This web page was created programmatically, to learn the article in its unique location you’ll be able to go to the hyperlink bellow:
https://pmc.ncbi.nlm.nih.gov/articles/PMC13415940/
and if you wish to take away this text from our website please contact us
Abstract
Background
eHealth literacy (eHL) has emerged as an necessary competency for selling well being fairness. Vocational school college students are sometimes thought-about a digitally weak inhabitants. Understanding the heterogeneity of eHL on this inhabitants and the underlying psychological mechanisms is important for designing efficient well being promotion interventions.
Objective
This research aimed to determine latent profiles of eHL amongst vocational school college students, examine variations in health-promoting way of life ranges throughout profiles, and look at the mediating function of well being self-efficacy.
Methods
A cross-sectional survey was performed amongst vocational school college students from February to March 2026. The eHealth Literacy Scale, the Health Self-Efficacy Scale, and the Health-Promoting Lifestyle Profile-II had been used to gather knowledge. Latent profile evaluation (LPA) was used to determine sub-groups, and bias-corrected bootstrap analyses had been carried out to check the mediation results.
Results
A complete of 594 vocational school college students had been included. The imply complete eHL rating was 25.21 ± 5.26. Three distinct profiles had been recognized: Low Application-High Confidence (Profile 1, 15%), High Application-Low Critical Thinking (Profile 2, 52%), and High eHealth Literacy (Profile 3, 33%). For health-promoting way of life, college students in Profiles 2 and three scored considerably increased than these in Profile 1 (each p < 0.001), whereas no important distinction was discovered between Profile 2 and Profile 3 (p > 0.05). Mediation evaluation confirmed that well being self-efficacy partially statistically mediated the affiliation between eHL profile and health-promoting way of life when evaluating Profile 1 with Profile 2 (impact = 0.463, normal error [SE] = 0.091, 95% confidence interval [CI] 0.345, 0.586) and Profile 1 with Profile 3 (impact = 0.697, SE = 0.126, 95% CI 0.464, 0.963). The proportion mediated elevated from 43.62% for the transition from low to average eHL to 64.37% for the transition from low to excessive eHL.
Conclusion
eHL amongst vocational school college students was heterogeneous, and well being self-efficacy was statistically examined as a big mediator within the affiliation between eHL profiles and health-promoting way of life, although causal inference is restricted by the cross-sectional design. These findings recommend that vocational school college students shouldn’t be handled as a homogeneous inhabitants in well being promotion observe. Instead, person-centered interventions needs to be tailor-made to particular eHL profiles.
Keywords: eHealth literacy, health-promoting way of life, latent profile evaluation, self-efficacy, vocational school college students
1. Introduction
The speedy evolution of digital applied sciences has considerably reworked healthcare supply (1, 2). These developments could enhance entry to well being companies, cut back well being inequalities, and promote progress towards common well being protection (3). However, a well-recognized “digital paradox” stays: populations that would profit most from digital well being advances usually face the best limitations to entry and utilization (4, 5). A key necessary cause for this paradox is variation in people’ eHealth literacy (eHL) (5, 6). eHL is outlined as the flexibility to have interaction with digital applied sciences in efficient, secure, and useful methods to realize well being objectives (7). It has emerged as an necessary competency for selling well being fairness and narrowing the digital divide (8, 9).
The college years symbolize a pivotal interval for shaping college students’ health-promoting life (HPL), which have lasting results on long-term well being outcomes, together with the alleviation of subhealth situations (10) and diminished all-cause mortality (11). Notably, practically 42% of worldwide most cancers deaths in 2023 had been attributable to modifiable way of life elements (12), highlighting the worth of selling wholesome behaviors from early maturity as an economical public well being technique. Despite this significance, vocational school college students, who account for practically half of the upper training inhabitants in China, stay underrepresented within the literature (13). In basic, they exhibit decrease ranges of HPL than college students in complete universities (14). Previous research have proven a average correlation between eHL and health-related behaviors (15, 16). Individuals with increased ranges of eHL had a 3.4-fold decrease danger of unfavorable health-promoting way of life behaviors than these with decrease ranges (14). Currently, person-centered approaches to enhancing eHL have been more and more beneficial (5, 17). However, earlier research usually concentrate on the common degree of eHL and deal with school college students as a homogeneous group, thereby overlooking the heterogeneity inside this inhabitants (18).
To tackle this subject, our research adopted a person-centered method utilizing latent profile evaluation (LPA) to determine distinct eHL profiles amongst vocational school college students. In addition, it’s crucial to look at the psychological mechanism linking eHL profiles to HPL. According to Social Cognitive Theory, particular person conduct is influenced not solely by expertise but additionally by cognitive mediators (19). Within this framework, well being self-efficacy (HSE) (20), outlined as the idea in a single’s functionality to carry out health-related actions, could act as a important mediator between eHL and HPL (21, 22). However, it stays unclear whether or not this mediating function operates equally throughout totally different eHL profiles.
Therefore, this research aimed to (1) determine latent profiles of eHL amongst vocational school college students; (2) look at variations in HPL and HSE throughout these profiles; and (3) take a look at the mediating function of HSE within the affiliation between eHL profiles and HPL. These findings could present an empirical foundation for designing focused digital well being interventions for vocational school college students and contribute to efforts to advertise well being fairness within the digital period.
2. Methods
2.1. Study design
A cross-sectional survey was performed at a college of expertise in Wuxi, China, from February to March 2026. A multistage stratified cluster random sampling methodology was used. Students had been first stratified by school (eight schools) and educational 12 months (3 years). Within every college-year stratum, the variety of lessons to be sampled was decided proportionally primarily based on pupil enrollment to make sure representativeness throughout strata. A complete of 24 lessons (roughly 40 college students per class) had been randomly chosen utilizing a computer-generated random quantity methodology, with lessons serving because the sampling clusters. All eligible college students within the chosen lessons had been invited to take part. Data had been collected throughout scheduled class conferences. With the help of head academics and sophistication screens, college students accomplished the questionnaires in a supervised classroom setting. Participation was voluntary, and college students had been knowledgeable that they may decline participation with out penalty.
2.2. Ethical concerns
The research protocol was reviewed and accepted by the Ethics Committee of Changzhi Medical College (Approval No.: RT2026001). Electronic knowledgeable consent was obtained from all individuals earlier than survey completion. For individuals youthful than 18 years, knowledgeable consent was moreover obtained from their dad and mom or authorized guardians. On the primary web page of the net survey, individuals had been knowledgeable of the research function, the voluntary nature of participation, the nameless and confidential dealing with of their knowledge, and that the info could be used for analysis functions solely.
2.3. Participants
Participants had been eligible in the event that they had been full-time vocational school college students, had been capable of learn and full the questionnaire in Chinese independently, and supplied voluntary knowledgeable consent. Students had been excluded in the event that they self-reported a extreme bodily sickness or psychiatric dysfunction that would have an effect on questionnaire completion, or in the event that they had been concurrently collaborating in one other health-related research. A complete of 960 college students had been invited to take part on this survey. Among them, 822 questionnaires had been returned, with a response fee of 85.6%. Of these, 228 responses had been excluded on account of completion time lower than 3 min (n = 221) and lacking important indicators (n = 7). A complete of 594 college students had been included within the ultimate evaluation. A participant move diagram is offered in Figure 1.
