Digital health literacy among urban community-dwelling older adults in China: Current status and determinants

Abstract

Objective

To investigate the current status and influencing factors of digital health literacy (DHL) among older adults in China, providing evidence to guide interventions for improving DHL in this population.

Methods

A cross-sectional survey was conducted in November 2024 in Xuzhou using a multistage stratified cluster sampling method. A total of 1,005 valid questionnaires were collected. The survey included demographic information, the Chinese version of the DHL scale, self-rated health status, and activities of daily living. Random forest modeling and binary logistic regression were employed to identify predictors of DHL.

Results

A total of 1,005 adults aged 60 years and older were recruited; the mean DHL score was 24.75 ± 11.74. Only 40.6% of the participants (n = 408) reached the adequate level. High DHL was observed among older adults with smartphone use proficiency (β = 2.684, p < 0.001), higher educational attainment (β = 0.890, p = 0.013), and very good self-rated health status (β = 1.628, p = 0.040). In contrast, low DHL was associated with the absence of daily use of digital health services (β = −0.996, p < 0.001) and a lack of household Internet access (β = −2.220, p < 0.001).

Conclusions

The results revealed that DHL among older adults is relatively low, primarily due to a substantial gap between the rapid expansion of Internet services and limited digital competencies. Targeted interventions, including enhancing household Internet access, implementing age-friendly DHL training, and promoting supportive and inclusive community-based activities, are recommended.

Keywords

digital health literacy older adults internet access logistic models random forest model

1. Introduction

China is home to the largest population of older adults in the world and is undergoing a rapid aging process.^1,2 According to data from the National Bureau of Statistics, as of the end of 2023, the number of people aged 60 years and above in China is 297 million (21.1% of the population), out of whom 217 million are aged 65 years and above (15.4%).³ Against this background, China has put forward a national strategy for healthy aging. In the promotion of healthy aging, the health literacy of older adults has emerged as a critical factor and a key determinant that influences public health outcomes and healthcare costs,⁴ thus making it a core indicator for achieving healthy aging. China’s 14th Five-Year Plan for Healthy Aging explicitly calls for improvement in health literacy among older adults. In recent years, driven by a global wave of digital innovation, healthcare delivery and management have shifted toward digital platforms—telemedicine, electronic health records, mobile health applications, and online appointment systems have become widespread.⁵ For this reason, traditional notions of health literacy have become insufficient for fully addressing the requirements for information acquisition and service use within digital environments. Against this background, digital health literacy (DHL), as an extension of health literacy into digital contexts, has increasingly become a core competency that enables the effective use of digital health resources and has become especially important for older adults. DHL refers to an individual’s ability to search for, locate, understand, and appraise health information from electronic sources and to apply the acquired knowledge to health-related problems.⁶ On this basis, DHL is more closely linked to healthy aging. In digital contexts, promoting the inclusion of older adults in digital societies is a crucial means of achieving healthy aging; moreover, improving older adults’ DHL is a key objective of the national strategy to proactively address population aging.⁷

Globally, policy attention to improving of health literacy and DHL among older adults has been steadily increasing. In the Global Health Literacy Action Plan, the World Health Organization (WHO) explicitly calls for strengthening older adults’ capacities to access, understand, and apply health information in digital and nondigital settings.⁸ In China, policy documents such as the Healthy China 2030 Blueprint and the National Medium- and Long-Term Plan for Actively Responding to Population Aging emphasize improving health literacy among older adults, strengthening their self-health management capacities, and promoting the age-friendly development of health information service systems.^9,10 These policy orientations indicate that accurately understanding older adults’ DHL and its determinants is a prerequisite for promoting healthy aging and a core strategy for improving equity in digital health services.

With the rapid development of science and technology, China has made substantial progress in the construction of Internet infrastructure. The 55th Statistical Report on China’s Internet Development demonstrates that the number of Internet users aged 60 years and above increased from 7.3 million in 2009 to 157.25 million in 2024.¹¹ In addition, relevant studies indicate that 60.61% of adults and 64.69% of minors in China possess at least basic digital literacy and skills.¹² Despite significant advances in digital health and the increasingly frequent use of digital resources among older adults,¹³ the digital divide within this population remains substantial. Older adults generally exhibit lower levels of DHL compared with younger groups due to age-related declines in cognitive, sensory, and motor functions,¹⁴ combined with limited exposure to technology and low levels of confidence in operating digital platforms.^15,16 At the same time, older adults are frequently identified on social media platforms as a vulnerable group with respect to misinformation and fraudulent activities and are more likely to become victims of online scams.¹⁷ They face substantial barriers in searching for, understanding, and applying online health information. These challenges not only limit their effective use of digital health services but may also further widen the digital divide in the field of digital health.

Existing research indicates that studies on DHL among older adults in China remain at an early stage, while overall levels of DHL have been partially established.¹⁸ Several studies have examined the DHL levels of older adults living in cities such as Beijing,¹⁹ Sichuan,²⁰ and Henan.²¹ Meanwhile, domestic and international studies on older adults’ DHL have mainly focused on measurement and status assessment, associated factors, and existing challenges. A substantial portion of the existing literature focuses primarily on the sociodemographic characteristics and health-related conditions of study samples.^22,23 In contrast, characteristics of Internet use, particularly smartphone proficiency and the frequency with which individuals utilize digital health services, have received comparatively limited scholarly attention. Scholars who analyzed influencing mechanisms have generally utilized univariate and multivariate linear or nonlinear models, which exhibit limited capacity in identifying the relative importance of different factors.

