Latent profile analysis of digital health literacy among community-dwelling older adults and its influencing factors

Abstract

Introduction

Digital health advances universal health coverage, yet insufficient digital health literacy exacerbates the digital divide and health inequalities among older adults. Existing studies lack in-depth exploration of within-group digital health literacy heterogeneity among Chinese community-dwelling older adults. This study aimed to identify digital health literacy latent profiles and associated factors to inform targeted interventions.

Methods

A cross-sectional survey was conducted among 535 community-dwelling older adults (aged ≥60 years) from urban and suburban communities in China. Latent profile analysis was used to identify digital health literacy subgroups based on three core subdomains: Self-Perception (SP), Information Acquisition (IA), and Interactive Judgment (IJ). Multinomial logistic regression was applied to examine factors associated with profile membership.

Results

Three distinct digital health literacy profiles were identified: Low SP-IA-IJ (38.88%), Medium SP-IA & Low IJ (47.85%), and High SP-IA & Medium IJ (13.27%). Notably, Interactive Judgment (IJ) was a universal weak dimension across all groups. Older age, cognitive decline, depressive symptoms, and loneliness were significant risk factors; higher education, better economic status, favorable self-rated health, and stronger social interaction were protective factors, with differential associations across profiles.

Conclusions

Overall digital health literacy among Chinese community-dwelling older adults is generally low with significant within-group heterogeneity. Targeted stratified interventions, universal Interactive Judgment (IJ) capacity building, and age-friendly digital environment optimization can help narrow the digital divide and promote equitable access to digital health services for older adults. This study is limited by its cross-sectional design and regional urban/suburban sample, restricting causal inference and generalizability to rural or other regions.

Keywords

digital divide digital health literacy latent profile analysis older adults China

Introduction

Digital health, which integrates health resources via the internet, has emerged as a vital approach for achieving universal health coverage.¹ Digital health enhances health service accessibility,² promotes public engagement in health management, improves population health outcomes,^3–5 and offers particular value for populations with limited in-person health service access.⁶ Simultaneously, digital health is reshaping health information interactions, shifting individuals from passive health service recipients to active participants in their own health management.⁷ The ability of individuals to seek out, access, understand, and evaluate health information from digital channels, and to process and apply this information to address health-related problems, is defined as digital health literacy.⁸

While digital technologies have unlocked new opportunities for the digital health advancement, older adults, who are disproportionately affected by the digital divide, may face exacerbated health inequalities.⁹ The digital divide framework has evolved significantly over the past decade, from an initial focus on the first-order divide (inequalities in physical access to digital devices and internet infrastructure), to the second-order divide (inequities in digital use, skills, motivation, and usage purposes), and more recently to the third-order divide (inequalities in real-world health, social, and economic outcomes driven by gaps in digital access and skills).¹⁰ Under this framework, insufficient digital health literacy among older adults is a core driver of persistent digital divides in this population, directly shaping their ability to translate digital access into tangible health benefits.^11,12

Previous studies have shown that older adults face multiple barriers to digital health adoption, including difficulties with device operation, information navigation, on-screen text reading, and privacy protection.^13,14 Older adults with low digital health literacy are more likely to experience anxiety and distress when faced with unverified online health information,^15,16 and face higher risks of adverse health outcomes,¹⁷ while those with higher literacy show greater motivation to engage with digital health resources and exhibit more positive health beliefs and behaviors.^18,19 Digital health literacy is therefore a major barrier hindering older adults’ integration into digital society and their access to high-quality digital health services,^20,21 making it critical to identify its influencing factors and develop targeted interventions to advance healthy aging.^22–24

Numerous factors associated with older adults’ digital health literacy have been identified, including demographic characteristics, health-related factors, and psychosocial and functional factors.^25–31 However, most existing studies conceptualize older adults’ digital health literacy as a homogeneous trait across the older adult population, with limited in-depth exploration of its within-group heterogeneity. Exploring multi-dimensional individual subgroups to achieve nuanced analysis of this heterogeneity is critical for addressing digital divides among older adults.³² As a person-centered analytical approach that accounts for inter-individual heterogeneity,³³ latent profile analysis can identify distinct, internally homogeneous subgroups of older adults with similar multi-dimensional patterns of digital health literacy. This method facilitates a more precise characterization of the distribution of digital health literacy among older populations, which in turn informs the design of targeted, stratified interventions.

China is currently undergoing simultaneous rapid population aging and digital transformation.³⁴ With 264 million adults aged 60 and above (18.7% of the national population)³⁵ and a 44.5% internet penetration rate among this group (15.6% of national internet users),³⁶ as the Chinese government continues to promote digital transformation of community health services and digital social governance for older adult inclusion,^37,38 the digital divide among Chinese older adults has accordingly shifted to the second- and third-order divides. Correspondingly, research on digital health literacy heterogeneity among Chinese community-dwelling older adults remains scarce.³⁰ Existing studies using latent profile analysis in this field have primarily focused on the impact of demographic characteristics and community digital environment, as well as stepwise differences in overall literacy levels across latent profiles, and have paid insufficient attention to the role of individual psychosocial factors, and have not fully explored asynchronous variation patterns in multi-dimensional literacy across subgroups.³⁹

Therefore, grounded in the three-tiered digital divide framework, this study focuses on community-dwelling older adults in China as the study population. Using latent profile analysis, this study aims to identify latent profiles of their digital health literacy, examine multi-dimensional influencing factors associated with profile membership including psychosocial indicators, and explore potential associations linking these factors to profile membership. This work contributes to the characterization of the within-group heterogeneity of digital health literacy among older adults, and provides preliminary evidence to inform for the development of differentiated, targeted intervention programs, the reduction of the digital divide among Chinese older adults, as well as relevant policy formulation and grassroots community practice.

Methods

Participants and quality Control

A research project examining the health status and social lives of older adults was implemented across multiple communities in Chengdu and adjacent cities in Sichuan Province, China, from August 2022 to July 2025. The cross-sectional survey presented in this manuscript, a sub-study of this larger project, was conducted between May and July 2025. To recruit older adults with diverse living contexts, two established urban communities, two newly developed urban communities, and one suburban community were selected for inclusion. Participant inclusion criteria were as follows: (1) aged ≥ 60 years; (2) no severe physical or mental illness, and ability to use digital devices (including mobile phones, computers, and tablets) in daily life; (3) ability to understand the survey questions and written informed consent form, and voluntary willingness to participate.

The sample size was calculated using the formula for estimating a population mean: $n = {(\frac{z_{1 - \frac{α}{2}} \cdot σ}{δ})}^{2}$ , with a two-sided significance level of α = 0.05.⁴⁰ Based on a previous large-scale survey, the standard deviation of digital health literacy was σ = 7.67.⁴¹ Thus, with an allowable error δ = 1, the minimum required sample size was 226. Accounting for potential invalid questionnaires and invalid responses, the sample size was increased by 20%, yielding in a final minimum target sample size of 272.

Prior to survey administration, all research staff completed standardized training. All participants were fully informed of the survey’s purpose, procedures, and confidentiality protocols, and signed the written informed consent form voluntarily. A total of 562 older adults participated in the survey, with 535 valid questionnaires included in the final analysis, corresponding to a valid response rate of 95.2%.