Figure 1.

Participant recruitment and screening move diagram.
2.4. Sample measurement
Schumacker and Lomax (23) beneficial 250–500 individuals for mediation evaluation. In addition, Meyer and Morin (24) advised that LPA typically achieves secure and interpretable outcomes with pattern sizes starting from 200 to 500, whereas pattern measurement bigger than 500 could yield extra strong outcomes. Accordingly, the ultimate pattern measurement within the current research was thought-about ample for the deliberate analyses.
2.5. Data assortment
Data had been collected utilizing an nameless on-line questionnaire distributed by way of class instructors or pupil representatives. Based on earlier research (6, 25), a self-designed questionnaire, together with demographic info and measures of eHealth literacy, well being self-efficacy, and health-promoting way of life, was used to gather knowledge.
2.5.1. Demographic variables
The demographic info included gender, educational 12 months, main sort, hometown, household financial standing, parental training degree, and self-rated well being. Major sort was categorized into two teams primarily based on college students’ educational disciplines: Science and Engineering, and Arts and Humanities.
2.5.2. eHL
The eHL was assessed utilizing the Simplified Chinese model of the eHealth Literacy Scale (SC-eHEALS). The unique eHEALS was developed by Norman et al. in 2006 to guage people’ perceived expertise to search out, consider, and apply web-based well being info to handle well being points (26). Xu et al. (27) revised and validated the simplified Chinese model by way of rigorous cross-cultural adaptation and psychometric testing, demonstrating passable reliability and validity in Chinese populations. The SC-eHEALS consists of eight objects, every rated on a 5-point Likert scale (1 = strongly disagree, 5 = strongly agree), with complete scores starting from eight to 40. Higher scores point out higher perceived eHL. Previous research have generally used a complete rating of 32 or increased to point excessive degree of eHL. In this research, the Cronbach’s α coefficient for the SC-eHEALS was 0.821, indicating good inner consistency.
2.5.3. HSE
The HSE was measured utilizing the Self-Rated Abilities for Health Practices–Adolescent Version (SRAHP-A), developed by Chilton et al. (28) to evaluate adolescents’ perceived self-efficacy for health-promoting behaviors. The scale is grounded in self-efficacy principle and well being promotion principle, that are according to the goals of this research. The SRAHP-A consists of 28 objects throughout 4 domains: vitamin (six objects), train (seven objects), accountable well being practices (seven objects), and psychological well-being (eight objects). Each merchandise is rated on a 5-point Likert scale (0 = in no way, 4 = utterly), with complete scores starting from zero to 112. Higher scores mirror stronger perceived HSE. The scale has been utilized in a number of research and has demonstrated passable reliability and validity (29, 30). In this research, the Cronbach’s α coefficient for the overall scale was 0.968, with subscale coefficients of 0.898 for vitamin, 0.886 for train, 0.899 for accountable well being practices, and 0.939 for psychological well-being, indicating wonderful inner consistency.
2.5.4. HPL
The Health-Promoting Lifestyle Profile II (HPLP-II) was used to evaluate HPL. The unique scale was developed by Walker et al. (31). This research used the mainland China–tailored model validated by Cao et al. (32), which accommodates 40 objects throughout six domains: interpersonal relations (5 objects), well being accountability (11 objects), stress administration (5 objects), vitamin (six objects), bodily exercise (eight objects), and non secular progress (5 objects). Responses are rated on a 4-point Likert scale (1 = by no means, 4 = routinely), with complete scores starting from 40 to 160. Higher scores point out a more healthy way of life. In this research, the Cronbach’s α coefficient for the overall scale was 0.970, with subscale coefficients of 0.929 for interpersonal relations, 0.929 for well being accountability, 0.907 for stress administration, 0.884 for vitamin, 0.924 for bodily exercise, and 0.916 for SG.
2.6. Quality management
To guarantee knowledge high quality, a pilot survey was performed with 30 college students to guage merchandise readability and research feasibility. The common completion time was 6 min, and minor wording revisions had been made primarily based on the pilot outcomes. In the formal survey, responses had been excluded if completion time was lower than 3 min (pre-specified primarily based on pilot research) or in the event that they contained lacking knowledge on key variables, together with the first final result (eHealth literacy scale) and important demographic traits. To encourage participation, all individuals obtained a field of surgical masks and had been eligible for a random prize draw.
Because the core variables on this research (e.g., eHL, HSE, and HPL) had been measured concurrently utilizing self-report questionnaires, widespread methodology bias was thought-about a possible methodological concern. Several procedural measures had been carried out to attenuate this danger, together with nameless participation, voluntary responses, and standardized survey administration procedures. In addition, Harman’s single-factor take a look at was carried out by way of exploratory issue evaluation to look at the extent of widespread methodology bias (33). Consistent with earlier research, extreme widespread methodology bias was thought-about current if the primary issue defined greater than 40% of the overall variance (34). In the current research, the primary unrotated issue accounted for 37.72% of the overall variance, suggesting that widespread methodology bias was unlikely to have considerably affected the research findings, though its presence can’t be utterly dominated out.
Two sensitivity analyses had been performed to look at the robustness of the findings. First, latent profile analyses had been re-estimated utilizing the total pattern to evaluate the steadiness of the recognized profiles. Second, mediation fashions had been re-run with further adjustment for age, hometown, parental training, and main sort.
This research is reported following the rule of Strengthening the Reporting of Observational Studies in Epidemiology (STROBE).
2.7. Data evaluation
All statistical analyses had been performed utilizing SPSS 26.0 and Mplus 8.3. Descriptive statistics had been used to summarize individuals’ demographic traits and the principle research variables. Reliability evaluation of the used scales and correlation evaluation between variables had been then carried out.
The LPA was performed to determine eHL profiles amongst vocational school college students. In the LPA, fashions with one to 5 profiles had been estimated. Item means had been freely estimated throughout profiles, whereas merchandise variances had been held equal throughout profiles. The optimum mannequin was chosen primarily based on a number of mannequin match indices, together with the Akaike Information Criterion (AIC), Bayesian Information Criterion (BIC), and the adjusted Bayesian Information Criterion (aBIC) (35). Lower values of AIC, BIC, and aBIC point out higher mannequin match (36). Entropy worth was used to evaluate classification accuracy, with values better than 0.8 indicating good classification accuracy. Model choice additionally thought-about the sensible significance and interpretability of every resolution. In addition, every profile was required to account for no less than 5% of the overall pattern (37). Average posterior possibilities had been used to evaluate classification high quality. Local independence was evaluated by analyzing residual correlations amongst indicators.
Mediation evaluation was carried out utilizing the PROCESS macro (Model 4) to look at the mediating function of HSE within the affiliation between eHL profiles and HPL. eHL profile membership was entered as a categorical impartial variable, with the Low Application–High Confidence profile specified because the reference group. Direct, oblique, and complete results had been estimated utilizing a bias-corrected bootstrap methodology with 5,000 resamples, and 95% confidence intervals (CIs) had been calculated. If the 95% CI didn’t comprise zero, the corresponding impact was thought-about statistically important.
