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.
Background
With the widespread use of the internet, the number of older internet users is rapidly increasing. However, the role of individual factors such as self-efficacy, health literacy, and physical exercise in the chain of influence between internet use and the health of older adults is unclear.
Objective
On the basis of the media effects theory and the health belief model, we aimed to explore the relationship between internet use and self-rated health among Chinese older adults. We also analyzed the mediating roles of self-efficacy, health literacy, and physical exercise and examined how these mediating effects varied by sex, residence, and per capita monthly household income groups. In addition, we investigated the moderating effect of education level.
Methods
We included 1147 participants aged ≥60 years from the 2021 Psychology and Behavior Investigation of Chinese Residents project in this study. Using a combination of multiple linear regression and bootstrap testing, we constructed a chain mediation model and a moderated chain mediation model to examine how internet use affects older adults’ health through self-efficacy, health literacy, and physical exercise. In addition, we explored the marginal effects of education level within each mediation pathway.
Results
Internet use significantly improved the self-rated health of older adults (B=2.183; P<.001), and this improvement was exclusively mediated by self-efficacy (B=0.502; P<.001), health literacy (B=5.415; P<.001), and physical exercise (B=3.449; P<.001). These three factors could act as independent mediators or form sequential chain mediation pathways. In heterogeneity analysis, the total indirect effects were more pronounced among female participants (B=1.965; P<.05) and individuals with middle (B=1.971; P<.05) to high (B=2.710; P<.05) income levels. Furthermore, education level moderated the relationships between internet use and self-efficacy (B=0.452; P=.003) and between internet use and self-rated health (B=1.284; P=.01). This suggests that the positive influence of internet use on self-rated health was more pronounced among older adults with higher education.
Conclusions
The findings of this study suggested that internet use can positively influence older adults’ self-rated health through the chain mediated effects of self-efficacy, health literacy, and physical exercise. This chain mediated effect was more pronounced among those with higher levels of education. In the future, efforts should be made to promote internet use among older persons by developing age-friendly digital platforms and expanding digital training and health education. Moreover, older persons should be encouraged to participate in volunteer activities to increase their self-efficacy and improve their health.
chain mediation effectdigital engagementolder adult populationsubjective health statusIntroductionBackground
The wave of population aging is sweeping across the globe, and China, the second most populous country, is also experiencing an accelerated aging process. By the end of 2023, the population of individuals aged ≥60 years in China reached 296.97 million, accounting for 21.1% of the total population [1]. Meanwhile, along with the development and popularization of the internet, the number of older internet users is growing. As of December 2023, the number of internet users in China aged ≥60 years reached 170 million, accounting for 15.6% of the country’s internet users, and this share is still increasing [2].
The internet is playing an increasingly important role in the daily lives of older adults. Older adults can use the internet to contact friends and relatives, read news, and stay in touch with society. In addition, they can use the internet to watch videos, access health-related information, and engage in online shopping to facilitate their lives. Therefore, scholars have shown keen interest in understanding how internet use affects the health of older adults. Most studies have found positive effects of internet use on the health of older adults, such as reduced loneliness and depression [3], increased well-being [4], enhanced cognitive functioning [5], and better chronic disease management [6]. Internet use has also been associated with healthy behaviors, including physical exercise and preventive health actions [7,8]. However, some studies suggest that excessive internet use may have negative effects on older adults, such as reducing real-life social interactions [9] and lowering life satisfaction [10].
Beyond these general associations, researchers have explored the underlying mechanisms through which internet use affects health. Most existing studies have examined, from the perspective of social participation, the pathways through which social capital influences the relationship between internet use and health among older adults, focusing on factors such as social engagement, social trust, and social support [11-13]. In contrast, less attention has been paid to the individual-level psychological and behavioral mechanisms through which internet use affects health. Although studies have pointed out the positive effects of internet use on older adults’ self-efficacy [14], health literacy [14], and lifestyle behaviors [15], few studies have systematically examined how these factors interact within a chain mediation framework. Moreover, the moderating role of education in shaping these mechanisms has received limited attention in existing research, despite its potential importance in understanding disparities in health outcomes.
Addressing these gaps, we aimed to investigate the association between internet use and health among older adults in China, focusing on the chain mediation roles of self-efficacy, health literacy, and physical exercise. We also examined whether education level moderates these mediation paths. Drawing on nationally representative data from the 2021 Psychology and Behavior Investigation of Chinese Residents (PBICR) project, this study contributes to a more comprehensive understanding of how digital engagement supports healthy aging and offers policy implications at both the individual and societal levels.
Theoretical FoundationMedia Effects Theory
Media effects theory explores how media use affects the cognition, emotions, attitudes, and behaviors of individuals or groups [16,17]. The theory posits that such influences are often indirect, meaning that media use first affects psychological factors such as attitudes and beliefs, which in turn shape behavior [16]. In the digital age, the internet has become an important medium through which older adults learn new skills, maintain social connections, and access health-related information. These experiences help enhance their self-efficacy [18] and improve their health knowledge [19]. Moreover, internet use can improve older adults’ health literacy [20], enabling them to better understand and apply health information, thereby facilitating the adoption of health-promoting behaviors [21].
Health Belief Model
The health belief model (HBM) is a sociopsychological model developed by social psychologists in the US Public Health Service in the 1950s to explain and predict individuals’ health-related behaviors [22]. It has been widely used in the field of digital health. For example, researchers have used the HBM to examine factors influencing women’s adoption of diet and exercise content on social media during the COVID-19 pandemic [23] and to analyze determinants of menopausal women’s intention to use mental health applications [24]. In addition, the HBM has been used to investigate the role of social media in promoting public health behaviors [25]. In the HBM, when individuals possess a high level of self-efficacy, they are more confident in their ability to successfully execute health behaviors and achieve healthy outcomes. Consequently, they are more likely to overcome obstacles to behavior change and adopt healthy behaviors. For instance, studies have shown that improving students’ self-efficacy can lead to better oral hygiene behaviors and improved oral health outcomes [26]. Furthermore, self-efficacy has been identified as a significant predictor of older adults’ online health information–seeking behavior [27], further supporting its crucial role in facilitating health behavior.
In summary, the media effects theory explains how internet use can influence individuals’ attitudes (such as self-efficacy) and their ability to internalize health information (health literacy). The HBM illustrates how self-efficacy drives the transformation of these cognitive factors into health behaviors (such as physical exercise). By integrating the media effects theory and the HBM, this study constructs a logical pathway from media use to cognitive and attitudinal changes, ultimately leading to health behaviors and outcomes. This framework explains how internet use indirectly influences self-rated health through the mediating roles of self-efficacy, health literacy, and physical exercise (Figure 1).
Chain mediation pathway of self-efficacy, health literacy, and physical exercise.
Research HypothesesRelationship Between Internet Use and Older Adults’ Health
Existing research has revealed that internet use among older adults impacts their self-rated health, mental health, physical health, and chronic disease management. Internet use can improve cognitive function by increasing activity participation [28], and it has also been shown to positively influence objective physical health [29]. In chronic disease management, internet use effectively increases physical exercise, aiding in the prevention of obesity, heart disease, hypertension, diabetes, and other chronic conditions [30]. Among various health indicators, self-rated health is a comprehensive measure and a powerful predictor of well-being [31]. Research has shown that internet use may enhance self-rated health. For instance, Zhou et al [32] found a positive correlation between internet use and self-rated health; Liu et al [13] discovered that Chinese older adult internet users reported better health. Therefore, we focused on self-rated health and proposed the following hypothesis:
Hypothesis 1: internet use positively influences the self-rated health among older adults.
Mediating Role of Self-Efficacy
Self-efficacy refers to an individual’s perceived ability and confidence to handle challenges in different situations, and it is considered a relatively stable psychological trait [33]. Previous research has identified a significant positive correlation between internet use and self-efficacy in older adults [34]. The use of information and communication technology allows older adults to learn new skills and remain socially active, which helps to increase their self-efficacy and reduce their social isolation [18]. Furthermore, according to the HBM, self-efficacy can promote health-related behaviors. For example, one study [27] found that higher self-efficacy can positively influence health information–seeking intentions by enhancing expectations of health-related outcomes, thus encouraging positive health management behaviors. To summarize, these findings suggest that internet use may improve older adults’ health by enhancing their self-efficacy. Therefore, the following research hypothesis was proposed:
Hypothesis 2: self-efficacy is a mediator between internet use and self-rated health among older adults.
Mediating Role of Health Literacy
Health literacy is defined as the ability to acquire, process, understand, and communicate information related to health. This ability facilitates healthy decision-making and the management of health conditions [35]. Higher levels of health literacy are advantageous for both physical and mental well-being [36,37]. Internet use has the potential to elevate health literacy by enhancing access to health information [38]. It provides older adults with convenient channels to obtain scientific and authoritative health knowledge and information. This information not only boosts their health literacy but also guides them to adopt healthier lifestyles, such as balanced diets and moderate exercise, thereby promoting their overall well-being. Moreover, individuals with higher health literacy are more likely to understand and adhere to medical advice and protocols, leading to improved health outcomes [39]. Studies have found a significant correlation between frequent internet use, good health status, and higher health literacy [40]. Therefore, the following research hypothesis was proposed:
Hypothesis 3: health literacy is a mediator between internet use and self-rated health among older adults.