Measures

Digital health literacy

Digital health literacy was assessed using the m-eHEALS scale, developed and validated by Wu et al. for Chinese populations.⁴¹ This scale is grounded in the theoretically and psychometrically validated eHEALS framework developed by Norman and Skinner,⁸ ensuring a consistent theoretical foundation and direct comparability of findings with related international studies. It is further optimized for mobile health (mHealth) use scenarios among Chinese populations, with concise, easy-to-understand items aligned with China’s current internet healthcare context and suitable for community-dwelling older adults. Notably, this scale was developed and validated in a large, nationally representative sample of over 3000 individuals across 23 provinces in China, and exhibits more robust psychometric properties than most comparable scales developed and validated in smaller, regionally limited samples.

The m-eHEALS includes 12 items across three dimensions: Self-Perception (SP), Information Acquisition (IA), and Interactive Judgment (IJ). Self-Perception (SP) includes 3 items measuring autonomy and competence in selecting digital health technologies for health support. Information Acquisition (IA) includes 5 items assessing the ability to obtain and understand information using digital health technologies. Interactive Judgment (IJ) includes 4 items evaluating the ability to communicate with others and conduct critical analysis using digital health technologies. Items are rated on a 5-point Likert scale, ranging from 1 (“strongly disagree”) to 5 (“strongly agree”). Total scores range from 12 to 60, with higher scores indicating higher levels of digital health literacy. The scale has demonstrated good reliability and validity, with a Cronbach’s α of 0.91 in the original validation study.⁴¹

The three dimensions of the m-eHEALS were selected as indicators for latent profile construction, as they capture the full core construct of digital health literacy within the three-tiered digital divide framework, and can fully reflect the multi-dimensional heterogeneity of digital health literacy among older adults. Based on existing literature, this study selected three categories of variables that have been widely verified to be associated with digital health literacy among older adults as predictors of latent profile membership, including demographic characteristics, health status, and psychosocial factors.

Sociodemographic characteristics

Sociodemographic variables included age (coded as 1=60-64, 2=65-69, 3=70-74, 4=75-79, 5=≥80), gender (1=male, 2=female), education level (1=primary school or below, 2=junior high, 3=senior high, 4=bachelor’s or above), marital/living status (1=cohabiting with spouse, 2=not cohabiting with spouse), economic status (1=very insufficient, 2=somewhat insufficient, 3=medium, 4=somewhat sufficient, 5=very sufficient), and chronic diseases (1=none, 2=present).

Self-rated health

Self-rated health, representing subjective health assessment, was measured on a 5-point scale from 1 (very unhealthy) to 5 (very healthy), with higher scores indicating better perceived health.

Independence in daily life

Independence in instrumental activities of daily living was assessed using the 6-item scale developed by Lawton and Brody.⁴² This tool assesses tasks including cooking, housekeeping, medication management, telephone use, shopping, and financial management. Each item is scored dichotomously (1=performs independently; 0=requires assistance/unable). A total score ≥1 indicates inability to maintain complete independence in daily life.

Cognitive function

Cognitive function was evaluated using the Chinese version of 8-item Ascertaining Dementia instrument, validated by Cai et al.⁴³ Participants reported perceived changes in eight cognitive domains using “changed,” “unchanged,” or “uncertain.” Responses were coded as 1 for “changed” and 0 for “unchanged/uncertain.” A total score ≥2 suggests cognitive decline.

Depression symptom

Depressive symptoms were assessed using the 11-item Chinese version of the Short Form Geriatric Depression Scale, validated by Tang et al.,⁴⁴ with “yes” or “no” responses (yes=1, no=0). Scores were summed across items (range 0-11), with higher scores indicating more severe symptoms.

Loneliness

Loneliness was measured using the 6-item Chinese version of University of California Los Angeles Loneliness Scale, adapted by Zhou et al.⁴⁵ It employs a 4-point frequency scale (never=1, rarely=2, sometimes=3, always=4), with total scores ranging from 6 to 24. Higher scores indicate greater loneliness.

Social interaction

Social interaction was assessed using the Index of Social Interaction, validated by Anme et al.^46,47 This 18-item scale evaluates five domains: active independent living, curiosity about society, communication with others, participation in society, and perceived safety within the surrounding society. Items have positive/negative choices scored 1 or 0 (e.g., “always/often”=1; “rarely/basically not”=0). Total scores range from 0 to 18, with higher scores indicating better social interaction.

Statistical analysis

Latent profile analysis (LPA) was conducted using Mplus 7.4, as this method is well-suited for the cross-sectional study design and continuous digital health literacy indicators, and enables objective determination of the optimal number of profiles using multiple standardized fit indices.³³ Model selection was guided by the Akaike Information Criterion (AIC), Bayesian Information Criterion (BIC), and sample-size adjusted BIC (aBIC), with lower values indicating better model fit. The Lo-Mendell-Rubin adjusted likelihood ratio test (LMR-LRT) and Bootstrap likelihood ratio test (BLRT) were used to compare model fit between k-profile and k-1-profile models, to determine the optimal number of profiles. Entropy, ranging from 0 to 1, was used to assess classification accuracy, with values closer to 1 indicating higher classification precision. The final model was required to have a minimum class proportion of at least 5% for all profiles. Descriptive analyses, chi-square tests, analysis of variance (ANOVA), and multinomial logistic regression were conducted using SPSS 26.0, with latent profile membership as the dependent variable. p < .05 was considered statistically significant.

Ethical approval and compliance

The study was approved by the ethical review committee of Sichuan Academy of Chinese Medicine Sciences (No. Ke202207-15-313). All research tools and questionnaires used in this study comply with academic publishing norms, with their original sources properly cited in the manuscript, and no copyright infringement is involved.

Results

Descriptive analysis of variables among older adults

All 535 participants completed the full survey. The sample was predominantly composed of young-old adults aged 60–69 years, with a slightly higher proportion of female participants. Most participants lived with their spouse, had an education level below senior high school, and more than half reported having at least one chronic condition. Most respondents maintained full independence in activities of daily living, while over one-third reported perceived cognitive decline. Detailed descriptive statistics of demographic and health-related characteristics are presented in Table 1.

Table 1.

Descriptive analysis of categorical variables among older adults (N=535).

Variables	Category	Value
Variables	Category	n	%
Age	60-64 years old	123	22.99
	65-69 years old	221	41.31
	70-74 years old	99	18.50
	75-79years old	60	11.21
	80 years old and above	32	5.98
Gender	Female	226	42.24
Gender	Male	309	57.76
Education Level	Primary school or below	226	42.24
	Junior high school	176	32.90
	Senior high school	96	17.94
	Bachelor’s degree or above	37	6.92
Marital/Living Status	Living with spouse	99	18.50
Marital/Living Status	Living without spouse	436	81.50
Economic Status	Very insufficient	33	6.17
	Relatively insufficient	54	10.09
	Medium	280	52.34
	Relatively sufficient	110	20.56
	Very sufficient	58	10.84
Chronic Diseases	None	247	46.17
Chronic Diseases	Present	288	53.83
Self-rated Health	Very poor	35	6.54
	Relatively poor	65	12.15
	Medium	240	44.86
	Relatively healthy	126	23.55
	Very healthy	69	12.90
Independence in Daily Life	Independence	392	73.27
Independence in Daily Life	No Independence	143	26.73
Cognitive Function	maintained	340	63.55
Cognitive Function	declined	195	36.45

The mean scores for self-perception (SP), information acquisition (IA), and interactive judgment (IJ) were 8.12±2.41, 14.03±3.15, and 9.15±1.78, respectively, with a total digital health literacy score of 31.05±5.82. Scores of psychosocial indicators including depressive symptoms, loneliness and social interaction also are shown in Table 2.