The mediation mannequin consisted of three regression equations: (1) HPL was regressed on eHL profiles to estimate the overall impact; (2) HSE was regressed on eHL profiles to estimate the affiliation between eHL profiles and the mediator; and (3) HPL was regressed on each eHL profiles and HSE to estimate the direct and oblique results. Covariate choice was guided primarily by theoretical relevance and prior empirical proof, with preliminary univariate analyses used solely as supplementary info (6, 25, 38). The identical demographic covariates had been included in all three regression fashions to attenuate potential confounding.
3. Results
3.1. Sample traits
After excluding invalid questionnaires, a ultimate pattern of 594 college students was included within the evaluation. No important variations in demographic traits had been noticed between included and excluded individuals (all p > 0.05). The ultimate pattern consisted of 594 college students. The imply age was 19.8 ± 1.6 years, starting from 16 to 25 years. Gender, household financial standing, and self-rated well being had been considerably related to HPL (p < 0.05). Table 1 exhibits extra detailed demographic traits of individuals.
Table 1.
Demographic traits of individuals (N = 594).
| Variables | N (%) | Outcome variable: HPL | |
|---|---|---|---|
| (Mean ± SD) | P | ||
| Gender | <0.001 | ||
| Male | 340 (57.2%) | 119.12 ± 24.70 | |
| Female | 254 (42.8%) | 111.07 ± 23.42 | |
| Academic Year | 0.992 | ||
| Freshman | 283 (47.6%) | 115.62 ± 25.05 | |
| Sophomore | 179 (30.1%) | 114.06 ± 23.70 | |
| Junior | 132 (22.3%) | 118.01 ± 23.82 | |
| Major Type | 0.577 | ||
| Science and Engineering | 477 (80.3%) | 115.96 ± 24.64 | |
| Arts and Humanities | 117 (19.7%) | 114.55 ± 23.82 | |
| Place of Origin | 0.162 | ||
| Urban | 198 (33.3%) | 117.67 ± 25.32 | |
| Rural | 396 (66.7%) | 114.69 ± 24.00 | |
| Family Economic Status | 0.002 | ||
| Below common | 80 (13.5%) | 108.60 ± 22.35 | |
| Average | 474 (79.8%) | 116.11 ± 24.35 | |
| Above common | 40 (6.7%) | 124.78 ± 26.61 | |
| Father’s Educational Level | 0.175 | ||
| Vocational school or beneath | 457 (80.0%) | 115.00 ± 23.59 | |
| Bachelor’s diploma or above | 119 (20%) | 118.40 ± 27.64 | |
| Mother’s Educational Level | 0.583 | ||
| Vocational school or beneath | 528 (88.9%) | 115.48 ± 24.10 | |
| Bachelor’s diploma or above | 66 (11.1%) | 117.24 ± 28.02 | |
| Self-rated Health | <0.001 | ||
| Poor/Fair | 39 (6.6%) | 107.92 ± 22.46 | |
| Good | 368 (62.0%) | 111.78 ± 22.56 | |
| Excellent | 187 (31.5%) | 124.98 ± 25.91 | |
3.2. Descriptive statistics
The imply complete scores for eHL, HSE, and HPL had been 25.21 ± 5.26, 81.33 ± 20.12, and 115.68 ± 24.47, respectively. The imply eHL rating was beneath the generally used threshold of 32 for top eHL. Among the HSE dimensions, psychological well-being confirmed the very best merchandise imply rating (23.58 ± 6.38), whereas vitamin confirmed the bottom merchandise imply rating (17.53 ± 4.66). Among the HPL dimensions, well being accountability had the very best merchandise imply rating (29.95 ± 7.98), whereas non secular progress had the bottom merchandise imply rating (15.01 ± 3.88). Detailed descriptive statistics are offered in Table 2.
Table 2.
Descriptive statistics of eHL, HSE, and HPL.
| Variables | Possible vary |
SD |
|---|---|---|
| eHL complete rating | 8–40 | 25.21 ± 5.26 |
| HSE complete rating | 0–112 | 81.33 ± 20.12 |
| Exercise (HSE) | 0–28 | 20.02 ± 5.48 |
| Nutrition (HSE) | 0–24 | 17.53 ± 4.66 |
| RHP (HSE) | 0–28 | 20.19 ± 5.58 |
| PW (HSE) | 0–32 | 23.58 ± 6.38 |
| HPL complete rating | 40–160 | 115.68 ± 24.47 |
| IR (HPL) | 5–20 | 16.13 ± 3.66 |
| HR (HPL) | 11–44 | 29.95 ± 7.98 |
| SM (HPL) | 5–20 | 15.09 ± 3.51 |
| Nutrition (HPL) | 6–24 | 17.89 ± 4.14 |
| PA (HPL) | 8–32 | 21.61 ± 6.14 |
| SG (HPL) | 5–20 | 15.01 ± 3.88 |
3.3. Correlation evaluation outcomes
As offered in Table 3, correlation evaluation confirmed that complete scores of the three predominant variables had been considerably positively correlated with one another. Specifically, the eHL complete rating was positively related to each the HSE complete rating (r = 0.46, p < 0.01) and the HPL complete rating (r = 0.35, p < 0.01). Additionally, a powerful constructive correlation was noticed between the HSE complete rating and the HPL complete rating (r = 0.64, p < 0.01).
Table 3.
Pearson correlation evaluation of eHL, HSE, and HPL.
| Variables | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| eHL complete rating | 1 | ||||||||||||
| HSE complete rating | 0.461** | 1 | |||||||||||
| Exercise (HSE) | 0.427** | 0.880** | 1 | ||||||||||
| Nutrition (HSE) | 0.434** | 0.908** | 0.746** | 1 | |||||||||
| RHP (HSE) | 0.449** | 0.925** | 0.730** | 0.813** | 1 | ||||||||
| PW (HSE) | 0.379** | 0.925** | 0.734** | 0.780** | 0.821** | 1 | |||||||
| HPL complete rating | 0.352** | 0.639** | 0.542** | 0.579** | 0.546** | 0.651** | 1 | ||||||
| IR (HPL) | 0.305** | 0.469** | 0.385** | 0.431** | 0.398** | 0.485** | 0.742** | 1 | |||||
| HR (HPL) | 0.324** | 0.559** | 0.481** | 0.505** | 0.487** | 0.556** | 0.884** | 0.552** | 1 | ||||
| SM (HPL) | 0.250** | 0.557** | 0.450** | 0.497** | 0.492** | 0.575** | 0.824** | 0.595** | 0.697** | 1 | |||
| Nutrition (HPL) | 0.260** | 0.543** | 0.438** | 0.490** | 0.474** | 0.562** | 0.858** | 0.644** | 0.680** | 0.720** | 1 | ||
| PA (HPL) | 0.298** | 0.515** | 0.445** | 0.474** | 0.426** | 0.524** | 0.838** | 0.489** | 0.671** | 0.582** | 0.646** | 1 | |
| SG (HPL) | 0.293** | 0.541** | 0.489** | 0.479** | 0.440** | 0.550** | 0.799** | 0.600** | 0.579** | 0.609** | 0.667** | 0.645** | 1 |
3.4. Results of latent profiles evaluation of eHL
As proven in Table 4, the AIC, BIC, and aBIC values decreased because the variety of profiles elevated. Considering mannequin match indices, entropy values, profile proportions, statistical assessments, and interpretability, the three-profile resolution was chosen because the optimum mannequin. Compared with the two-profile resolution, the three-profile mannequin demonstrated increased classification accuracy (Entropy = 0.912) and extra balanced profile proportions. Although the four-profile resolution confirmed barely higher match indices, the extra profile appeared to mirror a subdivision of an current profile and didn’t yield significant sensible distinctions. In the five-profile resolution, the LMR take a look at was not statistically important (p = 0.056). The common posterior possibilities for the three-profile mannequin had been 0.984, 0.967 and 0.940, respectively, all exceeding 0.80, indicating passable classification accuracy (Supplementary Table S1). No residual correlations exceeded |0.10|, supporting the belief of native independence. Furthermore, sensitivity evaluation utilizing the total pattern (N = 815) confirmed that the three-profile resolution remained optimum (Supplementary Table S2).