Mediating Role of Physical Exercise
Health behaviors typically refer to a series of positive lifestyle choices aimed at disease prevention and promotion of physical and mental well-being. Physical exercise is a crucial determinant of health, capable of preventing numerous diseases, including cardiovascular diseases [41]. Previous studies have indicated a link between high levels of internet use and high levels of physical exercise [42]. Guo et al [15], using data from the 2018 China Health and Retirement Longitudinal Study, found that internet use and its frequency significantly increased the probability and frequency of physical exercise participation among middle-aged and older adults. Physical exercise not only benefits physical health but also positively influences emotional states such as anxiety, stress, and depression through physiological and biochemical mechanisms, including the promotion of dopamine secretion [43], thereby contributing to mental well-being. Zhang and Zhang [44] pointed out that the internet facilitates physical exercise in multiple ways, and active exercise further impacts individuals’ mental health status. Liu et al [45] found that longer internet use predicted lower levels of depressive symptoms, and the frequency of physical activities, such as exercise, mediated the longitudinal relationship between internet use and depressive symptoms among older adults. Therefore, the fourth research hypothesis of this paper was proposed as follows:
Hypothesis 4: physical exercise is a mediator between internet use and self-rated health among older adults.
Chain Mediating Effects of Self-Efficacy, Health Literacy, and Physical Exercise
The effects of self-efficacy, health literacy, and physical activity on health do not exist independently but may interact to form a chain mediation pathway linking internet use and health.
First, self-efficacy has been shown to positively influence health literacy. Individuals with high self-efficacy tend to be more proactive in acquiring health-related knowledge and are more motivated to adopt healthy socialization and lifestyles [46,47].
Second, self-efficacy is a strong predictor of physical exercise [48,49]. Older adults with high self-efficacy are more confident in achieving health goals, which motivates them to initiate and maintain physical exercise routines [50].
Third, health literacy positively influences physical exercise [51,52]. Individuals with higher health literacy are better able to understand and act upon health information [53], which helps increase their participation in physical exercise [54]. High health literacy enables older adults to accurately assess their health status, adopt health advice, and serve as a basis for physical exercise.
These variables are not only interrelated but also follow a progressive pathway. Internet use may enhance older adults’ self-efficacy, which subsequently improves their health literacy, promotes engagement in physical exercise, and ultimately contributes to better health outcomes. Therefore, the following research hypotheses were proposed:
Hypothesis 5: self-efficacy and health literacy form a chain mediation pathway between internet use and self-rated health among older adults.
Hypothesis 6: self-efficacy and physical exercise form a chain mediation pathway between internet use and self-rated health among older adults.
Hypothesis 7: health literacy and physical exercise form a chain mediation pathway between internet use and self-rated health among older adults.
Hypothesis 8: self-efficacy, health literacy, and physical exercise form a chain mediation pathway between internet use and self-rated health among older adults.
Moderating Effect of Education Level
Previous studies have found that education level is a significant factor influencing internet use among older adults. Education levels reflect disparities in cultural capital [55], which may influence the relationship between internet use and health among older adults [56]. Compared to their less-educated counterparts, older adults with higher education levels have greater access to the internet and are more likely to acquire, understand, and learn health-related information online. Therefore, we argue that education level moderates the indirect pathways linking internet use to health. In other words, different levels of education may alter the strength of the mediation effects between internet use and self-rated health. On this basis, the following hypothesis was proposed:
Hypothesis 9: education level positively moderates the relationship between internet use and self-rated health among older adults.
The theoretical hypothesis model, on the basis of the aforementioned research hypotheses, is shown in Figure 2.
Hypothesis model. Hypothesis 1: internet use to self-rated health; hypothesis 2: internet use to self-efficacy to self-rated health; hypothesis 3: internet use to health literacy to self-rated health; hypothesis 4: internet use to physical exercise to self-rated health; hypothesis 5: internet use to self-efficacy to health literacy to self-rated health; hypothesis 6: internet use to self-efficacy to physical exercise to self-rated health; hypothesis 7: internet use to health literacy to physical exercise to self-rated health; hypothesis 8: internet use to self-efficacy to health literacy to physical exercise to self-rated health; hypothesis 9: adjustment of education level—internet use to self-rated health.
MethodsData Source
The data used in this study are the results of the 2021 survey of the PBICR project (the complete questionnaire used in the study can be found in Multimedia Appendix 1). Organized by the School of Public Health at Peking University, the PBICR project has conducted annual surveys since 2020, collecting data across 4 waves. The most recent publicly available data are from 2021. The project used a multistage sampling method to select 120 cities from 23 provinces, 5 autonomous regions, and 4 municipalities in China. All provincial capitals and municipalities were included by default. In addition, 2 to 6 noncapital prefecture-level cities in each province or region were selected using a random number table, ensuring randomization and regional diversity (detailed information on the survey regions can be found in Multimedia Appendix 2).
Within each selected city, quota sampling was used based on the 2021 national census data, matching local distributions in age, sex, and urban-rural residency. A total of 11,032 valid responses were collected. For this study, we focused on respondents aged ≥60 years, resulting in a final sample of 1147 older adults.
Ethical Considerations
This study was conducted in accordance with the ethical principles outlined in the Declaration of Helsinki and the International Ethical Guidelines for Biomedical Research Involving Human Subjects. Ethical approval was obtained from the Ethics Committee of the Shaanxi Provincial Research Center for Health Culture (approval JKWH-2021-01) and the Ethics Committee of Jinan University (approval JNUKY-2021-018). All participants provided written informed consent prior to participation. No compensation was provided to participants in this study. The research protocol was designed to ensure the protection of participants’ privacy and the confidentiality of the collected data.
Variable SettingExplained Variable
Self-rated health was chosen as the explained variable. The survey question used to assess self-rated health was, “How would you rate your health today?” Respondents were asked to rate their health on a scale from 0 to 100, with higher scores indicating better self-rated health.
Explanatory Variable
The explanatory variable in this study was internet use. The relevant question in the survey was, “How often do you typically use the following media?” Internet use was measured based on the reported frequency of using PCs (including tablets) or smartphones. Participants were categorized into a 5-category ordinal variable according to their use frequency. Detailed variable coding is presented in Table 1.
Definition of each variable.
Variable types and variable names
Variable definition
Explained variable
Self-rated health
Self-rated health score ranged from 0 to 100
Explanatory variable
Internet use
Never=1; occasionally (≤1 d per wk)=2; sometimes (2-3 d per wk)=3; frequently (4-5 d per wk)=4; almost every day (6-7 d per wk)=5
Control variables
Age group (y)
60-69=1; 70-79=2; ≥80=3
Sex
Male=1; female=0
Education level
Junior high school and below=1; high school or vocational school=2; university (including junior college) and above=3
Marital status
Married=1; other (including divorced, unmarried, and widowed)=0
Residence
Urban=1; rural=0
Per capita monthly household income (RMB)a
≤3000=1; 3001-6000=2; ＞6000=3a
Living situation
Living alone=1; other=0
Medical insurance
Participate in medical insurance (including urban employee medical insurance, urban and rural resident medical insurance, commercial medical insurance, and public medical insurance)=1; out-of-pocket spending=0
Chronic diseases (n)
Experiencing 1 or no chronic disease=1; experiencing 2 chronic diseases=2; experiencing 3 chronic diseases=3; experiencing 4 or more chronic diseases=4
The selection of control variables was informed by previous studies on the association between internet use and health in older adults [8,12,13]. These included individual characteristics (age, sex, and education level); family-related factors (marital status and living situation); and socioeconomic conditions (residence, per capita monthly household income, and medical insurance). In addition, the number of chronic diseases experienced was included as a key health-related covariate that may influence self-rated health. These variables were included to reduce potential confounding and improve the robustness of the estimated mediation effects. The specific settings for each variable are presented in Table 1.
Mediator Variables
The mediator variables selected for this study included self-efficacy, health literacy, and physical exercise (Table 1).
Self-efficacy was measured using the New General Self-Efficacy Scale, developed by Chen et al [57] in 2001. It consists of 8 items, with responses ranging from 1 (strongly disagree) to 5 (strongly agree). The total score ranges from 8 to 40, with higher scores reflecting greater levels of general self-efficacy.
Health literacy was assessed using the 12-item Short-Form Health Literacy Questionnaire [58], which consists of 3 assessment dimensions: health care, disease prevention, and health promotion. Each item is scored on a 4-point Likert scale (1=very difficult; 2=difficult; 3=easy; and 4=very easy). The health literacy index was standardized using the following formula: [index = (mean – 1) × (50/3)], where the “mean” represents the average score for each individual. The resulting health literacy scores ranged from 0 to 50, with scores of 0 to 25 classified as inadequate, 26 to 33 as problematic, 34 to 42 as sufficient, and 43 to 50 as excellent [59].