To address this limitation and provide a more comprehensive analytical perspective, the present study interprets the determinants of DHL through the framework of the Three-level Digital Divide Theory. The concept of the “digital divide” was first introduced in 1999 by the U.S. National Telecommunications and Information Administration (NTIA), which defined it as the disparity between individuals who are “information rich” and those who are “information poor.” Van Dijk and Riggins subsequently conceptualized this disparity as differences in physical access to Information and Communication Technologies (ICTs), a dimension now recognized as the first-level digital divide, or the “access divide.”²⁴ As Internet access expanded and became more widely available, scholarly attention shifted toward disparities in how individuals used digital technologies, giving rise to the notion of the second-level digital divide, or the “skills divide.” Further research then extended the framework to consider inequalities in the benefits derived from ICT use, resulting in the formulation of the third-level digital divide, or the “outcome divide.” Consequently, digital inequality is now commonly understood as comprising three interrelated dimensions: the access divide (differences in physical access to digital technologies), the skills divide (variations in technical proficiency and patterns of technology use), and the outcome divide (differences in the tangible benefits individuals obtain from digital engagement, including health-related outcomes).

To address this research gap, the current study recruited older adults from urban communities in Xuzhou City, Jiangsu Province, China. Its registered population aged 60 years and above is approximately 2.182 million, which accounts for 21.20% and is consistent with the national level of population aging.²⁵ To assess the DHL levels of older adults in Xuzhou and analyze the factors influencing DHL, this study applies a combined analytical approach, integrating random forest machine learning and binary logistic regression, to identify key determinants and their effect sizes, explore pathways among relevant factors, and preserve interpretability while accounting for multidimensional complexity. Unlike traditional regression models, which primarily estimate the magnitude and direction of effects, the random forest model is particularly suited to identifying and ranking the relative importance of multiple interrelated predictors. This capability enhances interpretability for prioritizing potential intervention targets. Binary logistic regression is subsequently employed to quantify effect directions and magnitudes, thereby ensuring statistical interpretability. The findings provide evidence to inform strategy optimization in future research.

2. Materials and methods

2.1. Study design and participants

This cross-sectional study was conducted in the urban communities of Xuzhou City in November 2024. To minimize selection bias, the participants were selected using a multistage stratified random cluster sampling method. First, based on the 2023 Statistical Yearbook of the Xuzhou Bureau of Statistics, the city’s seven urban administrative districts were classified into high, medium, and low economic levels based on gross domestic product, and one district was randomly selected from each stratum. Subsequently, 2–4 communities were randomly selected from each selected district, which resulted in a total of 10 communities as sampling sites. All permanent residents aged 60 years and above in the selected communities were invited to participate in the questionnaire survey.

In developing country contexts, the WHO defines older adults as individuals aged 60 years and above. Accordingly, the age threshold adopted in this study aligns with internationally recognized standards. The inclusion criteria were: age ≥60 years, residence in an urban community of Xuzhou for at least one year, and voluntary provision of written informed consent. The exclusion criteria included visual or hearing impairments, cognitive disorders affecting communication, or mental conditions that precluded effective communication. The minimum required sample size for this cross-sectional study was calculated using the formula: N = [Z²_1-α/2P(1-P)]/ε². With the level of significance (α) set to 0.05 (Z_1-α/2 = 1.96) and the margin of error (ε) specified as 0.02, the expected prevalence (P) was set at 0.11.²⁶ This estimate was derived from empirical evidence indicating that 11.1% of older adults in China possess adequate DHL. Based on these parameters, the minimum required sample size was calculated as 940 participants. To ensure adequate statistical power and account for potential invalid responses, a total of 1,050 questionnaires were distributed. Of these, 1,005 questionnaires were deemed valid, yielding a valid response rate of 95.71%.

To reduce information bias, investigators who had received standardized training administered the survey. The investigators explained the study objectives to the participants and conducted face-to-face questionnaire interviews after obtaining informed consent. To address potential issues related to missing data, a rigorous data-cleaning protocol was implemented. Questionnaires with incomplete responses (defined as more than 10% missing items) or substantial logical inconsistencies were excluded from the analysis (n = 45). The remaining 1,005 questionnaires were fully completed and contained no missing values for the core study variables. The Ethics Committee of Xuzhou Medical University approved the study (approval number: XZHMU-2021037).

2.2. Measurements

2.2.1. General characteristics of the study participants

The general characteristics of the participants included gender, age, marital status, educational attainment, household registration, number of children, monthly disposable income, occupation during employment, status of chronic diseases, regular medications, employment status, caring for grandchildren, household Internet access, frequency of digital health service use, life satisfaction, and smartphone use proficiency. Variables were selected and organized based on the theory of the three levels of the digital divide: (1) the “access divide,” including household Internet access, monthly disposable income, participation in pension insurance, and household registration; (2) the “skills divide,” including smartphone use proficiency, frequency of digital health service use, educational attainment, age, gender, marital status, occupation during employment, employment status, number of children, and caring for grandchildren; and (3) the “outcome divide,” including self-rated health status, status of chronic diseases, regular medications, activities of daily living (ADL), and life satisfaction.

2.2.2. Digital health literacy

This study employed the widely used eHealth Literacy Scale (eHEALS). The original English version was developed by Norman and H. Skinner,²⁷ while the Chinese version was translated by Guo et al.²⁸ Items were rated using a five-point Likert-type scale (1 = Strongly disagree, 5 = Strongly agree), with total scores ranging from 8 to 40; higher scores indicated higher DHL levels. Based on the total score, the participants were classified into two groups: adequate (≥32) and inadequate (<32).²⁹ The scale comprises eight items covering three dimensions: application of online health information and services; appraisal ability; and decision-making ability. In this study, the eHEALS demonstrated excellent reliability, with Cronbach’s α = 0.958.

2.2.3. Self-rated health status

Self-rated health status can serve as an indicator of physical health among older adults. Previous studies demonstrated a significant association between older adults’ physical health and self-rated health, thus supporting self-assessment as a reliable method for measuring functional health status in daily life.³⁰ Self-rated health status was assessed using the question, “How would you rate your overall health?” This item was rated using a five-point Likert-type scale, ranging from 1 = Very poor to 5 = Very good. Higher scores indicate higher levels of self-rated health.