Table 4.
Fit statistics for latent profile fashions of eHL.
| Profiles | AIC | BIC | aBIC | Entropy | LMRT(p) | BLRT(p) | Profile Proportions |
|---|---|---|---|---|---|---|---|
| 1-Profile | 13300.482 | 13370.672 | 13319.877 | 1 | |||
| 2-Profile | 12297.947 | 12407.619 | 12328.251 | 0.881 | <0.001 | <0.001 | 0.638/0.362 |
| 3-Profile | 11587.648 | 11736.802 | 11628.863 | 0.912 | <0.001 | <0.001 | 0.152/0.520/0.328 |
| 4-Profile | 11504.333 | 11692.969 | 11556.457 | 0.877 | 0.003 | <0.001 | 0.150/0.485/0.258/0.108 |
| 5-Profile | 11476.453 | 11704.571 | 11539.487 | 0.841 | 0.056 | <0.001 | 0.153/0.412/0.091/0.254/0.089 |
Based on the LPA and the response patterns illustrated in Figure 2, the three profiles had been labeled to mirror their distinct eHL patterns. Profile 1 (n = 89, 15%), labeled Low Application-High Confidence, was characterised by low scores on Finding assets, Available assets, and Using info (all < 2.0) and comparatively excessive scores on Decision confidence and Answering questions (all > 3.0). Profile 2 (n = 309, 52%), labeled High Application-Low Critical Thinking, demonstrated excessive scores on Where to search out (4.25) and Answering questions (3.90) however a notably low rating on Distinguishing high quality (1.97). Profile 3 (n = 196, 33%), labeled High eHL, exhibited constantly excessive scores throughout all objects, with imply merchandise scores starting from 3.36 to 4.51.
Figure 2.

Map of eHL classes of the individuals.
3.5. Differences in HSE and HPL throughout eHL profiles
Table 5 exhibits the numerous variations in HSE complete scores (F[2,591] = 67.067, p < 0.001) and HPL complete scores (Welch’s F[2,564] = 59.739, p < 0.001) throughout the three recognized eHL profiles. Specifically, HSE complete scores elevated stepwise throughout the three eHL profiles, with Profile 3 exhibiting the very best scores, adopted by Profile 2 and Profile 1; all pairwise comparisons had been statistically important (all p < 0.05). Regarding HPL, college students in Profile 2 and Profile 3 exhibited considerably increased HPL complete scores than these in Profile 1 (p < 0.001), whereas no important distinction was discovered between Profile 2 and Profile 3 (p > 0.05).
Table 5.
Comparison of eHL, HSE, and HPL scores throughout the three recognized eHL profiles.
| Variables | Profile 1 (n = 89) | Profile 2 (n = 309) | Profile 3 (n = 196) | F/Welch’s F | Post hoc assessments |
|---|---|---|---|---|---|
| eHL complete ratingb | 16.61 ± 1.73 | 24.07 ± 2.05 | 30.98 ± 2.55 | 1414.31** | 1 < 2 < 3 |
| HSE complete ratingb | 63.12 ± 16.89 | 81.17 ± 18.64 | 89.97 ± 18.06 | 67.067** | 1 < 2 < 3 |
| HPL complete ratinga | 91.98 ± 26.52 | 119.55 ± 22.81 | 120.49 ± 19.33 | 59.739** | 1 < 2, 1 < 3 |
3.6. Mediation impact of HSE
Gender, educational 12 months, household financial standing, and self-rated well being had been included as covariates in all mediation fashions to regulate for potential confounding. With the Low Application–High Confidence profile because the reference group, each the High Application–Low Critical Thinking profile (β = 1.062, p < 0.001) and the High eHL profile (β = 1.083, p < 0.001) had been positively related to HPL in Model 1. Both profiles had been additionally positively related to HSE in Model 2 (R2 = 0.279, p < 0.001; Table 6). After HSE was included within the mannequin, the associations between eHL profiles and HPL had been weakened however remained statistically important. The regression coefficients decreased to 0.599 for the High Application–Low Critical Thinking profile and 0.386 for the High eHL profile (Table 6).
Table 6.
The mediating impact of HSE on HPL.
| Models | Variables |
R 2 |
F (df) |
β | t |
|---|---|---|---|---|---|
| Model 1 | Dependent variable: HPL | 0.245 | 31.757 (6) | ||
| High Application – Low Critical Thinking | 1.062 | 10.014*** | |||
| High eHealth Literacy | 1.083 | 9.584*** | |||
| Model 2 | Dependent variable: HSE | 0.279 | 37.847 (6) | ||
| High Application – Low Critical Thinking | 0.837 | 8.082*** | |||
| High eHealth Literacy | 1.260 | 11.414*** | |||
| Model 3 | Dependent variable: HPL | 0.466 | 72.923 (7) | ||
| High Application – Low Critical Thinking | 0.599 | 6.360*** | |||
| High eHealth Literacy | 0.386 | 3.669*** | |||
| Health Self-efficacy | 0.553 | 15.549*** |
As proven in Figure 3 and Table 7, the oblique impact of HSE for the High Application–Low Critical Thinking profile was 0.463 (95% CI: 0.345–0.586), accounting for 43.62% of the overall impact. For the High eHL profile, the oblique impact was 0.697 (95% CI: 0.560–0.840), accounting for 64.37% of the overall impact. The 95% CIs for each oblique results didn’t embody zero. Sensitivity analyses with further adjustment for age, hometown, parental training, and main sort yielded constant outcomes (Supplementary Tables S3, S4), indicating robustness of the findings.
Figure 3.

The mediating impact of HSE between eHL profiles and HPL.
Table 7.