Physical exercise was measured based on the weekly total exercise time. Activities included fitness exercises (eg, dumbbell lifting), walking, swimming, cycling, using sports equipment (eg, treadmills), and other forms of aerobic exercise (eg, running and table tennis). The total exercise time (in minutes) was calculated by summing the duration of all activities, with ≥150 minutes per week considered a healthy level of physical exercise [60].
Moderating Variable
The moderating variable selected for this study was the education level of older adults. Education level was assessed using the questionnaire item, “Your highest level of education,” with responses categorized into the following groups: junior high school or below, high school or junior college, and bachelor’s degree or higher. These categories were determined based on the characteristics of the sample population and the available data.
Statistical Methods
The statistical analysis of this study was conducted in 5 main stages. First, a descriptive statistical analysis was performed. Second, a multiple linear regression model was used to preliminarily examine the relationship between internet use and self-rated health. Third, a chain mediation analysis was conducted to test whether self-efficacy, health literacy, and physical exercise sequentially mediated the relationship between internet use and self-rated health. Fourth, heterogeneity analyses were performed to examine whether the mediation effects varied by sex, residence, and per capita monthly household income groups. Fifth, a moderated mediation analysis was conducted to assess whether education level moderated the specific mediation paths within the chain model.
In addition, to assess the validity of self-rated health as an explained variable, an additional analysis was conducted to examine its association with BMI. This analysis was intended to examine whether subjective health perceptions aligned with objective health indicator.
All mediation and moderation analyses were conducted using the PROCESS macro (version 4.0), developed by Hayes and Rockwood [61], with 5000 bootstrap resamples. A mediation or moderation effect was considered statistically significant if the 95% CI did not include 0. All statistical analyses were performed using SPSS (version 25.0; IBM Corp), with 2-sided tests and a significance level set at P<.05.
ResultsDescriptive Analysis
Participant characteristics are presented in Table 2. The mean self-rated health score of the 1147 participants was 73.90 (SD 19.61). Among them, 356 (31.04%) participants used the internet almost daily. Most of the participants (525/1147, 45.77%) were aged between 70 and 79 years. Sex distribution was approximately equal (581/1147, 50.65% men and 566/1147, 49.35% women). Most older adults had a junior high school education or less (737/1147, 64.25%), and most were married (894/1147, 77.94%).
Descriptive statistical results of each variable (N=1147).
Variable types and variable names
Values
Explained variable
Self-rated health, mean (SD)
73.90 (19.61)
Explanatory variable
Internet use, n (%)
Never
317 (27.64)
Occasionally
97 (8.46)
Sometimes
183 (15.95)
Frequently
194 (16.91)
Almost every day
356 (31.04)
Control variables
Age group (y), n (%)
60-69
514 (44.81)
70-79
525 (45.77)
≥80
108 (9.42)
Sex, n (%)
Male
581 (50.65)
Female
566 (49.35)
Education level, n (%)
Junior high school and below
737 (64.25)
High school or vocational school
200 (17.44)
University and above
210 (18.31)
Marital status, n (%)
Married
894 (77.94)
Other
253 (22.06)
Residence, n (%)
Urban
654 (57.02)
Rural
493 (42.98)
Per capita monthly household income (RMB)a, n (%)
≤3000
471 (41.06)
3001-6000
406 (35.4)
＞6000
270 (23.54)
Living situation, n (%)
Living alone
136 (11.86)
Other
1011 (88.14)
Medical insurance, n (%)
Participate in medical insurance
1017 (88.67)
Out-of-pocket spending
130 (11.33)
Chronic diseases (n), n (%)
≤1
831 (72.45)
2
207 (18.05)
3
74 (6.45)
>4
35 (3.05)
Mediator variables
Self-efficacy, mean (SD)
28.10 (5.34)
Health literacy, n (%)
Inadequate
339 (29.56)
Problematic
574 (50.04)
Sufficient
146 (12.73)
Excellent
88 (7.67)
Physical exercise, n (%)
0 min
121 (10.55)
<150 min
362 (31.56)
>150 min
664 (57.89)
aRMB ¥1=US $0.155 (2021 average).
There were more urban residents (654/1147, 57.02%) than rural residents (493/1147, 42.98%). The average monthly per capita household income primarily fell into 2 categories: ≤RMB 3000 (US $465) or less (471/1147, 41.06%) and RMB 3001 (US $465.15) to RMB 6000 (US $930; 406/1147, 35.4%). Most participants did not live alone (1011/1147, 88.14%), and most participants had health insurance (1017/1147, 88.67%). In addition, 831 participants (72.45%) had at most one chronic medical condition.
The mean self-efficacy score for participants was 28.10 (SD 5.34). In terms of health literacy, most participants were classified as inadequate (339/1147, 29.56%) or problematic (574/1147, 50.04%). More than half of the older adults (664/1147, 57.89%) were engaged in at least 150 minutes of physical exercise per week.
The Relationship Between Internet Use and Self-Rated Health
Multiple linear regression analyses were conducted with self-rated health as the explained variable and internet use as the explanatory variable. Collinearity was also assessed. The results of the analyses are shown in Table 3. The results of collinearity diagnosis showed that the variance inflation factor values for all variables in both models A and B were <10, indicating the absence of significant multicollinearity.
Results of multiple linear regression analysis.
Variables
Model Aa (self-rated health)
Model Bb (self-rated health)
Bc
P value
VIFd
B
P value
VIF
Explanatory variable
Internet use
2.182
<.001
1.269
0.474
.23
1.480
Control variables
Age
−0.180
.84
1.093
−0.405
.63
1.097
Sex
1.075
.33
1.036
−0.099
.93
1.059
Education level
2.056
.01
1.404
0.976
.22
1.428
Marital status
3.399
.03
1.421
2.696
.07
1.425
Residence
0.871
.48
1.222
−0.641
.58
1.247
Per capita monthly household income
−0.409
.62
1.376
−0.653
.40
1.386
Living situation
4.694
.02
1.368
4.575
.01
1.370
Medical insurance
6.343
<.001
1.024
3.924
.02
1.045
Chronic diseases
−5.275
<.001
1.040
−4.584
<.001
1.053
Mediator variables
Self-efficacy
N/Ae
N/A
N/A
0.502
<.001
1.221
Health literacy
N/A
N/A
N/A
5.415
<.001
1.368
Physical exercise
N/A
N/A
N/A
3.449
<.001
1.268
Constant
62.414
<.001
N/A
40.091
<.001
N/A
aR2=0.116; F10, 1136=14.911; P<.001.
bR2=0.212; F13, 1133=23.461; P<.001.
cB: unstandardized regression coefficient.
dVIF: variance inflation factor.
eN/A: not applicable.
Compared with model A, model B had 3 new control variables: self-efficacy, health literacy, and exercise level. As shown in model A, internet use had a significant positive effect on self-rated health in older adults (B=2.182, P<.001), suggesting that internet use may enhance the self-rated health of older adults. Model B further showed that self-efficacy, health literacy, and physical exercise could positively influence older adults’ self-rated health (P<.001). Comparing models A and B, we found that, after adding the 3 variables of self-efficacy, health literacy, and physicalexercise, the significance of the regression coefficients of internet use on older adults’ self-rated health disappeared (B=0.474; P=.23). This finding suggests that self-efficacy, health literacy, and physical exercise may mediate the relationship between internet use and older adults’ self-rated health.
Chain Mediation Analysis
To explore the roles of self-efficacy, health literacy, and physical exercise in the relationship between internet use and self-rated health among older adults, we further conducted a chain mediation analysis [61]. In the chain mediation analysis, self-efficacy was a continuous variable, and all other variables were categorical. Therefore, self-efficacy scores were standardized before analysis.
The results of the chain mediation analysis for internet use and self-rated health are presented in Table 4. According to model 1, internet use significantly predicted self-rated health (B=2.183; P<.001), supporting hypothesis 1 and suggesting that internet use had a positive impact on older adults’ self-rated health. Models 2, 3, and 4 showed that internet use significantly predicted self-efficacy (B=0.868, P<.001); health literacy (B=0.130, P<.001); and physical exercise (B=0.088, P<.001).
Regression results of the relationship among internet use, self-efficacy, health literacy, physical exercise, and self-rated health.