2.2.4. Activities of daily living

The ability to perform ADL was assessed using the modified ADL scale. Originally proposed by Katz,³¹ this scale measures basic daily activities among older adults, including bathing, dressing, going to toilet, transferring, continence, and feeding, covering six domains in total. In this study, ADL was dichotomized: individuals who required partial or complete assistance in at least one of the abovementioned items were classified as ADL dependent. Those who required no assistance across all six items were classified as ADL independent. In this study, the ADL scale demonstrated good reliability, with Cronbach’s α = 0.826.

2.3. Statistical analysis

A database was created using EpiData 3.1 with double data entry for verification, and data were processed and analyzed using SPSS 25.0 and R 4.4.1, respectively. Continuous variables were summarized as mean ± standard deviation, while categorical variables were described using frequencies and percentages. Univariate analyses were performed using the chi-square test, and multivariate analyses were conducted using logistic regression models. To control for potential confounding effects, the model was adjusted for demographic and socioeconomic factors, including age, gender, and educational attainment, in order to identify the independent predictors of DHL. The level of statistical significance was set at α = 0.05. In addition, a random forest classification model was employed to rank the factors associated with adequate DHL. Model performance was evaluated using the area under the receiver operating characteristic curve (“AUC”).

3. Results

3.1. General characteristics of the study participants

Out of the 1,005 older residents, 472 (47%) were men, and 533 (53%) were women. The age distribution was as follows: 665 (66.2%), 276 (27.5%), and 64 (6.4%) participants were aged 60–69, 70–79, and 80 years or above, respectively. In terms of educational attainment, 256 (25.5%), 384 (38.2%), 244 (24.3%), and 121 (12%) participants had primary school education or below, junior high school, senior high school, and a college degree or above, respectively. Regarding household registration, 432 (43%) and 573 (57%) participants had rural and urban hukou, respectively. For marital status, 881 (87.7%) and 124 (12.3%) participants were married and unmarried, respectively. The number of children was distributed as follows: 17 (1.7%), 328 (32.6%), 412 (41%), and 248 (24.7%) participants had no children, one child, two children, and three or more children, respectively. Monthly disposable income was less than 1,000 CNY for 75 (7.5%) participants, between 1,000 and 2,999 CNY for 371 (36.9%), between 3,000 and 4,999 CNY for 284 (28.3%), and 5,000 CNY and above for 275 (27.4%). Regarding occupation during employment, 284 (28.3%), 373 (37.1%), 196 (19.5%), 63 (6.3%), and 89 (8.9%) participants were farmers, enterprise employees, administrative or public institution staff, self-employed, and freelancers, respectively. Table 1 presents detailed information.

Table 1.

Results of population distribution and univariate analysis.

Variables	DHL: inadequate (n = 597)	DHL: adequate (n = 408)	χ ²	p
Gender
Male	275 (58.3)	197 (41.7)	0.480	0.488
Female	322 (60.4)	211 (39.6)
Age (years)
60∼69	354 (53.2)	311 (46.8)	37.871	<0.001
70∼79	188 (68.1)	88 (31.9)
≥80	55 (85.9)	9 (14.1)
Marital status
Married	498 (56.5)	383 (43.5)	24.496	<0.001
Unmarried	99 (79.8)	25 (20.2)
Educational attainment
Primary school or below	204 (79.7)	52 (20.3)
Junior high school	231 (60.2)	153 (39.8)	86.185	<0.001
Senior high school	121 (49.6)	123 (50.4)
College degree or above	41 (33.9)	80 (66.1)
Household registration
Rural	313 (72.5)	119 (27.5)	53.513	<0.001
Urban	284 (49.6)	289 (50.4)
Number of children
0	9 (52.9)	8 (47.1)
1	158 (48.2)	170 (51.8)	42.476	<0.001
2	244 (59.2)	168 (40.8)
≥3	186 (75)	62 (25)
Monthly disposable income
<1,000	62 (82.7)	13 (17.3)
1,000∼2,999	248 (66.8)	123 (33.2)	53.744	<0.001
3,000∼4,999	167 (58.8)	117 (41.2)
≥5000	120 (43.6)	155 (56.4)
Occupation during employment
Farmer	221 (77.8)	63 (22.2)
Enterprise employee	188 (50.4)	185 (49.6)
Administrative or public institution staff	90 (45.9)	106 (54.1)	71.019	<0.001
Self-employed	45 (71.4)	18 (28.6)
Freelancer	53 (59.6)	36 (40.4)
Hypertension
Yes	135 (48.2)	145 (51.8)	20.149	<0.001
No	462 (63.7)	263 (36.3)
Emphysema
Yes	24 (80)	6 (20)	5.440	0.020
No	573 (58.8)	402 (41.2)
Tuberculosis
Yes	21 (87.5)	3 (12.5)	8.049	0.003
No	576 (58.7)	405 (41.3)
Chronic bronchitis
Yes	8 (100)	0 (0)	3.945	0.047
No	589 (59.1)	408 (40.9)
Nephritis
Yes	22 (81.5)	5 (18.5)	5.608	0.018
No	575 (58.8)	403 (41.2)
Lumbar disc herniation
Yes	29 (46)	34 (54)	4.963	0.026
No	568 (60.3)	374 (39.7)
Regular use of antihypertensive drugs
Yes	133 (46.5)	153 (53.5)	27.583	<0.001
No	464 (64.5)	255 (35.5)
Regular use of hypoglycemic drugs
Yes	115 (77.7)	33 (22.3)	24.101	<0.001
No	482 (77.6)	375 (22.4)
Regular use of anti-inflammatory drugs
Yes	67 (62.6)	40 (37.4)	0.513	0.474
No	530 (59)	368 (41)
Regular use of analgesics
Yes	35 (79.6)	9 (20.4)	7.741	0.005
No	562 (58.5)	399 (41.5)
Regular use of antitumor drugs
Yes	8 (80)	2 (20)	1.777	0.183
No	589 (59.2)	406 (40.8)
Regular use of psychotropic drugs
Yes	27 (90)	3 (10)	12.004	0.001
No	570 (58.5)	405 (41.5)
Participation in pension insurance
Yes	396 (53.1)	350 (46.9)	47.942	<0.001
No	201 (77.6)	58 (22.4)
ADL
Independent	411 (53.7)	355 (46.3)	44.124	<0.001
Dependent	186 (77.8)	53 (22.2)