Table 2.

Descriptive analysis of continuous variables among older adults (N=535).

Variables	Assessment	Value
Variables	Assessment	Mean	SD
Digital Health Literacy	Self-Perception	8.12	2.41
	Information Acquisition	14.03	3.15
	Interactive Judgment	9.15	1.78
	m-HEALS	31.05	5.82
Depression Symptoms	GDS-11	6.33	1.16
Loneliness	ULS-6	10.04	2.25
Social Interaction	Index of Social Interaction	13.29	2.05

Abbreviations: Standard division (SD), 11-item Chinese version of the Short Form Geriatric Depression Scale (GDS-11), 6-item Chinese version of University of California Los Angeles Loneliness Scale (ULS-6).

Latent profile analysis of digital health literacy

Fit indices for latent profile models with 2 to 5 profiles are presented in Table 3. The three-profile model demonstrated optimal fit, supported by the highest entropy value (0.806), lower AIC, BIC, and aBIC values compared to other models, and statistically significant LMR and BLRT results (p < .05). Although the four-profile and five-profile models had slightly lower AIC/BIC/aBIC values, their lower entropy values (0.798 and 0.785, respectively) indicated poorer fit. All three profiles in the three-profile model exceeded the minimum class size requirement of 5%, confirming its selection as the final model.

Table 3.

Fit indices for latent profile models.

Model	AIC	BIC	aBIC	Entropy	BLRT	LMRT	Probability
2	5990.739	6033.562	6001.818	0.792	<0.001	<0.001	0.834/0.166
3	5327.42	5387.372	5342.931	0.806	<0.001	<0.001	0.389/0.478/0.133
4	5099.585	5176.665	5119.528	0.798	<0.001	<0.001	0.389/0.467/0.053/0.091
5	4925.981	5020.191	4950.356	0.785	<0.001	0.001	0.376/0.352/0.128/0.048/0.096

Abbreviations: Information Criterion (AIC), Bayesian Information Criterion (BIC), Sample-size adjusted BIC (aBIC), Bootstrap Likelihood Ratio Test (BLRT), The Lo-Mendell-Rubin Likelihood Ratio Test (LMRT).

Three distinct latent profiles of digital health literacy were identified among the participants, as shown in Table 4 and Figure 1. Group 1 was designated the “Low SP-IA-IJ Group”, characterized by consistently low performance across all three dimensions (self-perception, information acquisition and interactive judgment) of digital health literacy. Group 2 was labeled the “Medium SP-IA and Low IJ Group”, with moderate levels of self-perception (SP) and information acquisition (IA), alongside low performance in interactive judgment (IJ). Group 3 was defined as the “High SP-IA and Medium IJ Group”, which showed the highest levels of self-perception (SP) and information acquisition (IA) among the three profiles, with a moderate level of interactive judgment (IJ). Notably, interactive judgment was the dimension with the lowest overall performance across all three identified profiles.

Table 4.

Three latent digital health literacy groups among older adults.

Digital health literacy	Group 1 (n=208, 38.88%)	Group 2 (n=256, 47.85%)	Group 3 (n=71, 13.27%)	Total
Digital health literacy	Low SP-IA-IJ	Medium SP-IA and low IJ	High SP-IA and medium IJ	Total
Self-Perception (SP)	6.35±1.38	8.21±1.62	12.25±1.55	8.12±2.41
Information Acquisition (IA)	11.43±1.56	14.75±1.70	19.71±1.65	14.03±3.15
Interactive Judgment (IJ)	8.50±1.40	9.23±1.64	11.20±1.56	9.15±1.78
m-HEALS	26.28±3.21	32.19±3.85	43.16±4.12	31.30±5.82

Abbreviations Note: Self-Perception (SP), Information Acquisition (IA), Interactive Judgment (IJ).

Figure 1.

The three latent digital health literacy groups among older adults.

Table 5 presents the mean posterior probability values for each profile in the final model, all of which exceeded 0.8, indicating high classification accuracy.

Table 5.

The Latent Class Probabilities of the three-profile model.

Profiles	Group 1	Group 2	Group 3
Group 1	0.929	0.053	0.018
Group 2	0.041	0.937	0.022
Group 3	0.013	0.039	0.948

Group 1: Low SP-IA-IJ, Group 2: Medium SP-IA and Low IJ, Group 3: High SP-IA and Medium IJ.

Abbreviations: Self-Perception (SP), Information Acquisition (IA), Interactive Judgment (IJ).

Table 6 shows differences in variables (e.g., age, gender, education) across the three latent profiles. Chi-square tests revealed statistically significant differences in age, education level, economic status, self-rated health, independence in daily life, and cognitive function among the profiles (p < .01). In contrast, no significant differences were found for gender, marital/living status, or chronic disease status.

Table 6.

Comparison of variables across the three latent profiles of digital health literacy.

Variables	Group 1		Group 2		Group 3		χ²	p
Variables	n	%	n	%	n	%	χ²	p
Age							28.65	<0.001
60-64 years old	32	15.38^b	68	26.56^a	23	32.39^a
65-69 years old	80	38.46^a	112	43.75^a	29	40.85^a
70-74 years old	44	21.15^a	44	17.19^a	11	15.49^a
75-79years old	30	14.42^a	24	9.38^a	6	8.45^a
80 years old and above	22	10.58^a	8	3.12^b	2	2.82^ab
Gender							0.33	0.849
Female	86	41.35	109	42.58	31	43.66
Male	122	58.65	147	57.42	40	56.34
Educational Level							45.72	<0.001
Primary school or below	122	58.65^a	99	38.67^b	15	21.13^c
Junior high school	60	28.85^a	90	35.16^a	26	36.62^a
Senior high school	20	9.62^b	48	18.75^a	19	26.76^a
Bachelor’s degree or above	6	2.88^a	19	7.42^ab	11	15.49^b
Marital/Living Status							2.16	0.34
Living with spouse	42	20.19	45	17.58	12	16.90
Living without spouse	166	79.81	211	82.42	59	83.10
Economic Status							22.39	0.004
Very insufficient	19	9.14^a	12	4.69^a	2	2.82^a
Relatively insufficient	32	15.38^a	26	10.16^a	5	6.90^a
Medium	104	50.00^a	138	53.91^a	38	53.52^a
Relatively sufficient	39	18.75^a	53	20.70^a	17	23.94^a
Very sufficient	14	6.73^a	27	10.55^a	9	12.68^a
Chronic Diseases							3.86	0.145
None	88	42.31	123	48.05	36	50.7
Present	120	57.69	133	51.95	35	49.3
Self-rated Health							18.95	0.015
Very poor	24	11.54^a	16	6.25^a	2	2.82^a
Relatively poor	34	16.35^a	30	11.72^a	6	8.45^a
Medium	88	42.31^a	119	46.48^a	32	45.07^a
Relatively healthy	44	21.15^a	61	23.83^a	19	26.76^a
Very healthy	18	8.65^a	30	11.72^a	12	16.9^a
Dependence in Daily Living							31.21	<0.001
Independence	129	62.02^b	201	78.13^a	61	85.92^a
No independence	79	37.98	55	21.87	10	14.08
Cognitive Function							48.57	<0.001
Maintained	106	50.96^b	175	68.75^a	57	80.28^a
Declined	102	49.04	81	31.25	14	19.72

Note. Group comparisons were performed using the chi-square test. For variables with significant overall differences, post hoc pairwise comparisons were conducted using the Bonferroni correction (significance level: α=0.05/3=0.0167, p < .0167). a,b,c Values in the same row without a common superscript are significantly different (p < 0.0167, Bonferroni-adjusted).