Variables
Model 1a (self-rated health)
Model 2b (self-efficacy)
Model 3c (health literacy)
Model 4d (physical exercise)
Model 5e (self-rated health)
Bf
P value
B
P value
B
P value
B
P value
B
P value
Explanatory variable
Internet use
2.183
<.001
0.868
<.001
0.13
<.001
0.088
<.001
0.474
.23
Mediator variables
Self-efficacy
N/Ag
N/A
N/A
N/A
0.046
<.001
0.005
.21
0.502
<.001
Health literacy
N/A
N/A
N/A
N/A
N/A
N/A
0.068
.006
5.415
<.001
Physical exercise
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
3.449
<.001
Control variables
Age (y)
−0.180
.84
0.469
.05
−0.013
.71
−0.019
.51
−0.405
.63
Sex
1.075
.33
0.400
.19
0.048
.27
0.171
<.001
−0.099
.93
Education level
2.056
.01
0.409
.07
0.084
.01
0.084
.002
0.976
.22
Marital status
3.400
.03
0.441
.31
0.079
.20
−0.025
.62
2.696
.07
Residence
0.871
.48
0.270
.42
0.141
.004
0.148
<.001
−0.641
.58
Per capita monthly household income
−0.410
.62
−0.179
.43
0.022
.50
0.075
.005
−0.653
.40
Living situation
4.694
.02
−0.494
.36
0.047
.55
0.069
.29
4.575
.01
Medical insurance
6.343
<.001
1.409
.003
0.118
.09
0.191
.001
3.924
.02
Chronic diseases
−5.275
<.001
0.019
.93
−0.096
.001
−0.047
.06
−4.584
<.001
Constant
62.414
<.001
−5.728
<.001
1.287
<.001
1.565
<.001
54.193
<.001
aR2=0.116; F10, 1136=14.911; P<.001.
bR2=0.098; F10, 1136=12.328; P<.001.
cR2=0.264; F11, 1135=37.055; P<.001.
dR2=0.211; F12, 1134=25.297; P<.001.
eR2=0.212; F13, 1133=23.461; P<.001.
fB: unstandardized regression coefficient.
gN/A: not applicable.
When internet use, self-efficacy, health literacy, and physical exercise were all included in the regression equation (model 5), the coefficients of the variables changed. Compared to model 1, the direct effect of internet use on self-rated health became nonsignificant in model 5 (B=0.474; P=.23). In contrast, self-efficacy, health literacy, and physical exercise remained significant predictors (all P<.001), which is consistent with the results of the multiple linear regression analysis.
Figure 3 and Table 5 together illustrate the results of the chain mediation analysis. In Figure 3, solid arrows indicate statistically significant paths, while dashed arrows represent nonsignificant relationships. The corresponding bootstrap test results are reported in Table 5.
Chain mediation analysis of the association between internet use and self-rated health. ****P<.001; ***P=.006; **P=.21; *P=.23.
Results of the chain mediation analysis.
Effect path
Ba (SE; 95% CI)
Proportion of effect (%)
Total effect
2.183 (0.384;1.429 to 2.936)
N/Ab
Direct effect
0.474 (0.392; −0.295 to 1.243)
N/A
Total indirect effect
1.709 (0.203; 1.325 to 2.129)
100
Internet use→self-efficacy→self-rated health
0.435 (0.124; 0.215 to 0.699)
25.48
Internet use→health literacy→self-rated health
0.703 (0.123; 0.470 to 0.957)
41.11
Internet use→physical exercise→self-rated health
0.303 (0.094; 0.136 to 0.496)
17.71
Internet use→self-efficacy→health literacy→self-rated health
0.215 (0.042; 0.139 to 0.301)
12.56
Internet use→self-efficacy→physical exercise→self-rated health
0.014 (0.013; −0.010 to 0.042)
0.81
Internet use→health literacy→physical exercise→self-rated health
0.030 (0.015; 0.007 to 0.065)
1.77
Internet use→self-efficacy→health literacy→physical exercise→self-rated health
0.009 (0.005; 0.002 to 0.020)
0.54
aB: unstandardized regression coefficient.
bN/A: not applicable.
As shown in Figure 3 and Table 5, the total effect of internet use on self-rated health was 2.183, with a 95% CI excluding 0, which was consistent with the result from model 1. This suggests that internet use had a positive impact on older adults’ self-rated health. The direct effect was 0.474, and its 95% CI included 0. This suggests that after controlling for the 3 mediating variables, internet use itself did not have a significant direct effect on older adults’ self-rated health. Instead, its positive influence operated entirely through the mediating roles of self-efficacy, health literacy, and physical exercise. The total indirect effect was 1.709, suggesting that these three factors may have a positive overall mediating effect on the relationship between internet use and self-rated health.
According to the study hypotheses, there were 7 indirect paths in the chain mediation model, as shown in Table 5. The analysis results for paths 4 to 6 indicated the significant mediating effects of self-efficacy, health literacy, and physical exercise, thereby supporting hypotheses 2, 3, and 4. At the 95% CI, the mediation effects were 0.435, 0.703, and 0.303, accounting for 25.48%, 41.11%, and 17.71% of the total indirect effect, respectively. Among the chain mediation paths (paths 7 to 10), paths 7, 9, and 10 were statistically significant, while path 8 was not. Thus, hypothesis 6 was not supported, while hypotheses 5, 7, and 8 were confirmed.
Heterogeneity Analysis
To explore whether the mediating effect of internet use on self-rated health varied by subgroup, we conducted heterogeneity analyses by sex, residence type, and per capita monthly household income. The results are shown in Table 6. In the total effect analysis, the positive association of internet use on self-rated health was statistically significant for all subgroups except for those with a per capita monthly household income ≤RMB 3000 (US $465; B=0.744; P>.05). In mediation analyses, the total indirect effect was consistently significant across groups, but this effect varied across mediation pathways. First, higher indirect effects and more effective mediation paths were observed in the female group than in the male group. Second, the effective mediating paths were generally consistent between rural and urban groups, suggesting that the ways in which internet use affected self-rated health were similar across these populations. Finally, the total indirect effect increased with monthly household income per capita, implying that the mediating effects of self-efficacy, health literacy, and physical exercise were more pronounced among the high-income group.
Results of heterogeneity analysis.
Effect path
Sex
Residence
Per capita monthly household income (RMB)a
Female, Bb
Male, B
Rural, B
Urban, B
≤3000, B
3001-6000, B
>6000, B
Total effect
2.391c
2.036c
2.012c
2.345c
0.744
2.605c
4.733c
Direct effect
0.426
0.563
0.388
0.648
−0.504
0.634
2.023
Total indirect effect
1.965c
1.473c
1.624c
1.697c
1.248c
1.971c
2.710c
Internet use→self-efficacy→self-rated health
0.695c
0.168
0.524c
0.348c
0.222
0.467c
0.804c
Internet use→health literacy→self-rated health
0.629c
0.776c
0.516c
0.840c
0.676c
0.840c
0.603c
Internet use→physical exercise→self-rated health
0.389c
0.212
0.267c
0.289c
0.068
0.448c
0.941c
Internet use→self-efficacy→health literacy→self-rated health
0.158c
0.290c
0.257c
0.179c
0.260c
0.158c
0.260c
Internet use→self-efficacy→physical exercise→self-rated health
0.000
0.021
0.028
0.004
0.012
−0.021
0.046
Internet use→health literacy→physical exercise→self-rated health
0.075c
0.004
0.021
0.031
0.008
0.065c
0.038
Internet use→self-efficacy→health literacy→physical exercise→self-rated health
0.019c
0.002
0.011
0.007
0.003
0.012c
0.017
aRMB ¥1=US $0.155 (2021 average).
bB: unstandardized regression coefficient.
cThe CI did not include 0.
Notably, the mediation paths through “health literacy” and “self-efficacy→health literacy” were significant across all subgroups. These findings highlight that self-efficacy and health literacy play a key role in the association between internet use and self-rated health.
Moderated Chain Mediation Analysis
The moderating effect of education level was tested, and the results are provided in Table 7. The results revealed that education level moderated the relationships between internet use, self-efficacy, and self-rated health (models 6 and 9). However, it did not moderate the relationships between internet use, health literacy, and physical exercise (models 7 and 8).
Moderated chain mediation test of internet use on self-rated health.
Variables
Model 6a (self-efficacy)
Model 7b (health literacy)
Model 8c (physical exercise)
Model 9d (self-rated health)
Be
P value
B
P value
B
P value
B
P value
Explanatory variable
Internet use
0.262
<.001
0.157
<.001
0.089
.001
−1.233
.12
Mediator variables
Self-efficacy
N/Af
N/A
0.046
<.001
0.005
.21
0.478
<.001
Health literacy
N/A
N/A
N/A
N/A
0.068
.006
5.463
<.001
Physical exercise
N/A
N/A
N/A
N/A
N/A
N/A
3.453
<.001
Moderating variables
Education level
−1.327
.03
0.163
.07
0.089
.24
−3.946
.07
Internet use×education level
0.452
.003
−0.021
.35
−0.001
.95
1.284
.01
Control variables
Age (y)
0.488
.04
−0.014
.69
−0.019
.51
−0.340
.69
Sex
0.456
.14
0.046
.30
0.171
<.001
0.066
.95
Marital status
0.419
.33
0.080
.20
−0.025
.62
2.641
.08
Residence
0.290
.39
0.140
.04
0.147
<.001
−0.586
.61
Per capita monthly household income
−0.137
.54
0.020
.54
0.075
.005
−0.540
.49
Living situation
−0.524
.33
0.048
.54
0.069
.29
4.475
.02
Medical insurance
1.398
.004
0.118
.09
0.191
.001
3.918
.02
Chronic diseases
0.028
.89
−0.097
.001
−0.047
.06
−4.554
<.001
Constant
−3.674
.001
1.195
<.001
1.559
<.001
59.826
<.001
aR2=0.105; F11, 1135=12.096; P<.001.
bR2=0.265; F12, 1134=34.037; P<.001.
cR2=0.211; F13, 1133=23.331; P<.001.
dR2=0.216; F14, 1132=22.311; P<.001.
eB: unstandardized regression coefficient.
fN/A: not applicable.
