, 2018;2(3):323-345. [CrossRef]
- d'Orsi E, Xavier AJ, Rafnsson SB, Steptoe A, Hogervorst E, Orrell M. Is use of the internet in midlife associated with lower dementia incidence? Results from the English Longitudinal Study of Ageing. Aging Ment Health. Nov 10, 2018;22(11):1525-1533. [FREE Full text] [CrossRef] [Medline]
- Shapira N, Barak A, Gal I. Promoting older adults' well-being through Internet training and use. Aging Ment Health. Sep 2007;11(5):477-484. [CrossRef] [Medline]
- La Rue A. Healthy brain aging: role of cognitive reserve, cognitive stimulation, and cognitive exercises. Clin Geriatr Med. Feb 2010;26(1):99-111. [CrossRef] [Medline]
- Fratiglioni L, Paillard-Borg S, Winblad B. An active and socially integrated lifestyle in late life might protect against dementia. The Lancet Neurology. Jun 2004;3(6):343-353. [CrossRef]
- Murman D. The impact of age on cognition. Semin Hear. Aug 9, 2015;36(3):111-121. [CrossRef] [Medline]
- Masliah E, Mallory M, Hansen L, DeTeresa R, Terry RD. Quantitative synaptic alterations in the human neocortex during normal aging. Neurology. Jan 1993;43(1):192-197. [CrossRef] [Medline]
- Dennis EL, Thompson PM. Functional brain connectivity using fMRI in aging and Alzheimer's disease. Neuropsychol Rev. Mar 23, 2014;24(1):49-62. [CrossRef] [Medline]
- Morrison JH, Baxter MG. The ageing cortical synapse: hallmarks and implications for cognitive decline. Nat Rev Neurosci. Mar 07, 2012;13(4):240-250. [CrossRef] [Medline]
- Churchill JD, Galvez R, Colcombe S, Swain RA, Kramer AF, Greenough WT. Exercise, experience and the aging brain. Neurobiol Aging. Sep 2002;23(5):941-955. [CrossRef] [Medline]
- Fillit HM, Butler RN, O'Connell AW, Albert MS, Birren JE, Cotman CW, et al. Achieving and maintaining cognitive vitality with aging. Mayo Clin Proc. Jul 2002;77(7):681-696. [CrossRef] [Medline]
- Jin Y, Jing M, Ma X. Effects of digital device ownership on cognitive decline in a middle-aged and elderly population: longitudinal observational study. J Med Internet Res. Jul 29, 2019;21(7):e14210. [FREE Full text] [CrossRef] [Medline]
- Tun PA, Lachman ME. The association between computer use and cognition across adulthood: use it so you won't lose it? Psychol Aging. Sep 2010;25(3):560-568. [CrossRef] [Medline]
- Lin L, Jing X, Lv S, Liang J, Tian L, Li H, et al. Mobile device use and the cognitive function and depressive symptoms of older adults living in residential care homes. BMC Geriatr. Feb 03, 2020;20(1):41. [FREE Full text] [CrossRef] [Medline]
- Wang G, Zhang W, Zeng R. WeChat use intensity and social support: The moderating effect of motivators for WeChat use. Computers in Human Behavior. Feb 2019;91:244-251. [FREE Full text] [CrossRef]
- Joe J, Demiris G. Older adults and mobile phones for health: a review. J Biomed Inform. Oct 2013;46(5):947-954. [FREE Full text] [CrossRef] [Medline]
- Liu X, Wang X, Zhang H, Pei M, Li N. Relationship between digital exclusion and cognitive impairment in Chinese adults. Front Aging Neurosci. Jul 3, 2023;15:1194348. [FREE Full text] [CrossRef] [Medline]
- Bennett DA, Wilson RS, Schneider JA, Evans DA, Mendes de Leon CF, Arnold SE, et al. Education modifies the relation of AD pathology to level of cognitive function in older persons. Neurology. Jun 24, 2003;60(12):1909-1915. [CrossRef] [Medline]
- Evans D, Beckett L, Albert M, Hebert L, Scherr P, Funkenstein H, et al. Level of education and change in cognitive function in a community population of older persons. Ann Epidemiol. Jan 1993;3(1):71-77. [FREE Full text] [CrossRef] [Medline]
- Lyketsos CG, Chen L, Anthony JC. Cognitive decline in adulthood: an 11.5-year follow-up of the Baltimore Epidemiologic Catchment Area study. Am J Psychiatry. Jan 1999;156(1):58-65. [CrossRef] [Medline]
- Carmelli D, Swan GE, LaRue A, Eslinger PJ. Correlates of change in cognitive function in survivors from the Western Collaborative Group Study. Neuroepidemiology. Dec 3, 1997;16(6):285-295. [CrossRef] [Medline]
- Hultsch DF, Hertzog C, Dixon RA, Small BJ. Memory change in the aged. New York, NY. Cambridge University Press; 1998.
- Alley D, Suthers K, Crimmins E. Education and cognitive decline in older Americans: results from the AHEAD sample. Res Aging. Jan 01, 2007;29(1):73-94. [CrossRef] [Medline]
- Proust-Lima C, Amieva H, Letenneur L, Orgogozo J, Jacqmin-Gadda H, Dartigues J. Gender and education impact on brain aging: a general cognitive factor approach. Psychol Aging. Sep 2008;23(3):608-620. [CrossRef] [Medline]
- Glauber R. Rural depopulation and the rural-urban gap in cognitive functioning among older adults. J Rural Health. Sep 07, 2022;38(4):696-704. [CrossRef] [Medline]
- Zhang T, Lu B, Wang X. Urban-rural disparity in cognitive performance among older Chinese adults: explaining the changes from 2008 to 2018. Front Public Health. Mar 23, 2022;10:843608. [FREE Full text] [CrossRef] [Medline]
- Wight R, Aneshensel C, Miller-Martinez D, Botticello A, Cummings J, Karlamangla A, et al. Urban neighborhood context, educational attainment, and cognitive function among older adults. Am J Epidemiol. Jun 15, 2006;163(12):1071-1078. [CrossRef] [Medline]
- Liu D, Li L, An L, Cheng G, Chen C, Zou M, et al. Urban-rural disparities in mild cognitive impairment and its functional subtypes among community-dwelling older residents in central China. Gen Psychiatr. Oct 27, 2021;34(5):e100564. [FREE Full text] [CrossRef] [Medline]
- 2020 National Tracking Survey. CHARLS. URL: https://charls.charlsdata.com/pages/data/111/zh-cn.html [accessed 2024-11-21]
Abbreviations
CHARLS: China Health and Retirement Longitudinal Study |
GEE: generalized estimating equations |
IPTW: inverse probability of treatment weighting |
KHB: Karlson-Holm-Breen method |
OW: overlap weighting |
PSM: propensity score matching |
Edited by N Cahill, T Leung; submitted 15.02.24; peer-reviewed by A Azizan, T Soluoku; comments to author 06.05.24; revised version received 31.05.24; accepted 12.11.24; published 02.12.24.
Copyright©Bowen Chen, Chun Yang, Shanshan Ren, Penggao Li, Jin Zhao. Originally published in the Journal of Medical Internet Research (https://www.jmir.org), 02.12.2024.
This is an open-access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in the Journal of Medical Internet Research (ISSN 1438-8871), is properly cited. The complete bibliographic information, a link to the original publication on https://www.jmir.org/, as well as this copyright and license information must be included.