Group 1: Low SP-IA-IJ, Group 2: Medium SP-IA and Low IJ, Group 3: High SP-IA and Medium IJ.

Abbreviations: Self-Perception (SP), Information Acquisition (IA), Interactive Judgment (IJ).

This study also tested differences in depressive symptoms, loneliness, and social interaction across the latent profiles. As shown in Table 7, ANOVA results indicated significant between-group differences in scores for depressive symptoms, loneliness, and social interaction (p < .01). Specifically, participants in Group 3 (the “High SP-IA and Medium IJ Group”) had lower depressive symptoms, lower loneliness, and better social interaction compared to the other two groups.

Table 7.

Differences in depressive symptoms, loneliness and social interaction.

Variables	Group 1	Group 2	Group 3	F	p
Depression Symptoms	7.45±0.51^a	5.97±0.58^b	4.35±0.49^c	128.753	<0.001
Loneliness	12.22±1.03^a	9.30±1.21^b	6.37±0.51^c	86.325	<0.001
Social Interaction	11.29±0.71^a	14.02±1.17^b	16.63±0.76^c	112.587	<0.001

Note. Group comparisons were performed using one-way ANOVA. For variables with significant overall differences, post hoc pairwise comparisons were conducted using Tukey’s HSD test. a,b,c Within each row, groups with different superscripts are significantly different from each other (p < .05, Tukey's HSD test).

Group 1: Low SP-IA-IJ, Group 2: Medium SP-IA and Low IJ, Group 3: High SP-IA and Medium IJ.

Abbreviations: Self-Perception (SP), Information Acquisition (IA), Interactive Judgment (IJ).

Multinomial logistic regression for the three latent profiles

A multinomial logistic regression analysis identified factors associated with digital health literacy profile membership, using Group 3 (“High SP-IA and Medium IJ”) as the reference.

As shown in Table 8, for the Low SP-IA-IJ Group, older age, perceived cognitive decline, impaired independence in daily living, higher levels of depressive symptoms, and greater loneliness were significantly associated with higher odds of profile membership. In contrast, higher education level, better economic status, better self-rated health, and better social interaction were significantly associated with lower odds of membership in this profile.

Table 8.

Multinomial logistic regression for digital health literacy among older adults.

Variables	Reference category	Group 1 vs Group 3			Group 2 vs Group 3
Variables	Reference category	OR	95%CI	p	OR	95%CI	p
Age	60-64 years old	3.22	2.15-4.82	<0.001	2.24	1.51-3.32	<0.001
Gender	Male	1.15	0.78-1.70	0.486	1.02	0.71-1.47	0.915
Education level	Primary school or below	0.68	0.55-0.84	<0.001	0.87	0.73-1.04	0.132
Marital/living status	Living without spouse	0.78	0.56-1.09	0.145	0.82	0.60-1.12	0.218
Economic status	Very insufficient	0.64	0.51-0.81	<0.001	0.85	0.70-1.03	0.096
Chronic Diseases	None	1.21	0.85-1.72	0.287	1.13	0.81-1.58	0.473
Self-rated Health	Very poor	0.76	0.63-0.92	0.005	0.81	0.68-0.97	0.023
Dependence in daily living	Independence	1.64	1.12-2.29	0.010	1.16	0.82-1.64	0.408
Cognitive Function	Maintained	2.21	1.53-3.19	<0.001	1.73	1.23-2.43	0.002
Depression Symptoms		1.51	1.32-1.73	<0.001	1.37	1.20-1.56	<0.001
Loneliness		1.32	1.18-1.43	<0.001	1.11	1.01-1.22	0.031
Social Interaction		0.73	0.65-0.82	<0.001	0.95	0.86-1.05	0.327

Group 1: Low SP-IA-IJ, Group 2: Medium SP-IA and Low IJ, Group 3: High SP-IA and Medium IJ.

Abbreviations: Self-Perception (SP), Information Acquisition (IA), Interactive Judgment (IJ), Odds Ratio (OR), 95% Confidence Interval (95%CI).

For the Medium SP-IA and Low IJ Group, older age, perceived cognitive decline, higher levels of depressive symptoms, and greater loneliness were significantly associated with higher odds of profile membership, while better self-rated health was significantly associated with lower odds of membership in this profile.

Gender, marital/living status and chronic disease status had no significant association with profile membership.

Discussion

Characteristics of latent profiles of digital health literacy

This study focused on community-dwelling older adults in China, with the aims of exploring the latent profiles of their digital health literacy and analyze the health-related factors associated with profile membership. Previous studies have indicated that older adults generally exhibit lower levels of digital health literacy compared with other age groups.³⁰ The findings further support this conclusion that the total digital health literacy score of the study participants was lower than that reported in general population samples,²⁷ and the high-literacy profile accounted for only a small proportion of the total sample. Latent profile analysis revealed significant within-group heterogeneity in digital health literacy among the study population, with three distinct latent profiles identified. While many previous studies on digital health literacy in Chinese older adults that focus on hierarchical differences in overall literacy levels,³⁹ this study identified asynchronous variation patterns across the subdimensions of digital health literacy within the profiles, which provides preliminary evidence to support the differentiated intervention strategies tailored to varying literacy levels advocated by existing research.³⁰

Consistent with comparative findings from studies in other populations using the m-eHEALS scale,⁴¹ older adults across all three latent profiles in this study did not achieve high scores in the interactive judgment dimension. Further analysis revealed that interactive judgment dimension is a universal weak dimension of digital health literacy across all identified profiles. Specifically, the lowest scores in this dimension were observed in Group 1 (Low SP-IA-IJ) and Group 2 (Medium SP-IA and Low IJ). Even in Group 3 (High SP-IA and Medium IJ), which showed relatively strong performance in the self-perception and information acquisition dimensions, a clear deficit remained in the interactive judgment dimension, with no synchronous improvement alongside the overall digital health literacy level.

From the perspective of the three-tiered digital divide framework, the interactive judgment dimension captures the core constructs of both the second-order skill divide and the third-order outcome divide in the field of digital health, which explains the universality of its deficit across different profiles. Specifically, the interactive judgment dimension includes both the operational skills of using digital tools for health information interaction (the core of the second-order skill divide) and the ability to critically evaluate health information and translate it into personal health decisions (the core of the third-order outcome divide).^10,32 Existing studies have shown that although the development of information technology and digital medicine has provided numerous online health resources for older adults, they often lack sufficient motivation to use mHealth tools.⁴⁸ In addition, concerns about online fraud, hacking and data privacy risks further limit their willingness to interact through digital channels.^49,50 Furthermore, the proliferation of false health information online makes it difficult for older adults to accurately analyze and evaluate various types of online health information.^51,52 These factors together hinder the development of older adults’ interactive judgment ability from both the skill and outcome levels of the digital divide, resulting in generally low scores in this dimension across heterogeneous groups.