The interaction term between internet use and education level (B=0.452; P=.003) significantly predicted self-efficacy. Education level positively moderated the relationship between internet use and self-efficacy, as shown in Figure 4. At higher education levels, the effect of internet use on self-efficacy was stronger (B=1.313; P<.001), whereas at lower education levels, the effect was weaker (B=0.714; P<.001), as presented in Table 8. The slope for individuals with higher education was significantly steeper than that for those with lower education, indicating that among older adults with higher education, internet use had a more pronounced influence on self-efficacy.
Education level as a moderator in the relationship between internet use and self-efficacy.
Moderating effect of education level.
Moderating variable
Education level
Self-efficacy
Self-rated health
Effect
P value
Effect
P value
Mean – SD
1.000
0.714
<.001
0.051
.90
Mean
1.541
0.959
<.001
0.745
.07
Mean + SD
2.325
1.313
<.001
1.752
.007
Similarly, the interaction term between internet use and education level (B=1.284; P=.01) was a predictor of self-rated health. Education level positively moderated the relationship between internet use and self-rated health, as shown in Figure 5. The results indicated that, at higher education levels, the effect of internet use on self-efficacy was stronger (B=1.752; P=.007), whereas at lower education levels, the moderating effect was not statistically significant (B=0.051; P=.90), as presented in Table 8. The slope for individuals with higher education was steeper than for those with lower education, indicating that internet use had a greater impact on self-rated health among older adults with higher education levels.
Furthermore, this study examined the moderating effect of education level on each mediating pathway (Table 9).
Education level as a moderator in the relationship between internet use and self-rated health.
Results of moderated chain mediation analysis.
Indirect effect path and moderating effect
Ba (SE; 95% CI)
Internet use→self-efficacy→self-rated health
Low level of education
0.341 (0.106; 0.151 to 0.573)
High level of education
0.627 (0.200; 0.279 to 1.051)
Index of moderated mediation
0.216 (0.104; 0.046 to 0.450)
Internet use→health literacy→self-rated health
Low level of education
0.745 (0.137; 0.496 to 1.033)
High level of education
0.596 (0.151; 0.315 to 0.918)
Index of moderated mediation
−0.113 (0.115; −0.347 to 0.106)
Internet use→physical exercise→self-rated health
Low level of education
0.304 (0.096; 0.133 to 0.514)
High level of education
0.299 (0.109; 0.107 to 0.535)
Index of moderated mediation
−0.004 (0.063; −0.135 to 0.117)
Internet use→self-efficacy→health literacy→self-rated health
Low level of education
0.180 (0.040; 0.108 to 0.264)
High level of education
0.330 (0.074; 0.196 to 0.482)
Index of moderated mediation
0.114 (0.047; 0.029 to 0.210)
Internet use→self-efficacy→physical exercise→self-rated health
Low level of education
0.012 (0.011; −0.008 to 0.035)
High level of education
0.021 (0.020; −0.013 to 0.067)
Index of moderated mediation
0.007 (0.008; −0.004 to 0.027)
Internet use→health literacy→physical exercise→self-rated health
Low level of education
0.032 (0.015; 0.007 to 0.066)
High level of education
0.025 (0.014; 0.005 to 0.058)
Index of moderated mediation
−0.005 (0.006; 0.018 to 0.005)
Internet use→self-efficacy→health literacy→physical exercise→self-rated health
Low level of education
0.008 (0.004; 0.002 to 0.016)
High level of education
0.014 (0.007; 0.003 to 0.031)
Index of moderated mediation
0.005 (0.003; 0.001 to 0.012)
aB: unstandardized regression coefficient.
As shown in Table 9, education level significantly moderated the magnitude of the indirect effects along paths 1, 4, and 7 in the relationship between internet use and self-rated health. However, no significant moderating effect was observed in paths 2, 3, 5, and 6. Education level positively moderated the mediating effects of paths 1, 4, and 7 (moderating effects of 0.216, 0.114, and 0.005, respectively), which means that the indirect effects of mediating paths 1, 4, and 7 were stronger at higher education levels and weaker at lower education levels. Overall, hypothesis 9 was supported. Education level may positively moderate the relationship between internet use and self-rated health among older adults.
Additional Analysis
To ensure the robustness of the findings, we conducted an additional analysis in this study, which aimed to assess the correlation between self-rated health and other measurable indicators and to reduce the subjective reporting bias of self-rated health scores. If the correlation is high, the self-rated health can be considered a more realistic response to the health status of older adults. Due to data availability, the BMI indicator was chosen for analysis in this study. BMI is an important indicator of physical health in older adults, and its recommended range varies according to age. On the basis of the recommendations from the Dietary Guidelines for Chinese Residents and the BMI and Weight Management Guidelines for the Chinese Oldest Old, we set different recommended values for various age groups of older adults. Specifically, the optimal BMI range for older adults aged 60 to 64 years was 18.5-23.9 kg/m2; for the age group of 65 to 79 years, the range was 20.0-26.9 kg/m2; and for those aged ≥80 years, it was 22.0-26.9 kg/m2. Table 10 summarizes the recommended BMI values for specific age groups.
Age-specific recommended BMI ranges.
Age group (y)
Recommended BMI range (kg/m2)
60-64
18.5-23.9
65-79
20.0-26.9
≥80
22.0-26.9
In this study, BMI values were coded as a binary variable. BMI values within the recommended range were coded as 1, and values outside the range were coded as 0. To investigate the association between BMI classification and self-rated health, the Spearman rank correlation test was performed in this study.
The analysis revealed a significant positive correlation (ρ=0.097; P=.001) between BMI in the recommended range and better self-rated health. This suggests that older adults with BMI values within the recommended range tend to report better self-rated health. These findings further support the representativeness of self-rated health as an indicator of health status in older adults and justify its use as an explanatory variable in this study.
DiscussionPrincipal Findings
This study provides new insights into the mechanisms linking internet use and health outcomes among older adults, drawing upon media effects theory and the HBM. The findings indicate that internet use enhances self-rated health indirectly through a sequential chain involving self-efficacy, health literacy, and physical exercise. These three factors not only function as independent mediators but also interact cumulatively, suggesting a structured pathway of influence. The results support the view that individual-level psychological and behavioral factors play a critical role in shaping the relationship between internet use and self-rated health among older adults. Furthermore, the mediating effects were more prominent among female participants and those not categorized as low-income. In addition, education level emerged as a significant moderator, affecting the strength of the associations between internet use, self-efficacy, and self-rated health.
The Association Between Internet Use and Self-Rated Health
The results support hypothesis 1, confirming a significant positive association between internet use and self-rated health among older adults. This finding is consistent with the conclusions drawn by Fjell et al [62] and Wang and Chen [63]. From the perspective of the media effects theory, the internet, as a modern and dynamic media form, facilitates access to diverse platforms such as social media, health-related websites, and e-commerce applications. For older adults, this form of media not only expands access to reliable health information but also enhances opportunities for social connectivity. These functions contribute to improved health perceptions and behaviors. For example, the internet enables timely communication with health care professionals [64], helping older adults overcome common barriers to medical access. As a result, their health conditions are more likely to improve through early intervention and continuous guidance [65]. In addition, the internet increases social interaction among older adults, helping them promote sustained social engagement and promoting their health [66].
Mediating Roles of Self-Efficacy, Health Literacy, and Physical Exercise
The chain mediating analysis reveals that self-efficacy, health literacy, and physical exercise fully mediate between internet use and the self-rated health of older adults. This finding confirms hypotheses 2, 3, 4, 5, 6, and 8. The result illustrates that the positive impact of internet use on older adults’ self-rated health is entirely manifested through these three mediating factors. Internet access alone does not directly improve health outcomes; it is the additional benefits of self-efficacy, health literacy, and physical exercise during internet use that improve older adults’ self-rated health. These findings align with the media effects theory [16,17], which posits that sustained media engagement can lead to attitudinal and behavioral changes by shaping individuals’ perceptions, knowledge, and actions.
First, internet use is positively associated with self-efficacy. This relationship reflects the media’s role in reinforcing individual agency and perceived control. Consistent with the media effects theory, higher media exposure, such as smartphones, is associated with higher self-efficacy [67]. After engaging with various media platforms, individuals can enhance their self-efficacy through self-motivation and self-regulation [68], as well as by increasing their perception of social support [69]. In addition, the internet offers new learning environments and opportunities through online courses, aiding individuals in gaining confidence and a sense of achievement from learning and mastering new skills, thereby further boosting self-efficacy [70].