Direct and oblique results of HSE on HPL.
| Path | Effect | Boot SE | Boot LLCI | Boot ULCI | Relative Effect (%) |
|---|---|---|---|---|---|
| Path 1: “High Application – Low Critical Thinking” profile → self-efficacy → HPL | |||||
| Direct Effect | 0.599 | 0.124 | 0.350 | 0.836 | 56.38% |
| Indirect Effect | 0.463 | 0.062 | 0.345 | 0.586 | 43.62% |
| Total Effect | 1.062 | 0.106 | 0.853 | 1.270 | |
| Path 2: “High eHealth Literacy” profile → self-efficacy → HPL | |||||
| Direct Effect | 0.386 | 0.129 | 0.127 | 0.632 | 35.63% |
| Indirect Effect | 0.697 | 0.072 | 0.560 | 0.840 | 64.37% |
| Total Effect | 1.083 | 0.113 | 0.861 | 1.305 | |
4. Discussion
To our data, that is the primary research to use LPA to determine distinct eHL profiles in vocational school college students. This research emphasised the heterogeneity of eHL on this inhabitants, which is according to the findings of Stellefson et al. (18). Through LPA, three profiles had been recognized: Low Application – High Confidence profile (Profile 1, 15%), High Application – Low Critical Thinking profile (Profile 2, 52%) and High eHL profile (Profile 3, 33%). Significant variations had been noticed in HPL and HSE throughout the three eHL profiles. HSE was statistically examined as a mediator between eHL profiles and HPL.
Furthermore, on this research, the imply complete rating of eHL (25.21) was beneath the brink of 26, which has been used to point a comparatively low degree of eHL (39). This rating was additionally decrease than the reported averages amongst college students typically universities in China (25) and people reported in a number of European international locations (6). Together, these findings recommend that vocational school college students could symbolize a digitally deprived subgroup and warrant focused consideration in efforts to advertise well being fairness.
Students within the Low Application – High Confidence profile had been characterised by excessive resolution confidence regardless of restricted sensible digital expertise. This sample could mirror overconfidence and is according to patterns described within the Dunning–Kruger literature. Specifically, people with decrease ranges of competence could overestimate their skills as a result of they’ve restricted consciousness of their very own data gaps and efficiency limitations (40). Consistent with this discovering, Bertolazzi et al. (41) recognized overconfidence in on-line well being info looking for amongst college college students, which can result in irrational self-care behaviors. Vocational school college students could possess intermediate ranges of health-related data, a stage at which overconfidence has been proven to peak (42). Moreover, this bias could also be additional bolstered by algorithm-driven personalization, which might enhance publicity to fragmented, homogeneous and low-quality well being info (43, 44). Therefore, interventions for this group ought to shift from basic motivational encouragement to extra rigorous coaching in info verification and digital competencies.
Turning to Profile 2, which represented the biggest subgroup (52%), these college students demonstrated sturdy expertise in finding assets and answering questions however had difficulties in critically evaluating the standard of well being info. This imbalance could stem from the vocational training surroundings, the place sensible expertise are emphasised over the event of important pondering (45). In addition, low HSE could hinder college students’ confidence of their means to guage well being info (43, 46). Furthermore, people with restricted eHL could also be extra more likely to encounter lower-quality or deceptive well being info on-line (47, 48). As a outcome, college students on this group could also be notably weak to pervasive misinformation and disinformation. Accordingly, digital well being interventions for this group shouldn’t solely educate college students find out how to discover well being info but additionally strengthen their means to critically consider well being info on digital platforms.
Finally, Profile 3 (33%, High eHL) exhibited constantly sturdy efficiency throughout software, info analysis, and decision-making dimensions. Students on this profile demonstrated comparatively sturdy perceived eHealth literacy throughout all assessed domains. Although this sample is broadly according to facets of important eHL proposed by Nutbeam (49), eHEALS doesn’t totally seize this assemble; due to this fact, this interpretation needs to be made with warning. At this stage, the worth of eHL could lengthen past particular person well being behaviors to broader population-level advantages (50). Notably, within the present digital panorama, a belief paradox has emerged: whereas scientifically rigorous establishments face growing skepticism, unvetted social media influencers are sometimes perceived as extra credible. A strategic response to this problem is to emphasise co-production within the design and supply of digital well being initiatives from the outset (3, 51). Therefore, college students on this profile could symbolize a useful useful resource for peer-led digital well being promotion initiatives. Interventions concentrating on these college students might empower them to behave as peer educators and co-creators in digital well being, doubtlessly facilitating the dissemination of dependable well being info amongst their friends (52).
A very notable discovering of this research is that HPL complete scores in Profile 3 didn’t considerably differ from these in Profile 2. This sample suggests a ceiling impact: enhancements in eHL past a sure threshold could not result in additional features in HPL. One believable clarification is that eHL acts as a midstream enabling issue fairly than a panacea for well being conduct (50). Healthy life are formed not solely by particular person digital expertise but additionally by broader structural elements, together with social, cultural, environmental, and financial influences, which are sometimes underemphasized in intervention research (53). Indeed, environmental and policy-related elements have been proven to exert stronger influences on school college students’ eHL than individual-level demographic traits (54).
From the angle of Social Cognitive Theory, well being conduct is influenced not solely by digital competencies but additionally by cognitive mediators comparable to HSE (19, 46). Our mediation evaluation additional revealed that the relative contribution of HSE differed between Profile 2 and Profile 3, though their HPL complete scores didn’t differ considerably. Higher general eHL in Profile 3 was related to a stronger mediating function of HSE, suggesting that enhanced digital literacy could promote more healthy practices partly by way of better psychological empowerment. Although causal relationships can’t be totally established because of the cross-sectional design, the findings assist the potential function of HSE as an necessary psychological pathway linking eHL and HPL. Importantly, though HSE and HPL share some widespread behavioral domains, they symbolize distinct constructs, reflecting confidence in partaking in health-promoting behaviors and the precise adoption of these behaviors, respectively. Thus, college students with distinct eHL profiles could present related ranges of HPL whereas differing within the psychological mechanisms underlying these behaviors. These findings spotlight the significance of mixing particular person functionality constructing with broader environmental assist, fairly than focusing solely on particular person elements.
Overall, the three recognized profiles reveal heterogeneity amongst vocational school college students in each general eHL and the configuration of subdomain competencies. For college students within the Low Application–High Confidence profile, focused interventions ought to prioritize sensible digital well being expertise coaching and consciousness of particular person data gaps. Students within the High Application–Low Critical Thinking profile could profit most from coaching in info appraisal and important analysis. In distinction, college students within the High eHL profile could also be effectively positioned to take part in peer-led well being promotion and digital well being co-creation initiatives. These findings recommend {that a} one-size-fits-all method could also be inadequate and that profile-specific intervention methods may very well be more practical in selling wholesome life amongst vocational school college students.
4.1. Strengths and limitations
This research has a number of strengths. First, to our data, it’s the first research to determine distinct eHL profiles amongst vocational school college students utilizing LPA, which captures the heterogeneity of eHL on this inhabitants. Second, grounded in Social Cognitive Theory, it additional clarifies the mediating function of HSE within the pathway from eHL profiles to HPL. Third, the identification of a ceiling impact highlights the significance of environmental and systemic elements, and the proposed tailor-made intervention methods for various subgroups could provide sensible implications for digital well being training.