3.2. eHEALS acore

The total eHEALS score among the participants was 24.75 ± 11.74. A total of 408 participants (40.6%) reached the cut-off score of ≥32, which indicates adequate DHL. Among the three dimensions, Items 6–8 obtained the lowest scores. Table 2 presents the detailed results.

Table 2.

Scores of eHEALS of the older adults.

Items	Scores
Application competence for online health information and services
1. I know how to search for useful health information on the Internet.	3.20 ± 1.56
2. I know how to use the Internet to address my health-related questions.	3.15 ± 1.53
3. I know what types of health information are available online.	3.15 ± 1.54
4. I know where to find useful health information on the Internet.	3.15 ± 1.55
5. I know how to use online health information to help myself.	3.13 ± 1.53
Assessment competence for online health information and services
6. I have the skills to evaluate the quality of online health information.	2.96 ± 1.55
7. I can distinguish between high- and low-quality health information on the Internet.	2.99 ± 1.56
Decision-making competence for online health information and services.
8. I am confident in using online information to make health-related decisions.	3.02 ± 1.58

3.3. Group comparisons

Chi-square analyses were conducted, and the results demonstrated that statistically significant differences (p < 0.05) were observed across age; marital status; level of education; household registration status; number of children; monthly disposable income; occupation during employment; diagnosis of hypertension, emphysema, tuberculosis, chronic bronchitis, nephritis, or lumbar disc herniation; regular use of antihypertensive and hypoglycemic drugs, analgesics, and psychotropic medications; participation in pension insurance; household Internet access; frequency of digital health service use; self-rated health status; life satisfaction; ADL; and smartphone use proficiency (Table 1).

3.4. Predictors of DHL

3.4.1. Importance ranking of study variables

Using adequacy of DHL as the dependent variable (inadequate = 0, adequate = 1), variables that reached statistical significance in the univariate analyses were entered into a random forest model. The tuning parameters of the random forest model included the number of variables randomly sampled at each split (“mtry”) and the minimum node size. For the classification task, mtry was set to the square root of the total number of predictors, and the minimum node size was set to 1. The number of trees was fixed at 1,000. The random forest model demonstrated acceptable discriminatory performance, with an AUC of 0.80 (Figure 1). The out-of-bag (OOB) error rate reached a minimum of approximately 0.27 when 14 variables were included, corresponding to a classification accuracy of 73% (Figure 2). The top 14 variables ranked by the Gini index were, in descending order: smartphone use proficiency, frequency of digital health service use, household Internet access, level of education, regular use of antihypertensive drugs, self-rated health status, hypertension, marital status, number of children, participation in pension insurance, regular use of analgesics, chronic bronchitis, monthly disposable income, and occupation during employment (Figure 3).

Figure 1.

Receiver operating characteristics (ROC) curve of the random forest model.

Figure 2.

Out-of-bag (OOB) classification error rate of the random forest model with increasing numbers of variables.

Figure 3.

Importance ranking of study variables.

3.4.2. Binary logistic regression analysis of DHL among older adults

Using adequacy of DHL as the dependent variable, binary logistic regression analysis was performed on the 14 variables ranked as most important in the random forest model. The model demonstrated good fit (Hosmer–Lemeshow test: χ² = 2.038, p = 0.980) and strong explanatory power (Nagelkerke R² = 0.448). Multicollinearity diagnostics further confirmed that all variance inflation factor (“VIF”) values were within an acceptable range (1.01–4.12). The results indicated that smartphone use proficiency, frequency of digital health service use, and household Internet access were significantly associated with DHL among older adults. Compared with those who were not proficient in smartphone use, older adults with moderate (OR = 8.219) and high (OR = 14.646) levels of proficiency were significantly more likely to possess adequate DHL. Older adults who used digital health services daily were more likely to exhibit adequate DHL compared with those who did not use such services on a daily basis (OR = 0.369). Older adults without household Internet access displayed an approximately 89.1% lower likelihood of having adequate DHL compared with those with Internet access at home (OR = 0.109).

In addition, educational attainment and self-rated health status were significantly associated with DHL. Compared with older adults with primary school education or below, those with junior high school (OR = 2.002), senior high school (OR = 2.154), and college degree or above (OR = 2.436) were more likely to exhibit adequate DHL. Older adults who rated their health as “very good” had a 5.092-fold higher likelihood of having adequate DHL than those who rated their health as “very poor” (Table 3).

Table 3.

Binary logistic regression analysis of digital health literacy among older adults.