Theoretically, building on existing views that digital health literacy among older adults presents heterogeneity at different levels,^39,53 this study extends existing research by identifying asynchronous variation patterns across dimensions of digital health literacy among Chinese community-dwelling older adults, thus enabling a more in-depth characterization of the structural features of the second- and third-order digital divides in this population. Practically, this study identifies interactive judgment as a universal core intervention target for all older adult populations. This means that intervention strategies should not only provide stratified support for subgroups with distinct literacy profiles^22,39 but also deliver targeted capacity-building interventions focused on interactive judgment for older adults, so as to fundamentally promote the improvement of digital health literacy among older adults. The heterogeneous profiles identified in this study can also deliver precise, actionable guidance for frontline practice. For older adults in Group 1 (Low SP-IA-IJ) with low literacy across all dimensions, bottom-line supportive interventions can be delivered through community-based one-to-one digital skill tutoring and themed training focused on basic health information acquisition. For those in Group 2 (Medium SP-IA and Low IJ) with moderate literacy in core dimensions but a prominent deficit in interactive judgment, priority can be given to targeted training covering critical evaluation of online health information and prevention of digital health fraud. For those in Group 3 (High SP-IA and Medium IJ) with relatively high overall literacy, the findings support their cultivation as core volunteers for community digital health mutual aid, while the profile characteristics also provide reference for the age-friendly interactive design of digital health platforms.

Factors associated with digital health literacy profile membership

Building on the identified heterogeneous latent profiles of digital health literacy, this study examined factors associated with older adults’ latent profile membership using multinomial logistic regression. Demographic characteristics, including age, gender, marital/residential status, education level, and economic income level, showed differential association patterns across the three identified profile groups.

Older age was significantly associated with a higher likelihood of being classified into Group 1 (Low SP-IA-IJ) and Group 2 (Medium SP-IA and Low IJ). This association may be related to the higher prevalence of technophobia among older adults, including anxiety, tension, and hesitation towards technology use,^54,55 as well as age-related physical adaptation difficulties, such as decreased vision and hearing, slowed movement and reaction, stiff muscles and joints, and hand tremors.^49,56 Higher education level and better economic status were significantly associated with a lower likelihood of membership in Group 1 (Low SP-IA-IJ), while no significant association was found with profile membership in Group 2 (Medium SP-IA and Low IJ). No significant association was identified between gender, marital/residential status and profile membership, and previous studies have not reached a consistent conclusion on the association between the above demographic characteristics and digital health literacy among older adults.^13,28,30

The findings of this study suggest that the association between demographic characteristics and older adults’ digital health literacy gradually weakens as their digital health literacy level increases, following a trend of diminishing marginal effect. From the perspective of the three-tiered digital divide framework, education level and economic status are associated with the first-order digital divide among older adults at the level of digital device and internet accessibility,⁵³ and lower education level and economic income are often core driving factors of digital exclusion.^10,57 Therefore, for older adults with low literacy, education level and economic status may be statistically significant key factors correlated with their digital health literacy. However, for older adults with medium or higher digital health literacy, their literacy level is more affected by psychosocial factors such as usage skills and social support within the second- and third-order digital divides,^27,28 leading to a gradual weakening of the association of demographic factors.

In addition to demographic characteristics, health-related factors also showed distinct differential associations with latent profile membership among older adults. Objective chronic disease status was not significantly associated with profile membership in this study, which is inconsistent with previous findings suggesting that health service needs are correlated with improved digital health literacy.^22,30,58 In contrast, self-rated health, as a subjective indicator of individual health perception, showed a robust and significant association with profile membership. Older adults with poorer self-rated health had a significantly higher likelihood of being classified into both the Group 1 (Low SP-IA-IJ) and Group 2 (Medium SP-IA and Low IJ) profiles. This differential association pattern between objective and subjective health indicators may be explained by two potential pathways. First, the lack of non-verbal emotional support in online health service interactions may fail to meet older adults’ needs for emotional support in chronic disease management,⁵⁹ which may reduce the attractiveness of such services for addressing chronic health conditions, and thus limit the association between chronic disease status and digital health literacy. Second, older adults with better self-rated health tend to be more proactive in health management and more inclined to seek internet-based health services,⁶⁰ which is associated with more frequent use of digital tools to access online health resources, and thus a higher likelihood of better digital health literacy.

With regard to physical function and cognitive status, independence in instrumental activities of daily living and cognitive function showed heterogeneous associations with latent profile membership. Impaired independence was significantly associated with a higher likelihood of membership in Group 1 (Low SP-IA-IJ), which may be related to difficulties in using digital devices caused by physical limitations,^56,61 and thus an association with lower digital health literacy. In contrast, well-maintained cognitive function was significantly associated with a lower likelihood of membership in both Groups, suggesting that cognitive decline is associated with a comprehensive reduction in the ability to understand and utilize digital health services.^49,62 This finding indicates that future interventions may integrate cognitive function training with digital health literacy improvement, for example, by incorporating content related to digital health service use and digital health tool operation into cognitive impairment prevention programs, to achieve synergy between functional support and literacy enhancement.

While most prior studies exploring the heterogeneity of digital health literacy among older adults have paid limited attention to psychosocial factors including depressive symptoms, loneliness, and social interaction, the findings of this study confirm the close associations between the severity of depressive symptoms, loneliness, social interaction, and latent profile membership of digital health literacy among older adults. These findings are also consistent with existing research conclusions that better mental health status is associated with higher digital health literacy in older adults.^63,64

Consistent with the results of latent profile analysis, greater severity of depressive symptoms and higher levels of loneliness were significantly associated with a higher likelihood of classification into both Group 1 (Low SP-IA-IJ) and Group 2 (Medium SP-IA and Low IJ). In contrast, lower levels of social interaction were only significantly associated with a higher likelihood of membership in Group 1 (Low SP-IA-IJ). One possible hypothesized explanation is that, as an internal and persistent negative mental state, depression and loneliness may be linked to reduced ability and motivation among older adults to analyze, judge, and communicate health-related information,^65–67 which could in turn correlate with variations in digital health literacy across all dimensions and different profile levels. On the other hand, social interaction mainly reflects the relationship between individual older adults and their social environment.⁶⁸ Within the second-order digital divide framework, social interaction is correlated with the use scenarios of basic digital tools and opportunities for peer learning, however, for older adults in Group 2 (Medium SP-IA and Low IJ), who have already achieved moderate levels of self-perception and information acquisition, the key limitation to their digital health literacy appear to shift to higher-order skills such as evaluation, interaction, and decision-making.^10,32 As a result, the strength of the association between social interaction and digital health literacy is significantly weakened, which explains the non-significant association observed in this group. It should be noted that this inference does not mean that social interaction is unrelated to older adults with moderate digital health literacy; rather, any potential association may be masked by more strongly correlated factors such as depressive symptoms, and thus no statistically significant association was detected in this study.

In summary, the three profiles exhibited notable differences in both digital health literacy levels and patterns of associated factors. With Group 3 as the reference, Group 1 (Low SP-IA-IJ) showed significant associations with all examined factors, pointing to an accumulation of multidimensional disadvantages. Group 2 (Medium SP-IA and Low IJ) was not significantly associated with education, economic income, or social interaction, suggesting that at this level of digital health literacy, the association of socioeconomic and social interaction factors weakens, while psychosocial factors emerge as more prominent. Meanwhile, even Group 3 (High SP-IA and Medium IJ) did not achieve optimal scores in interactive judgment, supporting that suboptimal interactive judgment is a universal limitation across all profiles and a pivotal link between the second-order skill divide and the third-order outcome divide.