However, a number of limitations needs to be acknowledged when deciphering these findings. First, the cross-sectional design limits causal inference, notably concerning the noticed mediation results. Future longitudinal, multicenter research with goal measures are due to this fact warranted. Second, all variables had been measured utilizing self-report questionnaires, which can introduce recall bias and customary methodology variance. Although Harman’s single-factor take a look at didn’t point out extreme widespread methodology bias, this chance can’t be utterly dominated out. Third, the research was performed in a single college in China, and participation was voluntary, which can restrict generalizability and introduce self-selection bias. In addition, potential class-level clustering results weren’t explicitly accounted for, which can have led to underestimation of ordinary errors. Finally, though knowledge cleansing procedures had been carried out to enhance knowledge high quality, the absence of embedded attention-check objects could have restricted the flexibility to determine insufficient-effort responses.
5. Conclusion
This research used a person-centered method to determine totally different eHL profiles in vocational school college students and to look at the mediating function of HSE between eHL and HPL. The findings point out that vocational school college students could symbolize a digitally deprived subgroup inside the broader college inhabitants and that their eHL is considerably heterogeneous. Three profiles had been recognized, and every profile confirmed totally different ranges of health-promoting behaviors and HSE. Importantly, HSE was statistically verified as a big mediator within the affiliation between eHL profiles and HPL. The findings spotlight the significance of tailoring digital well being training and well being promotion interventions to the precise wants of various eHL subgroups. However, causal inference stays restricted by the cross-sectional design.
Funding Statement
The creator(s) declared that monetary assist was obtained for this work and/or its publication. This research was supported by the Natural Science Foundation of the Jiangsu Higher Education Institutions of China [Grant No. 23KJD460008].
Footnotes
Edited by: Zehra Can, Bayburt University, Türkiye
Data availability assertion
The uncooked knowledge supporting the conclusions of this text will likely be made out there by the authors, with out undue reservation.
Ethics assertion
The research involving people had been accepted by the Ethics Committee of Changzhi Medical College (Approval No.: RT2026001). The research had been performed in accordance with the native laws and institutional necessities. The individuals supplied their written knowledgeable consent to take part on this research.
Author contributions
YN: Conceptualization, Writing – overview & modifying, Data curation, Investigation, Methodology, Project administration, Resources, Visualization, Writing – unique draft. WL: Conceptualization, Data curation, Writing – unique draft, Writing – overview & modifying, Formal evaluation. XJ: Investigation, Methodology, Project administration, Writing – overview & modifying. YH: Investigation, Methodology, Writing – overview & modifying, Funding acquisition, Resources. FA: Funding acquisition, Investigation, Methodology, Formal evaluation, Software, Writing – overview & modifying. DS: Formal evaluation, Writing – overview & modifying, Conceptualization, Supervision.
Conflict of curiosity
The creator(s) declared that this work was performed within the absence of any industrial or monetary relationships that may very well be construed as a possible battle of curiosity.
Generative AI assertion
The creator(s) declared that Generative AI was not used within the creation of this manuscript.
Any different textual content (alt textual content) supplied alongside figures on this article has been generated by Frontiers with the assist of synthetic intelligence and affordable efforts have been made to make sure accuracy, together with overview by the authors wherever potential. If you determine any points, please contact us.
Publisher’s word
All claims expressed on this article are solely these of the authors and don’t essentially symbolize these of their affiliated organizations, or these of the writer, the editors and the reviewers. Any product that could be evaluated on this article, or declare that could be made by its producer, is just not assured or endorsed by the writer.
Supplementary materials
The Supplementary materials for this text might be discovered on-line at:
References
-
1.Koebe P, Bohnet-Joschko S. The impression of digital transformation on inpatient care: blended strategies research. JMIR Public Health Surveill. (2023) 9:e40622. doi: 10.2196/40622,
[DOI] [PMC free article] [PubMed] [Google Scholar] -
2.Natakusumah Okay, Maulina E, Muftiadi A, Purnomo M. Digital transformation of well being high quality companies within the healthcare {industry} throughout disruption and society 5.0 period. Front Public Health. (2022) 10:971486. doi: 10.3389/fpubh.2022.971486,
[DOI] [PMC free article] [PubMed] [Google Scholar] -
3.Wong BLH, Maaß L, Vodden A, Van Kessel R, Sorbello S, Buttigieg S, et al. The daybreak of digital public well being in Europe: implications for public well being coverage and observe. Lancet Reg Health Eur. (2022) 14:100316. doi: 10.1016/j.lanepe.2022.100316
[DOI] [PMC free article] [PubMed] [Google Scholar] -
4.van Kessel R, Hrzic R, O’Nuallain E, Weir E, Wong BLH, Anderson M, et al. Digital well being paradox: worldwide coverage views to handle elevated well being inequalities for individuals residing with disabilities. J Med Internet Res. (2022) 24:e33819. doi: 10.2196/33819
[DOI] [PMC free article] [PubMed] [Google Scholar] - 5.Coetzer JA, Loukili I, Goedhart NS, Ket JCF, Schuitmaker-Warnaar TJ, Zuiderent-Jerak T, et al. The potential and paradoxes of eHealth analysis for digitally marginalised teams: a qualitative meta-review. Soc Sci Med 1982. (2024) 350:116895. doi: 10.1016/j.socscimed.2024.116895 [DOI] [PubMed] [Google Scholar]
-
6.Qiu CS, Lunova T, Greenfield G, Kerr G, Ergüven Ö, Beaney T, et al. Determinants of digital well being literacy: worldwide cross-sectional research. J Med Internet Res. (2025) 27:e66631. doi: 10.2196/66631,
[DOI] [PMC free article] [PubMed] [Google Scholar] -
7.Milanti A, Norman C, Chan DNS, So WKW, Skinner H. eHealth literacy 3.0: updating the Norman and Skinner 2006 mannequin. J Med Internet Res. (2025) 27:e70112. doi: 10.2196/70112,