Second, higher self-efficacy facilitates the improvement of health literacy among older adults. On one hand, the internet reduces information asymmetry and broadens access to various types of information. It enables older adults to learn about fitness and health knowledge online, significantly improving their health literacy and, consequently, their health levels [20]. On the other hand, within the HBM, self-efficacy plays a crucial role in influencing behavior. It empowers older adults to believe that they can improve their health through a series of actions, encouraging them to actively learn health knowledge, enhance health literacy, and attempt health behavior changes. As self-efficacy increases, older adults become more proactive in engaging in health-related social activities and adopting healthier lifestyles, behaviors that contribute to the enhancement of health literacy [47].
Third, the improvement of health literacy can further enhance the physical exercise of older adults, thereby promoting overall health improvements. Existing research confirms the positive role of the internet in enhancing physical exercise. Older adults can watch sports broadcasts, share exercise experiences, and participate in online sports communities, gaining a wealth of sports-related information and knowledge [71]. This digital exposure supports exercise planning [72] and increases awareness of the benefits of regular activity [73], encouraging consistent participation in physical exercise. Moreover, advancements in cognitive capacity enable behavioral change. Enhanced health literacy strengthens the ability to seek and understand health information, facilitating the decision-making process for engaging in physical exercise [74].
However, the expected pathway from self-efficacy to physical exercise (hypothesis 6) was not supported. One possible explanation is that physical exercise is influenced not only by internal motivation (eg, self-efficacy) but also by external factors such as social support [75]. Previous studies have shown that social support plays a critical role in shaping older adults’ willingness to engage in physical activity [76,77]. In addition, there is an interaction between self-efficacy and social support, and older adults’ participation in physical exercise increases only when both levels are high [78]. Compared with older adults in Western countries who regularly engage in individual exercise, older adults in East Asian societies are more likely to participate in group physical activities, such as square dancing [79], which are often more dependent on social support for motivation. Thus, the effect of self-efficacy on physical activity behavior may be weakened or masked in contexts where social support is lacking. Another possible explanation is that physical decline in older adults may diminish the predictive power of psychological constructs such as self-efficacy. For example, in a study by Liu et al [80], 41% of older adult respondents reported that their physical exercise was constrained by physical limitations. As physical functioning deteriorates, actual physical capability may replace self-efficacy as a more dominant determinant of physical exercise.
Differences in Mediation Effects Among Different Subgroups
Heterogeneity analyses revealed significant differences in mediation effects across subgroups. The stronger mediation effect observed among female participants may reflect sex differences in technology use patterns and health behavior responsiveness. Previous research suggests that female individuals tend to be more proactive in seeking health information and are more likely to use this information to promote their health [8,81].
Regarding income, findings suggest that older adults with higher per capita household incomes are more likely to improve their self-rated health through internet use. This may be because higher-income groups tend to have greater digital access, higher levels of education, and better health literacy [82-84]. As a result, they can use internet resources more effectively to improve their self-efficacy and physical exercise through internet use, thus enhancing their health.
In contrast, differences in mediating effects between urban and rural areas were relatively small. This suggests that the mechanisms that translate into better self-rated health through internet use were similar across residential settings. This finding provides promising evidence that digital health interventions have the potential to close the urban-rural health gap if internet access and training are ensured [85].
Moderating Role of Education Level
The results partially support hypothesis 9. Education level significantly moderates the relationship between internet use and both self-efficacy and self-rated health. Education level can influence literacy, cognitive levels, flexibility in searching for information, and technology acceptance [86], enabling older adults with higher education to more readily grasp internet-related knowledge and skills, thereby enhancing their self-efficacy. They are more likely to find suitable health resources and support services online, helping to improve their health.
Moderated mediation analysis further revealed that education strengthens three indirect pathways from internet use through (1) self-efficacy alone; (2) self-efficacy and health literacy; and (3) the full chain, including self-efficacy, health literacy, and physical exercise. The consistent presence of self-efficacy in all moderated pathways suggests that it is the key point at which education exerts its influence. This finding is in line with and further supports a previous study [11], which reported that those with higher education often have greater autonomy and control, indicating a strong and stable self-efficacy.
However, the education level did not moderate the effects of internet use on health literacy or physical exercise directly. According to the HBM, health literacy in this context may arise more from perceived usefulness and motivation to learn than from formal educational attainment. Similarly, participation in physical exercise among older adults may be influenced more by physical condition, environment, or routine than by educational background. In both cases, the effect of education may be indirectly working through self-efficacy rather than directly.
Policy Suggestions
On the basis of the findings, this study offers targeted recommendations to promote older adults’ health in the digital age from both societal and individual perspectives.
At the societal level, efforts should focus on building age-friendly digital environments. Governments can support the development of internet platforms tailored to older adults, especially in areas such as health information, telemedicine, and physical exercise. Community-based digital training programs should be expanded to improve not only technical skills but also confidence in internet use. Health institutions and family physician teams should strengthen health education by combining offline methods (eg, lectures) with digital formats (eg, short videos and official accounts), improving access to health information across education levels. In addition, it is suggested to use mobile apps to establish online exercise communities at the community level, creating an older adult–friendly and supportive social environment, thereby motivating older adults to engage more actively in physical exercise.
At the individual level, enhancing self-efficacy is key. Older adults should be encouraged to engage in structured social activities such as volunteering or digital learning, which reinforce confidence through task achievement and social support.
Limitations and Future Research
This study has several limitations. First, the use of cross-sectional data restricts the ability to infer causal relationships between internet use and health outcomes. Second, the outcome variable was based on self-rated health, introducing the potential for reporting biases. Third, the measurement of internet use was limited to frequency, without accounting for use types, which constrains the interpretation of how specific internet behaviors relate to health. Fourth, although key demographic and health-related variables were controlled, unmeasured confounders such as cognitive functioning, personality traits, and digital literacy may still affect the results. In addition, converting the health literacy scale into categorical levels may have reduced data sensitivity.
Future research should address these limitations by using longitudinal designs, incorporating multidimensional measures of internet engagement, using objective physiological health indicators, and controlling for a broader range of confounding variables. Furthermore, modeling health literacy as a continuous construct rather than converting it into discrete categories may help preserve its full variability and improve the sensitivity of mediation analyses.
Conclusions
The study indicates that internet use can impact the health of older adults through individual and chain mediating effects of self-efficacy, health literacy, and physical exercise, with education level moderating several mediating paths. The research holds theoretical and practical value for exploring positive aging from the perspective of internet use. First, it offers a new research perspective for proactive aging and the construction of digital older adult care systems. Second, most previous studies have only discussed individual mediating effects, neglecting the chain impact of multiple mediators. Moreover, the study considers the moderating role of education level on mediating paths, providing a new perspective for understanding the influence of cultural capital on health. Finally, this study proposes a few suggestions to support healthy aging, including promoting age-friendly digital platforms, expanding digital training and health education, and encouraging older adults to engage in volunteering or digital learning to enhance self-efficacy.
Selected questionnaire questions from the Psychology and Behavior Investigation of Chinese Residents 2021.
Detailed information on survey regions.
AbbreviationsHBM
health belief model
PBICR
Psychology and Behavior Investigation of Chinese Residents
This study was supported by the General Project of Ministry of Education Foundation on Humanities and Social Sciences, China (grant 22YJAZH082).
Data Availability
The Psychology and Behavior Investigation of Chinese Residents (PBICR) is an open access, large-sample, nationwide representative study. It regularly collects high-quality microdata from representative Chinese populations, including the general population, special groups, and patients with specific diseases, to analyze psychological and behavioral issues among Chinese residents. The study currently provides access to data from 2020 and 2021, which can be requested by following the detailed procedure outlined on the PBICR website [87]. On the website, applicants may click the section “Open Database” to view the full data access instructions, including contact information and required documents.
CM and WL contributed to the study design and formulation of overarching research goals. WL contributed to the design of the methodology, data collection, data analysis, and writing of the paper. CM revised and reviewed the manuscript. XL and ZS contributed to writing the initial draft.
None declared.