Notably, limited by the cross-sectional study design, bidirectional associations may exist for all findings reported herein. Specifically, negative mental states may correlate with lower digital health literacy levels, while difficulties in digital and social integration linked to insufficient digital health literacy may in turn be associated with exacerbated depressive symptoms and loneliness in this population. Future longitudinal studies are needed to clarify the direction and underlying mechanisms of these associations.

The findings of this study suggest that psychosocial factors may correlate with older adults’ digital health literacy via potential pathways related to the “construction of a digital and intelligent social interaction environment”. Social interaction may provide scenarios for the use of online health tools, such as community mutual-aid learning and joint operation with family members, which could support the practice and consolidation of digital health skills, rather than relying solely on one-way skill instruction.

Policy recommendations

Against the backdrop of deeply intertwined digital/intelligent development and population aging in China, the digital divide faced by older adults due to insufficient digital health literacy will persist long-term,⁶⁹ and its impact on their health rights and interests will become increasingly prominent.⁷⁰ This placed higher demands on the foresight and adaptability of policy formulation. It should be noted that the findings of this study are based on a sample of older adults from urban and suburban communities in a specific region of China, and the corresponding policy recommendations are first anchored to the urban and suburban community scenarios covered by the study, before presenting exploratory extended implications for broader reference.

For urban and suburban community scenarios directly covered by this study, it is recommended incorporating the improvement of digital health literacy among older adults into the core indicator systems for digital/intelligent social governance and primary community services,^71,72 clarify its priority in the aging development strategy, and ensure corresponding resource investment from the perspective of top-level policy design at the local level. Meanwhile, targeted strategies should be implemented with a focus on high-risk groups, prioritizing older adults with advanced age, poor economic conditions, weak physical function, or poor mental health, to reduce the risk of widening literacy gaps.^58,69 Additionally, to address deficiencies in interactive judgment capabilities, it is necessary to establish credible and authoritative health information channels on mainstream internet platforms,^73,74 and promote digital/intelligent assistant services within primary community welfare systems to enhance the accessibility and interactivity of online health services for older adults.^71,74,75

From a broader national-level perspective, the above recommendations based on the study findings can be used as exploratory reference for the formulation of national aging and digital inclusion policies. It should be clarified that the applicability of these recommendations will be affected by contextual factors such as regional digital infrastructure development, community service capacity, and the urban-rural divide. For example, in rural areas or regions with relatively weak digital infrastructure, the primary focus of intervention should first address the first-order digital divide related to digital access and equipment availability, before promoting targeted interventions to improve digital health literacy. The core findings of this study regarding the common weaknesses of digital health literacy and the influencing factors of its heterogeneity have certain universal reference value for the formulation of relevant policies in different regions of China.

Limitations and future research

This study has several potential limitations. First, the most critical potential limitation is selection bias associated with the study’s inclusion criteria, which only enrolled participants able to use digital devices. This criterion excludes older adults with minimal or no digital literacy, and may lead to an overestimation of overall digital health literacy levels in the broader older adult population. Importantly, this means the observed digital health literacy heterogeneity and within-group digital divide likely apply only to older adults who have crossed the first-order digital access divide. The actual overall digital divide among older adults is likely far more pronounced than our findings suggest, as the most digitally marginalized group is excluded from our sample. Second, the cross-sectional design adopted in this study cannot establish causal relationships between the examined influencing factors and digital health literacy among older adults. Third, all participants were recruited from urban and suburban communities in Sichuan Province, China, which significantly constrains the generalizability of the findings to rural areas and other regions of the country.

In response to the above limitations, future research can be further expanded and improved in targeted directions. Multi-wave longitudinal follow-up studies are warranted to explore the dynamic trajectories of digital health literacy latent profiles among older adults over time, clarify the causal associations between influencing factors and the transition of profile membership, and identify critical windows for targeted intervention to prevent the decline of digital health literacy in older adults. Future research could also develop targeted intervention strategies tailored to the distinct characteristics of different digital health literacy profiles, evaluate their effectiveness via randomized controlled trials, and expand the study sample to cover rural older adults and those with limited or no digital device use experience, so as to verify the generalizability of the current findings and provide more comprehensive evidence for the formulation of widely applicable digital health literacy promotion models for older adults.

Conclusions

Grounded in the three-order digital divide framework, this study identified three distinct digital health literacy latent profiles among Chinese community-dwelling older adults, revealing significant within-group heterogeneity with asynchronous variation across literacy subdomains. The core novel finding is that interactive judgment is a universal bottleneck across all profiles, even among older adults with strong basic digital health literacy performance. Demographic, health, and psychosocial factors showed differential associations with digital health literacy profile membership. Local digital social governance policies should incorporate digital health literacy improvement as a core indicator, prioritize high-risk groups, and implement profile-tailored stratified interventions. Targeted interactive judgment capacity building and age-friendly digital health environment optimization are key to narrowing the digital divide and promoting equitable access to digital health services for older adults.

Footnotes

Acknowledgement

We thank Master of Social Work Education Center at Chengdu University of Information Technology for their essential support in conducting this study. The center’s faculty and students contributed significantly to the research implementation.

ORCID iDs

Bailiang Wu

Chunyan Jin

Jin Zheng

Ethical considerations

The study was approved by the ethical review committee of Sichuan Academy of Chinese Medicine Sciences (No. Ke202207-15-313).

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by Scientific Research Startup Funds for Introduced Talent at Chengdu University of Information Technology [376/376646].

Declaration of conflicting interests

The authors declare no potential conflicts of interest with respect to the research, authorship, and/or publication of this article, and all authors have read and agreed to the published version of the manuscript.

Contributorship

Bailiang Wu: Investigation, Data Curation, Writing-original draft.

Chunyan Jin: Methodology, Investigation, Data Curation, Manuscript review.

Jin Zheng: Investigation, Ethical Supervision.

Guarantor

The first author, Bailiang Wu, serves as the guarantor for this manuscript.

References

Meskó

Drobni

Bényei

, et al. Digital health is a cultural transformation of traditional healthcare. Mhealth 2017; 3: 38. https://doi.org/10.21037/mhealth.2017.08.07

Narayan

Chung

Adedinsewo

, et al. Access to digital health technologies: personalized framework and global perspectives. Nature Reviews Cardiology 2025; 16: 1–4. https://doi.org/10.1038/s41569-025-01184-5

Vandelanotte

Müller

Short

, et al. Past, present, and future of eHealth and mHealth research to improve physical activity and dietary behaviors. Journal of nutrition education and behavior 2016; 48(3): 219–228. https://doi.org/10.1016/j.jneb.2015.12.006

Hoogland

Mansfield

Lafranchise

, et al. eHealth literacy in older adults with cancer. Journal of geriatric oncology 2020; 11(6): 1020–1022. https://doi.org/10.1016/j.jgo.2019.12.015

Biancuzzi

Dal Mas

Bidoli

, et al. Economic and performance evaluation of E-health before and after the pandemic era: a literature review and future perspectives. International Journal of Environmental Research and Public Health 2023; 20(5): 4038. https://doi.org/10.3390/ijerph20054038

Maita

Maniaci

Haider

, et al. The impact of digital health solutions on bridging the health care gap in rural areas: a scoping review. The Permanente Journal 2024; 28(3): 130–143. https://doi.org/10.7812/TPP/23.134

Lupton

. The digitally engaged patient: Self-monitoring and self-care in the digital health era. Social Theory & Health 2013; 11(3): 256–270. https://doi.org/10.1057/sth.2013.10

Norman

Skinner

. eHEALS: the eHealth literacy scale. Journal of medical Internet research 2006; 8(4): e507. https://doi.org/10.2196/jmir.8.4.e27

Jafar

Quick

Rimányi

, et al. Social media and digital inequity: Reducing health inequities by closing the digital divide. International journal of environmental research and public health 2024; 21(11): 1420. https://doi.org/10.3390/ijerph21111420

10.