[DOI] [PMC free article] [PubMed] [Google Scholar] -
8.Yao R, Zhang W, Evans R, Cao G, Rui T, Shen L. Inequities in well being care companies attributable to the adoption of digital well being applied sciences: scoping overview. J Med Internet Res. (2022) 24:e34144. doi: 10.2196/34144,
[DOI] [PMC free article] [PubMed] [Google Scholar] -
9.World Health Organization
Global Strategy on digital Health 2020–2025 (2021). Available on-line at: (accessed January 12, 2026) -
10.Chen J, Xiang H, Jiang P, Yu L, Jing Y, Li F, et al. The function of wholesome way of life within the implementation of regressing suboptimal well being standing amongst school college students in China: A nested case-control research. Int J Environ Res Public Health. (2017) 14:240. doi: 10.3390/ijerph14030240,
[DOI] [PMC free article] [PubMed] [Google Scholar] -
11.Ekelund U, Tarp J, Ding D, Sanchez-Lastra MA, Dalene KE, Anderssen SA, et al. Deaths doubtlessly averted by small modifications in bodily exercise and sedentary time: a person participant knowledge meta-analysis of potential cohort research. Lancet. (2026) 407:339–49. doi: 10.1016/S0140-6736(25)02219-6,
[DOI] [PubMed] [Google Scholar] -
12.Force LM, Kocarnik JM, May ML, Bhangdia Okay, Crist A, Penberthy L, et al. The international, regional, and nationwide burden of most cancers, 1990–2023, with forecasts to 2050: a scientific evaluation for the worldwide burden of illness research 2023. Lancet. (2025) 406:1565–86. doi: 10.1016/S0140-6736(25)01635-6
[DOI] [PMC free article] [PubMed] [Google Scholar] -
13.Li Q, Fang F, Zhang Y, Tu J, Zhu P, Xi L. eHealth literacy and its outcomes amongst postsecondary college students: systematic overview. J Med Internet Res. (2025) 27:e64489. doi: 10.2196/64489,
[DOI] [PMC free article] [PubMed] [Google Scholar] -
14.Chao D-P. Associations between sociodemographic traits, eHealth literacy, and health-promoting way of life amongst college college students in Taipei: cross-sectional validation research of the Chinese model of the eHealth literacy scale. J Med Internet Res. (2024) 26:e52314. doi: 10.2196/52314,
[DOI] [PMC free article] [PubMed] [Google Scholar] -
15.Jiang S, Ng JYY, Choi SM, Ha AS. Relationships amongst eHealth literacy, bodily literacy, and bodily exercise in Chinese college college students: cross-sectional research. J Med Internet Res. (2024) 26:e56386. doi: 10.2196/56386,
[DOI] [PMC free article] [PubMed] [Google Scholar] -
16.Kim Okay, Shin S, Kim S, Lee E. The relation between eHealth literacy and health-related behaviors: systematic overview and meta-analysis. J Med Internet Res. (2023) 25:e40778. doi: 10.2196/40778,
[DOI] [PMC free article] [PubMed] [Google Scholar] -
17.Iyamu I, Gorun P, Chang H-J, Sierra-Rosales R, Haag D, Pedersen H, et al. Development and validation of a revised multidimensional digital well being literacy scale: secondary evaluation utilizing cross-sectional knowledge from the 2022 GetCheckedOnline neighborhood survey in British Columbia, Canada. J Med Internet Res. (2025) 27:e78008. doi: 10.2196/78008,
[DOI] [PMC free article] [PubMed] [Google Scholar] -
18.Stellefson M, Hanik B, Chaney JD, Tennant B. Analysis of ehealth search views amongst feminine school college students within the well being professions utilizing Q methodology. J Med Internet Res. (2012) 14:e60. doi: 10.2196/jmir.1969,
[DOI] [PMC free article] [PubMed] [Google Scholar] - 19.Bandura A. Social cognitive principle of self-regulation. Organ Behav Hum Decis Process. (1991) 50:248–87. doi: 10.1016/0749-5978(91)90022-L [DOI] [Google Scholar]
- 20.Bandura A. “Self-efficacy”. In: Weiner IB, Craighead WE, editors. The Corsini Encyclopedia of Psychology. New York, NY, USA: Wiley; (2010). p. 1–3. [Google Scholar]
-
21.Rabenbauer LM, Mevenkamp N. Factors within the effectiveness of e-health interventions for persistent Back ache: how self-efficacy mediates e-health literacy and wholesome habits. Telemed E-Health. (2021) 27:184–92. doi: 10.1089/tmj.2019.0301,
[DOI] [PubMed] [Google Scholar] -
22.Liu HX, Chow BC, Hassel H, Huang YW, Liang W, Wang RB. Prospective affiliation of eHealth literacy and well being literacy with bodily exercise amongst Chinese school college students: a a number of mediation evaluation. Front Public Health. (2024) 12:1275691. doi: 10.3389/fpubh.2024.1275691,
[DOI] [PMC free article] [PubMed] [Google Scholar] -
23.Schumacker RE, Lomax RG. A newbie’s information to structural equation modeling. Technometrics. (2004) 47:522–2. doi: 10.1198/tech.2005.s328, 37536082
[DOI] [Google Scholar] - 24.Meyer JP, Morin AJS. An individual-centered method to dedication analysis: principle, analysis, and methodology. J Organ Behav. (2016) 37:584–612. doi: 10.1002/job.2085 [DOI] [Google Scholar]
-
25.Yu Y, Liang Z, Zhou Q, Tuersun Y, Liu S, Wang C, et al. Decomposition and comparative evaluation of urban-rural disparities in eHealth literacy amongst Chinese college college students: cross-sectional research. J Med Internet Res. (2025) 27:e63671. doi: 10.2196/63671,
[DOI] [PMC free article] [PubMed] [Google Scholar] -
26.Norman CD, Skinner HA. eHEALS: the eHealth literacy scale. J Med Internet Res. (2006) 8:e27. doi: 10.2196/jmir.8.4.e27,
[DOI] [PMC free article] [PubMed] [Google Scholar] -
27.Xu RH, Zhou L, Lu SY, Wong EL, Chang J, Wang D. Psychometric validation and cultural adaptation of the simplified Chinese eHealth literacy scale: cross-sectional research. J Med Internet Res. (2020) 22:e18613. doi: 10.2196/18613,
[DOI] [PMC free article] [PubMed] [Google Scholar] -
28.Chilton JM, Gosselin KP, Haas BK. Development of the self-rated skills for well being practices–adolescent model: a self-efficacy measure. J Nurs Meas. (2018) 26:134–41. doi: 10.1891/1061-3749.26.1.134,
[DOI] [PubMed] [Google Scholar] -
29.Ispriantari A, Lee H, Kim S, Lee H, Lee A, Park C. Determinants of health-promoting behaviors amongst Indonesian adolescents residing in little one welfare establishments: cross-sectional survey research. JMIR Public Health Surveill. (2025) 11:e75024. doi: 10.2196/75024,
[DOI] [PMC free article] [PubMed] [Google Scholar] -
30.Velez-Gomez P, Dodd SL, Quick M, Niraula SR. Adaptation and validation of a scale to measure self-efficacy for well being practices in adolescents. J Pediatr Nurs. (2025) 85:143–50. doi: 10.1016/j.pedn.2025.07.022,
[DOI] [PubMed] [Google Scholar] -
31.Walker SN, Sechrist KR, Pender NJ. The health-promoting way of life profile: growth and psychometric traits. Nurs Res. (1987) 36:76–81. doi: 10.1097/00006199-198703000-00002,