PingpingWThe total population has declined, and the high-quality development of the population has been effectiveNational Bureau of Statistics2024-06-07https://www.stats.gov.cn/xxgk/jd/sjjd2020/202401/t20240118_1946711.htmlThe 53rd statistical report on the development of the internet in ChinaChina Internet Network Information Center2024-06-08https://www3.cnnic.cn/n4/2024/0322/c88-10964.htmlLamSSMJivrajSScholesSExploring the relationship between internet use and mental health among older adults in England: longitudinal observational studyJ Med Internet Res20200728227e1568310.2196/1568332718913v22i7e15683PMC7420689DongXMengSChenDHow does the Internet enhance the subjective well-being of elderly individuals in China?Front Psychol2022101813103616910.3389/fpsyg.2022.103616936329732PMC9622754YuanWJiangMGongSHow to improve the cognitive health of middle-aged and elderly people: evidence from China family panel studiesFront Public Health20221079925510.3389/fpubh.2022.7992553518684035186840LiPZhangCGaoSZhangYLiangXWangCZhuTLiWAssociation between daily internet use and incidence of chronic diseases among older adults: prospective cohort studyJ Med Internet Res2023071725e4629810.2196/4629837459155v25i1e4629837459155PengYIChanYSDo internet users lead a healthier lifestyle?J Appl Gerontol2020033932778410.1177/07334648187857972995708729957087GuoELiJLuoLGaoYWangZThe effect and mechanism of internet use on the physical health of the older people-Empirical analysis based on CFPSFront Public Health202210191095285810.3389/fpubh.2022.9528583633914036339140KrautRPattersonMLundmarkVKieslerSMukopadhyayTScherlisWInternet paradox. A social technology that reduces social involvement and psychological well-being?Am Psychol19980953910173110.1037//0003-066x.53.9.101798415799841579HageEWortmannHvan OffenbeekMBoonstraAThe dual impact of online communication on older adults' social connectivityInf Technol People2016291315010.1108/ITP-09-2014-0216LiuQPanHWuYMigration status, internet use, and social participation among middle-aged and older adults in China: consequences for depressionInt J Environ Res Public Health202008181716600710.3390/ijerph1716600732824867ijerph17166007PMC7459605LiZYangMInternet use and depressive symptoms among Chinese older adults: the mediation and suppression effects of social capitalFront Psychol20219211272979010.3389/fpsyg.2021.72979034621224PMC8490646LiuNHeYLiZThe relationship between internet use and self-rated health among older adults in China: the mediating role of social supportInt J Environ Res Public Health2022111019221478510.3390/ijerph19221478536429504ijerph192214785PMC9690403YuYWuYHuangZSunXAssociations between media use, self-efficacy, and health literacy among Chinese rural and urban elderly: a moderated mediation modelFront Public Health202311110490410.3389/fpubh.2023.110490436969672PMC10034173GuoBZhangXZhangRChenGThe association between internet use and physical exercise among middle-aged and older adults-evidence from ChinaInt J Environ Res Public Health2022120719241640110.3390/ijerph19241640136554283ijerph192416401PMC9778177ValkenburgPMOliverMOliverMBRaneyAABryantJMedia effects theories: an overviewMedia Effects: Advances in Theory and Research. 4th edition2020New York, NYRoutledge1635ValkenburgPMPeterJWaltherJBMedia effects: theory and researchAnnu Rev Psychol2016673153810.1146/annurev-psych-122414-03360826331344ChenYRSchulzPJThe effect of information communication technology interventions on reducing social isolation in the elderly: a systematic reviewJ Med Internet Res20160128181e1810.2196/jmir.459626822073v18i1e18PMC4751336TangLWangJEffects of new media use on health behaviors: a case study in ChinaIran J Public Health2021055059495810.18502/ijph.v50i5.611234183953IJPH-50-949PMC8223554EstacioEVWhittleRProtheroeJThe digital divide: examining socio-demographic factors associated with health literacy, access and use of internet to seek health informationJ Health Psychol201910241216687510.1177/135910531769542928810415ZhangCMohamadEAzlanAAWuAMaYQiYSocial media and eHealth literacy among older adults: systematic literature reviewJ Med Internet Res2025032627e6605810.2196/6605840138684v27i1e66058PMC11982777JanzNKBeckerMHThe health belief model: a decade laterHealth Educ Q198411114710.1177/1090198184011001016392204DaviesCMannAFactors influencing women to accept diet and exercise messages on social media during COVID-19 lockdowns: a qualitative application of the health belief modelHealth Mark Q20234044153310.1080/07359683.2023.219307637021625Martin-KeyNAFunnellELBenacekJSpadaroBBahnSIntention to use a mental health app for menopause: health belief model approachJMIR Form Res202410168e6043410.2196/6043439412868v8i1e60434PMC11525080BakMChinCLChinJUse of health belief model-based deep learning to understand public health beliefs in breast cancer screening from social media before and during the COVID-19 pandemicAMIA Annu Symp Proc202320232808382223951342PMC10785880MohammadkhahFRamezankhaniAAtashpooshAAhmady MoghadamFBakhtiarMKhani JeihooniAEffect of training based on health belief model and behavioral intention on improving dental and oral self-care behavior in 9-12-year-old Iranian female studentsBMC Oral Health2022111922151510.1186/s12903-022-02552-03640297210.1186/s12903-022-02552-0PMC9675157FangZLiuYPengBEmpowering older adults: bridging the digital divide in online health information seekingHumanities and Social Sciences Communications20241228111174810.1057/S41599-024-04312-7KaminSTSeifertALangFRParticipation in activities mediates the effect of Internet use on cognitive functioning in old ageInt Psychogeriatr20210133183810.1017/S104161022000363433190668S1041-6102(24)03678-0WangJLiangCLiKImpact of internet use on elderly health: empirical study based on Chinese general social survey (CGSS) dataHealthcare (Basel)2020111284110.3390/healthcare804048233198418healthcare8040482PMC7712052McGarrigleLToddCPromotion of physical activity in older people using mHealth and eHealth technologies: rapid review of reviewsJ Med Internet Res202012292212e2220110.2196/2220133372894v22i12e22201PMC7803474JylhäMWhat is self-rated health and why does it predict mortality? Towards a unified conceptual modelSoc Sci Med2009086933071610.1016/j.socscimed.2009.05.01319520474S0277-9536(09)00292-5ZhouYBaiYWangJThe impact of internet use on health among older adults in China: a nationally representative studyBMC Public Health20240417241106510.1186/s12889-024-18269-43863258810.1186/s12889-024-18269-4PMC11022463LuszczynskaAScholzUSchwarzerRThe general self-efficacy scale: multicultural validation studiesJ Psychol20050913954395710.3200/JRLP.139.5.439-45716285214EricksonJJohnsonGMInternet use and psychological wellness during late adulthoodCan J Aging20110630219720910.1017/S071498081100010924650669S0714980811000109BerkmanNDDavisTCMcCormackLHealth literacy: what is it?J Health Commun201015 Suppl 291910.1080/10810730.2010.49998520845189926926618LiuYBLiuLLiYFChenYLRelationship between health literacy, health-related behaviors and health status: a survey of elderly ChineseInt J Environ Res Public Health2015081812897142510.3390/ijerph12080971426295246ijerph120809714PMC4555308ArsenovićSTrajkovićGPekmezovićTGazibaraTAssociation of health literacy with physical and mental health in people with chronic diseasesRev Epidemiol Sante Publique20230271110141910.1016/j.respe.2022.10141936563616S0398-7620(22)00890-2ArcuryTASandbergJCMeliusKPQuandtSALengXLatulipeCMillerDPSmithDABertoniAGOlder adult internet use and eHealth literacyJ Appl Gerontol2020023921415010.1177/073346481880746830353776PMC6698430LiuCWangDLiuCJiangJWangXChenHJuXZhangXWhat is the meaning of health literacy? A systematic review and qualitative synthesisFam Med Community Health202005823310.1136/fmch-2020-00035132414834fmch-2020-000351PMC7239702SinghHKolschenJSamkange-ZeebFBrandTZeebHSchüzBModifiable predictors of health literacy in working-age adults - a rapid review and meta-analysisBMC Public Health20220730221145010.1186/s12889-022-13851-03590656710.1186/s12889-022-13851-0PMC9338662ChenLWangSXuJCSurvey on physical fitness and cardiovascular function of the city elderly in different regular physical activities in ChinaJ Nutr Health Aging201822911071110.1007/s12603-018-1070-030379310S1279-7707(23)02139-5LinLLiuJCaoXWenSXuJXueZLuJInternet addiction mediates the association between cyber victimization and psychological and physical symptoms: moderation by physical exerciseBMC Psychiatry2020040320114410.1186/s12888-020-02548-63224544310.1186/s12888-020-02548-6PMC7118978MikkelsenKStojanovskaLPolenakovicMBosevskiMApostolopoulosVExercise and mental healthMaturitas201712106485610.1016/j.maturitas.2017.09.00329150166S0378-5122(17)30856-3ZhangSZhangYThe relationship between internet use and mental health among older adults in China: the mediating role of physical exerciseRisk Manag Healthc Policy202114469770810.2147/RMHP.S33818334866945338183PMC8633706LiuNWangYLiZInternet use, physical activity and depressive symptoms in older adults: a cross-lagged panel