Variables	β	Wald	P	OR(95%CI)
Skills Divide
Smartphone use proficiency (reference: not proficient)
Moderately proficient	2.106	47.670	<0.001	8.219 (4.520, 14.946)
Proficient	2.684	76.142	<0.001	14.646 (8.015, 26.765)
Frequency of digital health service use (reference: daily)
Not daily	−0.996	21.627	<0.001	0.369 (0.243, 0.562)
Educational attainment (reference: primary school or below)
Junior high school	0.694	8.272	0.004	2.002 (1.247, 3.212)
Senior high school	0.767	7.464	0.006	2.154 (1.242, 3.735)
College degree or above	0.890	6.199	0.013	2.436 (1.209, 4.909)
Access Divide
Household Internet access (reference: yes)
No	−2.220	17.247	<0.001	0.109 (0.038, 0.310)
Outcome Divide
Self-rated health status (reference: very poor)
Poor	0.480	0.361	0.548	1.616 (0.338, 7.724)
Fair	0.440	0.339	0.560	1.553 (0.353, 6.835)
Good	0.663	0.767	0.381	1.940 (0.440, 8.543)
Very good	1.628	4.219	0.040	5.092 (1.077, 24.062)

4. Discussion

The results demonstrated that older adults in Xuzhou exhibited low levels of DHL, and educational attainment, health status, household Internet access, smartphone use proficiency, and frequency of digital health service use were identified as major factors associated with DHL. Moreover, the mean DHL score among older adults reached 24.75, which was well below the adequacy threshold (≥32), thus indicating a need for greater attention to DHL in this population. Previous studies have consistently reported that DHL levels among older adults were significantly lower than those among younger populations,^32–34 further highlighting the urgent need to improve DHL among older adults. In particular, the lowest scores were observed for the items “I have the skills to evaluate the quality of online health information,” “I can distinguish between high-quality and low-quality health information on the Internet,” and “I am confident in using online information to make health-related decisions,” which reflect substantial difficulty in evaluating health information and decision-making among older adults. These findings are consistent with Tabak’s survey on eHealth literacy in Brazil.³⁵ Moreover, in contrast to the previous literature, the current study specifically emphasized the strength of the effect of each independent variable on the dependent variable, identifying key factors that significantly influence DHL among older adults (Figure 3).

The second-level digital divide (“skills divide”), which emphasizes digital competencies and usage patterns, emerged as the most critical dimension in this study. We categorized smartphone use proficiency and frequency of digital health service use as Internet use characteristics.³⁴ First, based on the results of the random forest model, Internet use characteristics ranked first in importance. Older adults with higher levels of smartphone use proficiency and more frequent use of digital health services exhibited higher levels of DHL, which is consistent with previous studies.^34,36,37 The current study also found a positive correlation between participants’ ability to use the Internet to address their health-related questions and DHL. Therefore, familiarity with and skills in Internet use influence their ability to effectively utilize the Internet for health-related purposes.³⁸ Moreover, the strong association between Internet use characteristics and DHL implies that, in China’s rapidly digitalizing healthcare system, such as Internet-based medical services, telemedicine, and electronic health insurance certificates. Internet use characteristics have become an indispensable determinant of access to health information. This finding corroborates that of Kong and Wang, who emphasized that practical digital skills, distinct from traditional demographic factors, serve as key indicators for predicting levels of information literacy among older adults.³⁹

Second, educational attainment ranked third in importance among the factors associated with DHL. Older adults with higher levels of educational attainment demonstrated greater DHL than those with lower levels of education.⁴⁰ The lower levels of DHL observed among older adults with low levels of education may be associated with multiple factors, including limited educational background and knowledge structure; insufficient capacity to acquire and process information; ²¹ poor digital skills and unfamiliarity with digital devices; and a lack of social support and resources.⁴¹ The combined effects of these factors are associated with lower levels of performance among older adults with low levels of education in accessing, understanding, evaluating, and applying digital health information compared with those with higher levels of education.

Among the factors that influence older adults’ DHL, household Internet access ranked second in importance, representing the first-level digital divide (“access divide”). The lack of a stable Internet connection is negatively correlated with older adults’ DHL. Internet access plays a key role in enhancing DHL skills and facilitating access to and interaction with various digital health tools, applications, and websites.⁴² According to official data, China’s information technology infrastructure is robust, with nearly 4.65 million 5G base stations and the number of users of China’s three major telecom operators and broadband networks reaching approximately 1.82 billion.⁴³ However, the current study found that 9% of older adults still lacked Internet access at home. A potential reason is that, without access to Internet infrastructure, even individuals with motivation and skills face practical barriers to participation. Furthermore, research indicates that older adults typically lack knowledge of accessing and using information in digital environments to facilitate effective health-related decision-making due to declining physical abilities, lack of digital skills, and low levels of Internet literacy.⁴⁴

The third level of the digital divide (“outcome divide”) is reflected in the significant association between health status and DHL. Health status ranked fourth in importance among the factors influencing DHL. Older adults with higher levels of self-rated health status exhibited higher levels of DHL than those with lower self-rated health.^45,46 A potential underlying reason is that older adults with poorer self-rated health typically face multiple concurrent challenges, including multimorbidity, declining physical function, increased psychological risks, and restricted access to health services, which correlate with significantly lower levels of DHL. In contrast, those with very good self-rated health often demonstrate a stronger awareness of health maintenance and promotion and exhibit a greater capacity to seek and use online health resources, which is associated with higher levels of DHL.^47,48

In the current study, the extremely low scores for DHL clearly revealed an urgent need to narrow the digital divide across all three levels. The WHO noted that improving DHL “can lead to better health outcomes and reduce inequalities in healthcare.” Taken together with the WHO’s observation that older adults frequently encounter difficulty in accessing digital solutions, the current findings further highlight the urgency of implementing targeted policy actions. Specifically, an urgent need exists to develop strategies for enhancing older adults’ access to, skills in, and confidence in using digital health resources.