McNeely

. Traversing the digital divide in concept and effect: Relative interpretations and orientations. Policy & Internet 2024; 16(2): 214–221. https://doi.org/10.1002/poi3.409

11.

Estrela

Semedo

Roque

, et al. Sociodemographic determinants of digital health literacy: a systematic review and meta-analysis. International Journal of Medical Informatics 2023; 177: 105124. https://doi.org/10.1016/j.ijmedinf.2023.105124

12.

Gallifant

Celi

Pierce

. Digital determinants of health: opportunities and risks amidst health inequities. Nature Reviews Nephrology 2023; 19(12): 749–750. https://doi.org/10.1038/s41581-023-00763-4

13.

Kebede

Ozolins

Holst

, et al. Digital engagement of older adults: scoping review. Journal of Medical Internet Research 2022; 24(12): e40192. https://doi.org/10.2196/40192

14.

Mizrachi

Shahrabani

Nachmani

, et al. Obstacles to using online health services among adults age 50 and up and the role of family support in overcoming them. Israel Journal of Health Policy Research 2020; 9(1): 42. https://doi.org/10.1186/s13584-020-00398-x

15.

Kim

Jeon

Byun

, et al. Exploring E-health literacy and technology-use anxiety among older adults in Korea. InHealthcare 2023; 11(11): 1556, MDPI. https://doi.org/10.3390/healthcare11111556

16.

Hung

Lyons

. Health information technology use among older adults in the United States, 2009–2018. Current medical research and opinion 2020; 36(5): 789–797. https://doi.org/10.1080/03007995.2020.1734782

17.

Zhu

Peng

. Factors influencing the accessibility and reliability of health information in the face of the COVID-19 outbreak—a study in rural China. Frontiers in Public Health 2021; 9: 706779. https://doi.org/10.3389/fpubh.2021.706779

18.

Cui

Yin

, et al. Health-promoting behaviors mediate the relationship between eHealth literacy and health-related quality of life among Chinese older adults: a cross-sectional study. Quality of Life Research 2021; 30(8): 2235–2243. https://doi.org/10.1007/s11136-021-02797-2

19.

Sheng

Simpson

. Seniors, health information, and the Internet: motivation, ability, and Internet knowledge. Cyberpsychology, Behavior, and Social Networking 2013; 16(10): 740–746. https://doi.org/10.1089/cyber.2012.0642

20.

van Kessel

Hrzic

O'Nuallain

, et al. Digital health paradox: International policy perspectives to address increased health inequalities for people living with disabilities. Journal of medical Internet research 2022; 24(2): e33819. https://doi.org/10.2196/33819

21.

Campanozzi

Gibelli

Bailo

, et al. The role of digital literacy in achieving health equity in the third millennium society: a literature review. Frontiers in public health 2023; 11: 1109323. https://doi.org/10.3389/fpubh.2023.1109323

22.

Cui

Yin

, et al. Associations between health literacy, digital skill, and eHealth literacy among older Chinese adults: a cross-sectional study. Digital Health 2023; 9: 20552076231178431, [Internet]. https://doi.org/10.1177/20552076231178431

23.

Buyl

Beogo

Fobelets

, et al. e-Health interventions for healthy aging: a systematic review. Systematic reviews 2020; 9(1): 128. https://doi.org/10.1186/s13643-020-01385-8

24.

Bevilacqua

Strano

Di Rosa

, et al. eHealth literacy: from theory to clinical application for digital health improvement. Results from the ACCESS training experience. International Journal of Environmental Research and Public Health 2021; 18(22): 11800. https://doi.org/10.3390/ijerph182211800

25.

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, [Internet]. https://doi.org/10.1177/20552076221102765

26.

Yuen

Winter

Savira

, et al. Digital health literacy and its association with sociodemographic characteristics, health resource use, and health outcomes: rapid review. Interactive journal of medical research 2024; 13(1): e46888. https://doi.org/10.2196/46888

27.

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(1): 996. https://doi.org/10.1186/s12889-022-13378-4

28.

Shi

, et al. Factors influencing digital health literacy among older adults: A scoping review. Frontiers in public health 2024; 12: 1447747. https://doi.org/10.3389/fpubh.2024.1447747

29.

Liu

Wang

, et al. Impact of digital health literacy on health-related quality of life in Chinese community-dwelling older adults: the mediating effect of health-promoting lifestyle. Frontiers in Public Health 2023; 11: 1200722. https://doi.org/10.3389/fpubh.2023.1200722

30.

Zhang

Rosli

Mohd Saat

, et al. Scoping review of the e-health literature in older adults. Digital Health 2025; 11: 20552076251362545. https://doi.org/10.1177/20552076251362545

31.

Van Kessel

Wong

Clemens

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

32.

Vassilakopoulou

Hustad

. Bridging digital divides: A literature review and research agenda for information systems research. Information systems frontiers 2023; 25(3): 955–969. https://doi.org/10.1007/s10796-020-10096-3

33.

Ferguson

SLG

Moore

Hull

. Finding latent groups in observed data: A primer on latent profile analysis in Mplus for applied researchers. International Journal of Behavioral Development 2020; 44(5): 458–468. https://doi.org/10.1177/0165025419881721

34.

Luo

. The aged society in the digital age: features, opportunities, and challenges. Journal of Jiangxi Normal University (Philosophy and Social Sciences Edition) 2024: 17–25.

35.

National Bureau of Statistics . Main Data of the Seventh National Population Census (Internet) 2021, Cited 2025 Aug 7. Available from. https://www.stats.gov.cn/sj/zxfb/202302/t20230203_1901080.html

36.

China Internet Network Information Center (CNNIC) . The 53rd Statistical Report on Internet Development in China (Internet). Cited 2025 Sep 7. Available from:. https://www3.cnnic.cn/n4/2024/0322/c88-10964.html (2024).

37.

Shao

Fan

Cai

, et al. Research on the Current Situation and Governance Path of the Digital Divide of the Elderly in the Context of Digital Medicine. Medicine & Philosophy 2022; 43(24): 73–76.

38.

Dai

. Building an age-friendly digital society. Journal of Southeast University (Philosophy and Social Sciences) 2025; 27(01): 105–114. +151–152.

39.

Zhai

Liao

, et al. Latent profile analysis of eHealth literacy of the elderly in Shandong province. Chinese nursing research 2025; 39(12): 1969–1975.

40.

Hajian-Tilaki

. Sample size estimation in epidemiologic studies. Caspian journal of internal medicine 2011; 2(4): 289–298.

41.

Zhang

Zhu

. The research on the evaluation of consumer’s electronic health literacy and application in mobile health. Consumer Economics 2017; 33(01): 90–96+55.

42.

Chen

Hao

Wei

, et al. Comparison of Three ADL Scales in Disability Assessment of Middle-aged and Elderly People in China-Based on The Analysis of CHARLS 2018. Modern Preventive Medicine 2021; 48(13): 2401–2413.