analysisJ Affect Disord202404013509374510.1016/j.jad.2024.01.20138278330S0165-0327(24)00220-9XuXYLeungAYChauPHHealth literacy, self-efficacy, and associated factors among patients with diabetesHealth Lit Res Pract20180422e677710.3928/24748307-20180313-0131294279PMC6607806LaraRVázquezMLOgallarAGodoy-IzquierdoDOptimism and social support moderate the indirect relationship between self-efficacy and happiness through mental health in the elderlyHealth Psychol Open202072205510292094790510.1177/20551029209479053296379810.1177_2055102920947905PMC7488902PateRRDowdaMDishmanRKSaundersRPCordanKLShullERBuckoAGColabianchiNDeterminants of change in physical activity in children during the transition from elementary to high schoolMed Sci Sports Exerc2024070156712758410.1249/MSS.00000000000034223845173900005768-202407000-00007PMC11178449StecaPAdorniRSerinoSD'AddarioMSelf-efficacy beliefs as key ingredients to healthy and sustainable lifestyles. A five-year longitudinal study on diet and physical activity habits of newly diagnosed patients with acute coronary syndromeInt J Psychol2024125968223110.1002/ijop.1315138840556Collado-MateoDLavín-PérezAMPeñacobaCDel CosoJLeyton-RománMLuque-CasadoAGasquePFernández-Del-OlmoMÁAmado-AlonsoDKey factors associated with adherence to physical exercise in patients with chronic diseases and older adults: an umbrella reviewInt J Environ Res Public Health20210219184202310.3390/ijerph1804202333669679ijerph18042023PMC7922504BujaARabensteinerASperottoMGrottoGBertoncelloCCocchioSBaldovinTContuPLoriniCBaldoVHealth literacy and physical activity: a systematic reviewJ Phys Act Health20201031171212597410.1123/jpah.2020-016133129198jpah.2020-0161LamMHLeungAYThe effectiveness of health literacy oriented programs on physical activity behaviour in middle aged and older adults with type 2 diabetes: a systematic reviewHealth Psychol Res2016062341559510.4081/hpr.2016.559527403464PMC4926030NutbeamDHealth literacy as a public health goal: a challenge for contemporary health education and communication strategies into the 21st centuryHealth Prom Int20001532596710.1093/heapro/15.3.259NutbeamDThe evolving concept of health literacySoc Sci Med20081267122072810.1016/j.socscimed.2008.09.05018952344S0277-9536(08)00457-7CuiYHeYXuXZhouLNutakorJACan cultural capital, cognitive ability, and economic capacity help rural older adults bridge the digital divide? Evidence from an empirical studyFront Public Health202412139365110.3389/fpubh.2024.139365139257946PMC11385861HanJZhaoXImpact of internet use on multi-dimensional health: an empirical study based on CGSS 2017 dataFront Public Health20211018974981610.3389/fpubh.2021.74981634733815PMC8558342ChenGGullySMEdenDValidation of a new general self-efficacy scaleOrgan Res Methods2001010141628310.1177/109442810141004DuongTVAringazinaAKayupovaGNurjanahPhamTVPhamKMTruongTQNguyenKTOoWMSuTTMajidHASørensenKLinIChangYYangSChangPWDevelopment and validation of a new short-form health literacy instrument (HLS-SF12) for the general public in six Asian countriesHealth Lit Res Pract20190432e9110210.3928/24748307-20190225-0131294310PMC6607763YdirinCSBHealth literacy and health-promoting behaviors among adults at risk for diabetes in a remote Filipino communityBelitung Nurs J202172889710.33546/bnj.129837469942BNJ-7-2-88PMC10353656MokAKhawKTLubenRWarehamNBrageSPhysical activity trajectories and mortality: population based cohort studyBMJ20190626365l232310.1136/bmj.l232331243014PMC6592407HayesAFRockwoodNJWrightAGHallquistMNMediation, moderation, and conditional process analysis: regression-based approaches for clinical researchThe Cambridge Handbook of Research Methods in Clinical Psychology2013Cambridge, MACambridge University Press396414FjellACronfalk SeigerBHermannMRongveAAßmusJKvingeLSeigerÅSkaugKBoströmAMFactors associated with self-rated health in a Norwegian population of older people participating in a preventive home visit program: a cross-sectional studyBMC Geriatr2020090420132310.1186/s12877-020-01733-23288755510.1186/s12877-020-01733-2PMC7650279WangYChenHEffects of mobile Internet use on the health of middle-aged and older adults: evidences from China health and retirement longitudinal studyBMC Public Health20240604241149010.1186/s12889-024-18916-w3883495910.1186/s12889-024-18916-wPMC11149348JinHLiLQianXZengYCan rural e-commerce service centers improve farmers’ subject well-being? A new practice of ‘internet plus rural public services’ from ChinaInt Food Agribus Manag Rev202012022356819610.22434/ifamr2019.0217ZeadallySSiddiquiFBaigZIbrahimASmart healthcare: challenges and potential solutions using internet of things (IoT) and big data analyticsPSU Res Rev20191018421496810.1108/prr-08-2019-0027JeonGSChoiKPurposes of internet use and its impacts on physical and psychological health of Korean older adultsHealthcare (Basel)2024011812224410.3390/healthcare1202024438255131healthcare12020244PMC10815879PekkalaKvan ZoonenWWork-related social media use: the mediating role of social media communication self-efficacyEur Manag J202202401677610.1016/j.emj.2021.03.004JaffeJMMedia interactivity and self-efficacy: an examination of hypermedia first aid instructionJ Health Commun199711242355210.1080/108107397127572JiangMMXiaoYWLiaoZLPathways of media contact to health literacy in middle-aged and older people: the chain mediation effect of perceived social support and self-efficacyJ Multidiscip Healthc2024171112110.2147/JMDH.S44822338205129448223PMC10778256WangYYasminFAkbarAImpact of the internet on English language learning among university students: mediating role of academic self-efficacyFront Psychol202314118418510.3389/fpsyg.2023.118418537868593PMC10588668JiMDengDYangXInfluence of internet usage on physical activity participation among Chinese residents: evidence from 2017 China General Social SurveyFront Public Health202412129369810.3389/fpubh.2024.129369838873316PMC11169592BoothAONowsonCAMattersHEvaluation of an interactive, internet-based weight loss program: a pilot studyHealth Educ Res2008062333718110.1093/her/cyn00718349032cyn007TomitaMRTsaiBMFisherNMKumarNAWildingGENaughtonBJImproving adherence to exercise in patients with heart failure through internet-based self-managementJ Am Geriatr Soc20081056101981310.1111/j.1532-5415.2008.01865.x19054217JGS1865RudolfKBiallasBDejongheLAGriebenCRückelLMSchallerAStassenGPfaffHFroböseIInfluence of health literacy on the physical activity of working adults: a cross-sectional analysis of the TRISEARCH trialInt J Environ Res Public Health201912061624494810.3390/ijerph1624494831817592ijerph16244948PMC6950634DishmanRKSallisJFOrensteinDRThe determinants of physical activity and exercisePublic Health Rep19851002158713920714PMC1424729ZhouJYangCYuJZhaoXWuJLiuZLiJGuYThe influence of social support on leisure-time physical activity of the elderly in the Chinese village of FuwenHealthcare (Basel)202308031115219310.3390/healthcare1115219337570433healthcare11152193PMC10418849ShenYLuCWangBThe impact of internet use on physical activity among Chinese older adults: the mediating role of social supportFront Public Health202513149218810.3389/fpubh.2025.149218839916702PMC11801014WarnerLMZiegelmannJPSchüzBWurmSSchwarzerRSynergistic effect of social support and self-efficacy on physical exercise in older adultsJ Aging Phys Act2011071932496110.1123/japa.19.3.24921727305DengCFengRKongLSquare dance the key factor of the elevating prevalence of physical activity in ChinaIran J Public Health20191048101920131850272PMC6908920LiuZKempermanATimmermansHSocial-ecological correlates of older adults’ outdoor activity patternsJ Transp Health2020031610084010.1016/j.jth.2020.100840BidmonSTerlutterRGender differences in searching for health information on the internet and the virtual patient-physician relationship in Germany: exploratory results on how men and women differ and whyJ Med Internet Res20150622176e15610.2196/jmir.412726099325v17i6e156PMC4526954ZhangJLiMDigital technology access, labor market behavior, and income inequality in rural ChinaHeliyon202407301014e3352810.1016/j.heliyon.2024.e3352839149026S2405-8440(24)09559-8PMC11325672YangDZhengGWangHLiMEducation, income, and happiness: evidence from ChinaFront Public Health20224121085532710.3389/fpubh.2022.85532735493390PMC9039002TangCWuXChenXPanBYangXExamining income-related inequality in health literacy and health-information seeking among urban population in ChinaBMC Public Health2019022119122110.1186/s12889-019-6538-23079188210.1186/s12889-019-6538-2PMC6385413MaitaKCManiaciMJHaiderCRAvilaFRTorres-GuzmanRABornaSLundeJJCoffeyJDDemaerschalkBMForteAJThe impact of digital health solutions on bridging the health care gap in rural areas: a scoping reviewPerm J202409162831304310.7812/TPP/23.13439135461PMC11404635LövdénMFratiglioniLGlymourMMLindenbergerUTucker-DrobEMEducation and cognitive functioning across the life spanPsychol Sci Public Interest2020081021164110.1177/152910062092057632772803PMC7425377Psychology and behavior investigation of Chinese residentsX-Mol2024-04-23https://www.x-mol.com/groups/pbicr