First, age-friendly DHL training programs should be implemented. These programs should specifically target older adults with low levels of education, limited smartphone use proficiency, and difficulty in using digital health services. Evidence from studies conducted in India proposes that such training programs can effectively alleviate the digital divide, promote digital equity, and are associated with improved health outcomes for older adults.^49,50 The relevant authorities should provide basic operational training, guiding older adults on the safe use of the Internet and health applications. In addition, electronic health tools should be optimized according to the actual capabilities of older adults, including the use of large and clear fonts, simplified navigation, and assistive functions (e.g., voice commands, text-to-speech, and high-contrast interfaces).^20,51 Second, household Internet access among older adults should be improved. Governments should increase investment in Internet infrastructure for groups unable to afford installation (e.g., empty-nest older adults and low-income households), including broadband networks and mobile communication base stations, to ensure convenient and stable Internet access for older adults.⁵¹ Finally, older adults should be encouraged to actively participate in community activities. Previous studies demonstrated that the use of digital health applications and online engagement is associated with improved self-rated health among older adults by enhancing social participation and strengthening social support.^52,53 Moreover, digital health services can also reduce loneliness and improve mental health, thereby further enhancing subjective health perceptions among older adults.⁵⁴ The present study has identified pathways for cultivating digital literacy and digital skills among older adults. Apart from this study’s contribution, the relevant departments should actively promote both online and offline social activities to effectively alleviate negative emotions, improve health status, and enhance social participation and social support. Furthermore, policymakers should optimize institutional safeguards within healthcare services and pension systems, while simultaneously encouraging community members, friends, and relatives to deliver social support through multiple channels and in diverse forms.^55,56

In summary, the findings, together with evidence from domestic and international research, point to a common conclusion: in the context of rapid digitalization, insufficient DHL among older adults has become a major barrier to health equity and healthy aging. Specifically, despite China’s world-leading digital infrastructure, a pronounced capability gap persists between the rapid expansion of technological supply and the limited digital competencies of older adults, which results in persistent barriers to accessing, understanding, and using digital health resources. Therefore, enhancing DHL among older adults is not only an issue of technological adaptation but also a systemic requirement at the levels of public health policy and social governance. To effectively address this gap, multidimensional interventions are required, including improvement in household Internet access among older adults, institutionalization of age-friendly DHL training programs, and the promotion of supportive and inclusive community environments, thereby achieving synergistic gains in capacity, confidence, and social connectedness.

This study has several limitations. First, the cross-sectional nature of the data limits causal inference, and the observed associations may reflect bidirectional relationships. Second, the use of self-reported data may have introduced recall bias or social desirability bias. Third, although the random forest approach helps mitigate overfitting and variable redundancy, variable importance scores may still be influenced by measurement scales or the number of categories. Fourth, as the study was conducted in a single urban city in China, regional variations in socioeconomic development, digital infrastructure, healthcare accessibility, and cultural context may limit the external validity of the findings. Therefore, caution should be exercised when generalizing the results to rural areas, other regions of China, or international populations. Future multicenter, longitudinal studies with more diverse samples are warranted to confirm and extend these findings.

5. Conclusion

This study found that levels of DHL among older adults in Xuzhou, China, are relatively low, with significant disparities observed across the three levels of the digital divide. Greater smartphone proficiency, more frequent use of digital health services, improved household Internet access, higher levels of education, and better self-rated health were identified as significant predictors associated with adequate DHL. Given these findings, the study recommends that relevant authorities implement targeted interventions to improve household Internet access among older adults, enhance their skills in using digital services, and actively promote community-based activities, thereby improving levels of DHL in this population. By integrating the random forest model with binary logistic regression analysis under the Three-level Digital Divide Theory, this study not only identified the factors associated with DHL among older adults but also quantified the relative degrees of the key determinants, thereby contributing to an in-depth understanding of complex health-related behaviors.

Footnotes

Acknowledgments

The authors express their sincere gratitude to all contributors involved in this study. Each author made substantial contributions to the conception and design of the study, data collection and analysis, and the drafting and critical revision of the manuscript. The collaboration, insightful discussions, and collective commitment of all authors were essential to the completion of this work. During the preparation of this manuscript, the authors used the ChatGPT tool (“ChatGPT”), developed by OpenAI, for language editing and proofreading in order to improve the clarity and flow of the text. Following the use of this tool, the authors carefully reviewed and edited the content and accept full responsibility for the final version of the manuscript.

ORCID iD

Xiliang Li

Ethical considerations

The study involving human participants was approved by the Ethics Committee of Xuzhou Medical University (XZHMU-2021037). The study was conducted in accordance with the local legislation and institutional requirements.

Consent to participate

The participants provided their written informed consent to participate in this study.

Funding

The authors declare that financial support was received for the research and/or publication of this article. The study was supported by the Philosophy and Social Science Major Project of the Jiangsu Province Educational Department (Grant No 2021SJZDA136).

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and publication of this article.

Contributorship

Xiliang Li proposed the research topic and direction, and was responsible for data collection and organization, study conception and design, and drafting the initial manuscript. Yichen Li and Wenhao Huang contributed to manuscript revision. Hongan Zhang proposed the research topic and conducted the study design and feasibility analysis. Zhaohui Qin and Yan Xu were responsible for quality control and critical review of the manuscript. All authors reviewed and approved the final manuscript.

Guarantor

Zhaohui Qin and Yan Xu.

References

Jiang

Song

. Health consequences of online social capital among middle-aged and older adults in China. Appl Res Qual Life 2022; 17: 2277–2297. https://doi.org/10.1007/s11482-021-10033-9

Chen

Giles

Yao

, et al. The path to healthy ageing in China: a Peking University–Lancet Commission. Lancet 2022; 400: 1967–2006. https://doi.org/10.1016/S0140-6736(22)01546-X

Wang

. Population size slightly declines, achievements in high-quality population development. China Economic Net, 2024. https://www.ce.cn/xwzx/gnsz/gdxw/202401/18/t20240118_38870849.shtml (accessed 25 September 2025).