43.

Cai

Jia

, et al. Strategy for the Choice of Appropriate Mild Cognitive Impairment Screening Scales for Community-dwelling Older Adults. Chinese General Practice 2022; 25(25): 3191.

44.

Tang

. Application of Short Form Geriatric Depression Scale (GDS-15) In Chinese Elderly. Chinese Journal of Clinical Psychology 2013; 21(03): 402–405.

45.

Zhou

, et al. Reliability and Avidity of ULS-8 loneliness scale in elderly samples in a rural community. https://doi.org/10.3969/j.issn.1672-7347.2012.11.008 (2012).37(11): 1124–1128.

46.

Anme

. Study to develop the Index of Social Interaction. J Natl Rehabil Cent Disabled 1992; 13: 1–7.

47.

Watanabe

Tanaka

Watanabe

, et al. Association between the older adults’ social relationships and functional status in Japan. Geriatrics & Gerontology International 2017; 17(10): 1522–1526. https://doi.org/10.1111/ggi.12909

48.

Clar

Dyakova

Curtis

, et al. Just telling and selling: current limitations in the use of digital media in public health: a scoping review. Public health 2014; 128(12): 1066–1075. https://doi.org/10.1016/j.puhe.2014.09.009

49.

Neugroschl

Zhu

, et al. Design considerations for mobile health applications targeting older adults. Journal of Alzheimer’s Disease 2021; 79(1): 1–8. https://doi.org/10.3233/jad-200485

50.

Andrews

Brown

Hawley

, et al. Older adults’ perspectives on using digital technology to maintain good mental health: interactive group study. Journal of Medical Internet Research 2019; 21(2): e11694. https://doi.org/10.2196/11694

51.

Zhang

Wang

. Fake vs. real health information in social media in China. Proceedings of the Association for Information Science and Technology 2017; 54(1): 742–743. https://doi.org/10.1002/pra2.2017.14505401139

52.

Melchior

Oliveira

. Health-related fake news on social media platforms: A systematic literature review. New Media & Society 2022; 24(6): 1500–1522. https://doi.org/10.1177/14614448211038762

53.

Mohan

Saleem

Voderhobli

, et al. Ensuring sustainable digital inclusion among the elderly: A comprehensive analysis. Sustainability 2024; 16(17): 7485. https://doi.org/10.3390/su16177485

54.

Di Giacomo

Ranieri

D’Amico

, et al. Psychological barriers to digital living in older adults: computer anxiety as predictive mechanism for technophobia. Behavioral Sciences 2019; 9(9): 96. https://doi.org/10.3390/bs9090096

55.

Kim

Freddolino

Greenhow

. Older adults’ technology anxiety as a barrier to digital inclusion: a scoping review. Educational Gerontology 2023; 49(12): 1021–1038. https://doi.org/10.1080/03601277.2023.2202080

56.

Fischer

David

Crotty

, et al. Acceptance and use of health information technology by community-dwelling elders. International journal of medical informatics 2014; 83(9): 624–635. https://doi.org/10.1016/j.ijmedinf.2014.06.005

57.

Zhang

. Measuring and applying digital literacy: implications for access for the elderly in rural China. Education and Information Technologies 2023; 28(8): 9509–9528. https://doi.org/10.1007/s10639-022-11448-z

58.

Zhang

Zhao

Liu

. Current situation and influencing factors of electronic health literacy of the elderly. Modern Preventive Medicine 2022; 49(09): 1642–1646+1652.

59.

Jiang

Sun

Chen

, et al. Patients’ and healthcare providers’ perceptions and experiences of telehealth use and online health information use in chronic disease management for older patients with chronic obstructive pulmonary disease: a qualitative study. BMC geriatrics 2022; 22(1): 9. https://doi.org/10.1186/s12877-021-02702-z

60.

Falk Erhag

Ahlner

Rydberg Sterner

, et al. Internet use and self-rated health among Swedish 70-year-olds: a cross-sectional study. BMC geriatrics 2019; 19(1): 365. https://doi.org/10.1186/s12877-019-1392-8

61.

Foster

Sethares

. Facilitators and barriers to the adoption of telehealth in older adults: an integrative review. CIN: Computers, Informatics, Nursing 2014; 32(11): 523–533. https://doi.org/10.1097/CIN.0000000000000105

62.

Wildenbos

Jaspers

Schijven

, et al. Mobile health for older adult patients: Using an aging barriers framework to classify usability problems. International journal of medical informatics 2019; 124: 68–77. https://doi.org/10.1016/j.ijmedinf.2019.01.006

63.

Nutakor

Zhou

Larnyo

, et al. Impact of health information seeking behavior and digital health literacy on self-perceived health and depression symptoms among older adults in the United States. Scientific Reports 2024; 14(1): 31080. https://doi.org/10.1038/s41598-024-82187-z

64.

Zhang

Lewis

Chen

, et al. Technology use and attitudes towards digital mental health in people with severe mental health problems: a survey study in China. Frontiers in Psychiatry 2023; 14: 1261795. https://doi.org/10.3389/fpsyt.2023.1261795

65.

Lapane

Lim

McPhillips

, et al. Health effects of loneliness and social isolation in older adults living in congregate long term care settings: A systematic review of quantitative and qualitative evidence. Archives of gerontology and geriatrics 2022; 102: 104728. https://doi.org/10.1016/j.archger.2022.104728

66.

Liu

Song

Zhu

. Exploring the relationship between older adults’ online health information seeking, negative emotions and prevention behaviors in the pandemic context: a two-wave longitudinal study. Frontiers in Public Health 2024; 12: 1377017. https://doi.org/10.3389/fpubh.2024.1377017

67.

Akhter-Khan

Prina

Wong

, et al. Understanding and addressing older adults’ loneliness: the social relationship expectations framework. Perspectives on Psychological Science 2023; 18(4): 762–777. https://doi.org/10.1177/17456916221127218

68.

Zou

van den Berg

Wong

, et al. Digital, social, and built environment influences on loneliness among older adults in China. Health & Place 2025; 96: 103581. https://doi.org/10.1016/j.healthplace.2025.103581

69.

Ren

Zhu

. The Trends and Socio-economic Differentials of Gray Digital Divide in China. Global Journal of Media Studies 2025; 03: 48–70.

70.

Xue

. The association between the digital divide and health inequalities among older adults in China: nationally representative cross-sectional survey. Journal of Medical Internet Research 2025; 27: e62645. https://doi.org/10.2196/62645

71.

Zuo

Cheng

Peng

, et al. Differences analysis of community residents' e-Health literacy level and influencing factors between urban and rural. Chinses Nursing Research 2022: 587–593.

72.

Wang

Xie

. Global artificial intelligence governance research in the digital and intelligent era: advances, trends and countermeasures. Journal of Knowledge Management 2025.

73.

Zhang

Chen

Chang

. Correlation between Peer Effects and eHealth Literacy in Older People. Journal of Nursing (China) 2022; 12: 75–78.

74.

Jin

. Research progress of eHealth literacy among older adults. Chinses Nursing Research 2024; 38(04): 620–623.

75.

Miller

Callegari

Abah

, et al. Digital literacy training for Low-Income older adults through undergraduate Community-Engaged learning: Single-Group Pretest-Posttest study. JMIR aging 2024; 7: e51675. https://doi.org/10.2196/51675