[DOI] [PubMed] [Google Scholar] - 32.Cao WJ, Guo Y, Ping WW, Zheng JZ. Development and psychometric assessments of a Chinese model of the HPLP-II scales. Chin J Dis Control Prev. (2016) 20:286–9. doi: 10.16462/j.cnki.zhjbkz.2016.03.018 [DOI] [Google Scholar]
-
33.Podsakoff PM, MacKenzie SB, Lee J-Y, Podsakoff NP. Common methodology biases in behavioral analysis: a important overview of the literature and beneficial cures. J Appl Psychol. (2003) 88:879–903. doi: 10.1037/0021-9010.88.5.879,
[DOI] [PubMed] [Google Scholar] - 34.Kock F, Berbekova A, Assaf G. Understanding and managing the specter of widespread methodology bias: detection, prevention and management. Tour Manag. (2021) 86:104330. doi: 10.1016/j.tourman.2021.104330 [DOI] [Google Scholar]
- 35.Nylund-Gibson Okay, Grimm RP, Masyn KE. Prediction from latent lessons: an indication of various approaches to incorporate distal outcomes in combination fashions. Struct Equ Model Multidiscip J. (2019) 26:967–85. doi: 10.1080/10705511.2019.1590146 [DOI] [Google Scholar]
-
36.Kim SY. Determining the variety of latent lessons in single- and multiphase progress combination fashions. Struct Equ Model Multidiscip J. (2014) 21:263–79. doi: 10.1080/10705511.2014.882690,
[DOI] [PMC free article] [PubMed] [Google Scholar] - 37.Williams GA, Kibowski F. “Latent class analysis and latent profile analysis”. In: Jason LA, Glenwick DS, editors. Handbook of Methodological Approaches to Community-Based Research. New York, NY, USA: Oxford University Press; (2016). p. 143–51. [Google Scholar]
-
38.Tönnies T, Schlesinger S, Lang A, Kuss O. Mediation evaluation in medical analysis. Dtsch Arzteblatt Int. (2023) 120:681–7. doi: 10.3238/arztebl.m2023.0175,
[DOI] [PMC free article] [PubMed] [Google Scholar] -
39.Richtering SS, Hyun Okay, Neubeck L, Coorey G, Chalmers J, Usherwood T, et al. eHealth literacy: predictors in a inhabitants with moderate-to-high cardiovascular danger. JMIR Hum Factors. (2017) 4:e4. doi: 10.2196/humanfactors.6217,
[DOI] [PMC free article] [PubMed] [Google Scholar] -
40.Canady BE, Larzo M. Overconfidence in managing well being considerations: the dunning-Kruger impact and well being literacy. J Clin Psychol Med Settings. (2023) 30:460–8. doi: 10.1007/s10880-022-09895-4,
[DOI] [PMC free article] [PubMed] [Google Scholar] - 41.Bertolazzi A, Lombi L, Lovari A, Ducci G, D’Ambrosi L. I do know that I do know: on-line well being info looking for, self-care and the overconfidence effect1,2. Sociol Forum. (2023) 38:793–812. doi: 10.1111/socf.12925 [DOI] [Google Scholar]
-
42.Lackner S, Francisco F, Mendonça C, Mata A, Gonçalves-Sá J. Intermediate ranges of scientific data are related to overconfidence and detrimental attitudes in direction of science. Nat Hum Behav. (2023) 7:1490–501. doi: 10.1038/s41562-023-01677-8,
[DOI] [PubMed] [Google Scholar] -
43.Taba M, Allen TB, Caldwell PHY, Skinner SR, Kang M, McCaffery Okay, et al. Adolescents’ self-efficacy and digital well being literacy: a cross-sectional blended strategies research. BMC Public Health. (2022) 22:1223. doi: 10.1186/s12889-022-13599-7,
[DOI] [PMC free article] [PubMed] [Google Scholar] -
44.Bahg G, Sloutsky VM, Turner BM. Algorithmic personalization of knowledge could cause inaccurate generalization and overconfidence. J Exp Psychol Gen. (2025) 154:2503–22. doi: 10.1037/xge0001763,
[DOI] [PMC free article] [PubMed] [Google Scholar] - 45.Wang YM, Huang FQ, Chen KZ, Zheng TH. The 2024-2035 grasp plan on constructing China into a number one nation in training promotes the high-quality growth of vocational training industry-education integration. Chin J ICT Educ. (2025) 31:12–20. [Google Scholar]
-
46.Bao X, Chen D, Shi L, Xia Y, Shi Z, Wang D. The relationship between COVID-19-related prevention cognition and wholesome way of life behaviors amongst college college students: mediated by e-health literacy and self-efficacy. J Affect Disord. (2022) 309:236–41. doi: 10.1016/j.jad.2022.04.044,
[DOI] [PMC free article] [PubMed] [Google Scholar] -
47.Chen X, Hay JL, Waters EA, Kiviniemi MT, Biddle C, Schofield E, et al. Health literacy and use and Trust in Health Information. J Health Commun. (2018) 23:724–34. doi: 10.1080/10810730.2018.1511658,
[DOI] [PMC free article] [PubMed] [Google Scholar] -
48.Yuen E, Winter N, Savira F, Huggins CE, Nguyen L, Cooper P, et al. Digital well being literacy and its affiliation with sociodemographic traits, well being useful resource use, and well being outcomes: speedy overview. Interact J Med Res. (2024) 13:e46888. doi: 10.2196/46888,
[DOI] [PMC free article] [PubMed] [Google Scholar] - 49.Nutbeam D. Health literacy as a public well being aim: a problem for up to date well being training and communication methods into the twenty first century. Health Promot Int. (2000) 15:259–67. doi: 10.1093/heapro/15.3.259 [DOI] [Google Scholar]
-
50.Nutbeam D, Lloyd JE. Understanding and responding to well being literacy as a social determinant of well being. Annu Rev Public Health. (2021) 42:159–73. doi: 10.1146/annurev-publhealth-090419-102529,
[DOI] [PubMed] [Google Scholar] -
51.Ienca M, Gounder C, Emanuel EJ. The paradox of belief in well being care within the age of social media. Lancet. (2026) 407:909–14. doi: 10.1016/S0140-6736(25)02556-5,
[DOI] [PubMed] [Google Scholar] -
52.Yang A. Adolescent Cocreation in digital well being: from passive topics to energetic stakeholders. J Med Internet Res. (2025) 27:e70020. doi: 10.2196/70020,
[DOI] [PMC free article] [PubMed] [Google Scholar] -
53.Glanz Okay, Bishop DB. The function of behavioral science principle in growth and implementation of public well being interventions. Annu Rev Public Health. (2010) 31:399–418. doi: 10.1146/annurev.publhealth.012809.103604,
[DOI] [PubMed] [Google Scholar] -
54.Chen SC-I, Yu M, Yu Y, Wang R, Zhu Z, Liu S, et al. The impression of e-health literacy on danger notion amongst college college students. Healthcare. (2025) 13:265. doi: 10.3390/healthcare13030265,
[DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This part collects any knowledge citations, knowledge availability statements, or supplementary supplies included on this article.
Supplementary Materials
Data Availability Statement
The uncooked knowledge supporting the conclusions of this text will likely be made out there by the authors, with out undue reservation.
This web page was created programmatically, to learn the article in its unique location you’ll be able to go to the hyperlink bellow:
https://pmc.ncbi.nlm.nih.gov/articles/PMC13415940/
and if you wish to take away this text from our website please contact us