Nutbeam

Lloyd

. Understanding and responding to health literacy as a social determinant of health. Annu Rev Public Health 2021; 42: 159–173. https://doi.org/10.1146/annurev-publhealth-090419-102529

Van Kessel

Wong

BLH

Clemens

, et al. Digital health literacy as a super determinant of health: more than simply the sum of its parts. Internet Interv 2022; 27: 100500. https://doi.org/10.1016/j.invent.2022.100500

World Health Organization . Strengthening digital health literacy to empower people in the digital age. https://www.who.int/europe/news-room/events/item/2024/11/21/default-calendar/strengthening-digital-health-literacy-to-empower-people-in-the-digital-age (2024, accessed 20 September 2025).

Zhu

. Bridging the digital divide: dilemma and breakthrough of digital survival of the elderly groups. Inner Mongolia Soc Sci 2023; 44: 162–169.

World Health Organization . Global strategy on digital health 2020–2025. World Health Organization, 2021. https://www.who.int/publications/i/item/9789240020924 (accessed 25 September 2025).

State Council of the People’s Republic of China . Healthy China 2030 blueprint. https://www.gov.cn/zhengce/2016-10/25/content_5124174.htm (2016, accessed 25 September 2025).

10.

State Council of the People’s Republic of China . National medium- and long-term plan for actively responding to population aging. https://www.gov.cn/zhengce/2019-11/21/content_5454347.htm (2019, accessed 25 September 2025).

11.

China Internet Network Information Center . The 55th statistical report on internet development in China. https://www.cnnic.net.cn/n4/2025/0117/c88-11229.html (2025, accessed 25 September 2025).

12.

State Council of the People’s Republic of China . Survey report on the digital literacy and skills development level of the whole population. https://www.gov.cn/lianbo/bumen/202410/content_6983266.htm (2024, accessed 25 September 2025).

13.

Tennant

Stellefson

Dodd

, et al. eHealth literacy and Web 2.0 health information seeking behaviors among baby boomers and older adults. J Med Internet Res 2015; 17: e70. https://doi.org/10.2196/jmir.3992

14.

Liu

Zhang

, et al. Current situation and influencing factors of e-health literacy among rural older adults in Zhengzhou. Mod Prev Med 2020; 47: 283–309.

15.

Choi

DiNitto

. The digital divide among low-income homebound older adults: internet use patterns, eHealth literacy, and attitudes toward computer/Internet use. J Med Internet Res 2013; 15: e93. https://doi.org/10.2196/jmir.2645

16.

Watkins

Xie

. eHealth literacy interventions for older adults: a systematic review of the literature. J Med Internet Res 2014; 16: e225. https://doi.org/10.2196/jmir.3318

17.

Christiansen

Lindberg

Sanmartin Berglund

, et al. Using mobile health and the impact on health-related quality of life: perceptions of older adults with cognitive impairment. Int J Environ Res Public Health 2020; 17: 2650. https://doi.org/10.3390/ijerph17082650

18.

Wang

Luan

. Research progress on digital health literacy of older adults: a scoping review. Front Public Health 2022; 10: 906089. https://doi.org/10.3389/fpubh.2022.906089

19.

Cui

Zhang

, et al. Associations between digital skill, eHealth literacy, and frailty among older adults: evidence from China. J Am Med Dir Assoc 2024; 25: 105275. https://doi.org/10.1016/j.jamda.2024.105275

20.

Zhang

Yuan

Jiang

. Digital health literacy among older adults in China: a cross-sectional study on prevalence and influencing factors. Front Public Health 2025; 13: 1661177. https://doi.org/10.3389/fpubh.2025.1661177

21.

Xue

Wang

, et al. E-health literacy in stroke patients: latent profile analysis and influencing factors. J Adv Nurs 2025; 81: 1388–1398. https://doi.org/10.1111/jan.16351

22.

Bardus

Keriabian

Elbejjani

, et al. Assessing eHealth literacy among internet users in Lebanon: a cross-sectional study. Digit Health 2022; 8: 20552076221119336. https://doi.org/10.1177/20552076221119336

23.

Lee

Tak

. Factors associated with eHealth literacy focusing on digital literacy components: a cross-sectional study of middle-aged adults in South Korea. Digit Health 2022; 8: 20552076221102765. https://doi.org/10.1177/20552076221102765

24.

Van Dijk

. Digital divide research, achievements and shortcomings. Poetics 2006; 34: 221–235. https://doi.org/10.1016/j.poetic.2006.05.004

25.

Xuzhou released the 2023 annual report on elderly population information and aging development. Xuzhou Municipal People’s Government. https://www.xz.gov.cn/001/001001/20241011/c191852c-db86-48f2-b29f-7700e097780d.html (2024, accessed 25 September 2025).

26.

Cui

. Analysis of eHealth literacy and its influencing factors among the elderly (in Chinese). Chin J Dis Control Prev 2019; 23: 1318–1322.

27.

Norman

Skinner

. eHEALS: the eHealth literacy scale. J Med Internet Res 2006; 8: e507. https://doi.org/10.2196/jmir.8.4.e27

28.

Guo

Sun

, et al. Adaptation and evaluation of Chinese version of eHEALS and its usage among senior high school students. Chin J Health Educ 2013; 29: 106–108.

29.

Jiang

Wang

Leng

, et al. The level of electronic health literacy among older adults: a systematic review and meta-analysis. Arch Public Health 2024; 82: 204. https://doi.org/10.1186/s13690-024-01428-9

30.

Knapik

Brzęk

Famuła-Wąż

, et al. The relationship between physical fitness and health self-assessment in elderly. Medicine 2019; 98: e15984. https://doi.org/10.1097/MD.0000000000015984

31.

Katz

Ford

Moskowitz

, et al. Studies of illness in the aged: the index of ADL: a standardized measure of biological and psychosocial function. JAMA 1963; 185: 914–919. https://doi.org/10.1001/jama.1963.03060120024016

32.

Rueda-Medina

Gómez-Urquiza

Casas-Barragán

, et al. eHealth literacy, internet use and health-related behaviour among health sciences students: a cross sectional study. Contemp Nurse 2023; 59: 143–152. https://doi.org/10.1080/10376178.2023.2166547

33.

Dadaczynski

Okan

Messer

, et al. Digital health literacy and web-based information-seeking behaviors of university students in Germany during the COVID-19 pandemic: cross-sectional survey study. J Med Internet Res 2021; 23: e24097. https://doi.org/10.2196/24097

34.

Zhao

Huang

Cheng

, et al. Digital health literacy and associated factors among internet users from China: a cross-sectional study. BMC Public Health 2024; 24: 908. https://doi.org/10.1186/s12889-024-18324-0

35.

Tabak

Froner

Cardoso

DHR

, et al. Understanding digital health literacy in Brazil: findings from the eHEALS survey. Public Health 2025; 246: 105828. https://doi.org/10.1016/j.puhe.2025.105828

36.

Mengestie

Yilma

Beshir

, et al. eHealth literacy of medical and health science students and factors affecting eHealth literacy in an Ethiopian university: a cross-sectional study. Appl Clin Inform 2021; 12: 301–309. https://doi.org/10.1055/s-0041-1727154

37.

Arcury

Sandberg

Melius

, et al. Older adult internet use and eHealth literacy. J Appl Gerontol 2020; 39: 141–150. https://doi.org/10.1177/0733464818807468

38.

Xie

. Multi-level factors influencing eHealth literacy among Chinese older adults: a longitudinal study. J Adv Nurs 2025; 81: 5831–5844. https://doi.org/10.1111/jan.16702

39.

Kong

Wang

. Exploring the influential factors and improvement strategies for digital information literacy among the elderly: an analysis based on integrated learning algorithms. Digit Health 2024; 10: 20552076241286635. https://doi.org/10.1177/20552076241286635

40.

Estrela

Semedo

Roque

, et al. Sociodemographic determinants of digital health literacy: a systematic review and meta-analysis. Int J Med Inform 2023; 177: 105124. https://doi.org/10.1016/j.ijmedinf.2023.105124

41.

Carroll

Moorhead

Bond

, et al. Who uses mobile phone health apps and does use matter? A secondary data analytics approach. J Med Internet Res 2017; 19: e125. https://doi.org/10.2196/jmir.5604

42.

Chen

McGaughey

Zeltmann

, et al. How seniors in Taiwan use information technology: computer and cell phones. Int J Hum Comput Interact 2018; 34: 166–176. https://doi.org/10.1080/10447318.2017.1335445

43.

Xinhua . China home to nearly 4.65 million 5G base stations. The State Council of the People’s Republic of China. https://english.www.gov.cn/archive/statistics/202509/23/content_WS68d296f4c6d06a009cb724e8.html (2025, accessed 25 September 2025).

44.

Wilson

Heinsch

Betts

, et al. Barriers and facilitators to the use of e-health by older adults: a scoping review. BMC Public Health 2021; 21: 1556. https://doi.org/10.1186/s12889-021-11623-w

45.

Bíró

Vincze

Nagy-Pénzes

, et al. Investigation of the relationship of general and digital health literacy with various health-related outcomes. Front Public Health 2023; 11: 1229734. https://doi.org/10.3389/fpubh.2023.1229734

46.

Ning

Zheng

Liu

, et al. Profiles of digital health literacy among university students and its association with mental health during the COVID-19 pandemic: a latent profile analysis. Digit Health 2024; 10: 20552076231224596. https://doi.org/10.1177/20552076231224596

47.

Ghazi

Berner

Anderberg

, et al. The prevalence of eHealth literacy and its relationship with perceived health status and psychological distress during Covid-19: a cross-sectional study of older adults in Blekinge, Sweden. BMC Geriatr 2023; 23: 5. https://doi.org/10.1186/s12877-022-03723-y

48.

Liu

Zhao

, et al. Current status and influencing factors of digital health literacy among community-dwelling older adults in Southwest China: a cross-sectional study. BMC Public Health 2022; 22: 996. https://doi.org/10.1186/s12889-022-13378-4

49.

Bansal

Choudhary

. Fostering digital equity: evaluating impact of digital literacy training on internet outcomes in rural marginalised communities in India. Int J Lifelong Educ 2024; 43: 473–493. https://doi.org/10.1080/02601370.2024.2347327

50.

Choudhary

. From margins to mainstream: paving the way to digital equity through digital literacy training programs. Eval Program Plann 2025; 109: 102533. https://doi.org/10.1016/j.evalprogplan.2024.102533

51.

Shao

Gao

. The impact of digital literacy on the health behavior of rural older adults: evidence from China. BMC Public Health 2025; 25: 919. https://doi.org/10.1186/s12889-025-21964-5

52.

Wang

Shen

Cheng

, et al. The impact of informal social support on the health poverty vulnerability of the elderly in rural China: based on 2018 CHARLS data. BMC Health Serv Res 2022; 22: 1122. https://doi.org/10.1186/s12913-022-08468-3

53.

Yang

Niu

. The impact of internet use on health status among older adults in China: the mediating role of social support. Front Public Health 2023; 11: 1108096.

54.

Wilson

. Is it love or loneliness? Exploring the impact of everyday digital technology use on the wellbeing of older adults. Ageing Soc 2018; 38: 1307–1331. https://doi.org/10.1017/s0144686x16001537

55.

Sun

Zhao

, et al. Internet use and self-rated health among Chinese older adults: the role of social engagement and social support. Front Public Health 2025; 12: 1504940. https://doi.org/10.3389/fpubh.2024.1504940

56.

Yoo

. Empowering older adults: improving senior digital literacy. In: American Association for Adult and Continuing Education Conference, Online, October 27–30, 2020, pp. 281–287.