Abstract
Background
Digital health literacy is increasingly relevant to health behavior among college students, but its relationships with sports participation and associated psychological factors remain incompletely understood. Guided by social cognitive theory and an adaptive cognitive-processing perspective, this study examined a prespecified cross-sectional path model involving digital health literacy, cognitive flexibility, physical activity self-efficacy, and sports participation.
Methods
A cross-sectional survey was conducted among 600 Chinese college students. Participants completed the DHLI 2.0, CFI, PASE, and PARS-3. Pearson correlation analysis, structural equation modeling, and bootstrap estimation of model-specific path-coefficient products using 5,000 resamples were performed with SPSS 26.0 and AMOS.
Results
Digital health literacy was positively associated with cognitive flexibility, physical activity self-efficacy, and sports participation (all p < 0.001). The structural model demonstrated a good fit to the data (χ2/df = 1.984, RMSEA = 0.016, SRMR = 0.033, CFI = 0.986, TLI = 0.993). Within the prespecified model, digital health literacy was positively associated with cognitive flexibility (β = 0.504), physical activity self-efficacy (β = 0.213), and sports participation (β = 0.191). The three-path coefficient product was small, although its bootstrap confidence interval excluded zero (standardized estimate = 0.011, 95% CI [0.002, 0.031]).
Conclusion
The observed covariance pattern was consistent with the prespecified cross-sectional path model involving cognitive flexibility and physical activity self-efficacy. However, because all variables were measured at a single time point, the findings do not establish temporal ordering or causal mediation. Longitudinal and experimental studies are needed to examine temporal ordering and causal relationships among these variables.
Keywords: cognitive flexibility, college students, digital health literacy, physical activity, self-efficacy, sports participation
1. Introduction
Digital technology is profoundly changing the health behaviors of college students (Jiang et al., 2024). The World Health Organization reports a widespread decline in physical activity among adolescents and young adults globally, with sedentary behavior and digital dependence becoming significant health risks (Bull et al., 2020). Insufficient physical activity remains a prominent issue among college students. With the rapid popularization of mobile internet, smart wearable devices, and online fitness platforms, an increasing number of university students are accessing health and exercise information through digital media, such as using fitness apps to record exercise data, learning fitness techniques through short video platforms, or participating in online sports interactions via social media (Cho et al., 2014; Singh et al., 2025). Against this backdrop, digital health literacy is gradually becoming an important factor in promoting healthy behaviors among young people (Ridwan et al., 2022; Zhu et al., 2024). Existing research shows that digital health literacy not only relates to an individual’s ability to access and understand health information but also further influences their health decisions and behaviors (Dadaczynski et al., 2021; Norman and Skinner, 2006). For college students, higher digital health literacy can help them more effectively filter exercise information, discern the authenticity of online health content, and enhance their initiative in participating in physical activities (Bayati et al., 2018).
The relationship between digital health literacy and health promotion has received widespread attention (Liu H. X. et al., 2024). Studies have found significant associations between digital health literacy and physical activity, healthy lifestyles, and mental health (Kühn et al., 2022; Li and Yang, 2024). Particularly among college students, individuals with higher digital health literacy tend to develop positive health behavior patterns, including regular exercise, healthy eating, and proactive health management (Paige et al., 2018). However, existing research focuses primarily on the direct impact of digital health literacy on general health behaviors, lacking in-depth exploration of the underlying psychological mechanisms by which it affects college students’ participation in sports. In fact, whether college students truly participate in sports depends not only on their ability to acquire sports knowledge but also on their cognitive processing abilities and behavioral beliefs (Acar et al., 2024). As defined by Berkman and colleagues, health literacy (HL) is “the degree to which individuals can obtain, process, understand, and communicate about health-related information needed to make informed health decisions” (Berkman et al., 2010). Therefore, explaining the impact of digital health literacy on sports participation solely from the perspective of “information acquisition” remains limited.
First proposed in cognitive psychology, cognitive flexibility emphasizes an individual’s ability to adjust cognitive strategies, shift thinking patterns, and adapt to new situations in complex environments (Dennis and Vander Wal, 2010). In a digital environment, individuals are bombarded with a vast amount of health and exercise information daily. A high level of digital health literacy can help college students process complex information more effectively and enhance their cognitive flexibility (Hohl and Dolcos, 2024; Liu L. et al., 2024). Existing research has shown a significant association between cognitive flexibility and health behavior, self-regulation, and proactive decision-making (Martin and Rubin, 1995; Tong et al., 2023). College students with higher cognitive flexibility are more likely to adjust their exercise plans, overcome barriers to exercise, and try different forms of physical activity (Shokrkon and Nicoladis, 2022; Tong et al., 2023). Accordingly, cognitive flexibility was examined as a possible statistical link between digital health literacy and sports participation.
Furthermore, physical activity self-efficacy is also an important psychological factor influencing college students’ sports participation. Based on Bandura’s social cognitive theory, self-efficacy refers to an individual’s confidence in their ability to perform a certain behavior (Bandura, 2013). In the field of sports, physical activity self-efficacy is mainly reflected in an individual’s belief that they can persevere in exercising even when faced with situations such as insufficient time, academic pressure, or physical fatigue. Numerous studies have confirmed a significant positive correlation between physical activity self-efficacy and sports participation (Dishman et al., 2005; Shudayfat et al., 2023). Meanwhile, individuals with higher cognitive flexibility typically possess stronger problem-solving and behavioral regulation abilities, which contribute to increased control and confidence (Zhang et al., 2026). Self-efficacy is recognized as an important component of self-management and a key factor in promoting behavioral change and maintaining optimal functioning (Huang et al., 2026). In the present model, physical activity self-efficacy was therefore positioned as the more proximal correlate of sports participation.
Existing research consistently demonstrates that cognitive flexibility enhances an individual’s sense of behavioral control and self-regulation, factors closely related to the formation of self-efficacy. A study by Bebeley et al. (2017) in a university student population found that exercise self-efficacy plays a significant mediating role between exercise motivation and physical activity, indicating that an individual’s perception of their own exercise capabilities further influences actual exercise behavior. Similarly, Mohammadkhani et al. (2022) also pointed out that cognitive flexibility can further promote positive behavioral performance by improving an individual’s psychological adaptability.
Behavioral participation depends not only on the acquisition of external information but also on an individual’s cognitive interpretation and self-regulatory beliefs (Bandura, 2013). Digital health literacy can first serve as a cognitive resource, enabling individuals to assess, organize, and adaptively process digital health information. These associations may be reciprocal, and physical activity itself may strengthen self-efficacy or cognitive flexibility. The proposed order should therefore be understood as a theory-informed specification rather than an empirically established temporal process.
Social cognitive theory provides the basis for treating physical activity self-efficacy as a behavior-specific belief closely associated with participation (Bandura, 2013). The adaptive cognitive-processing perspective, informed by the CFI framework, supports examining cognitive flexibility as a general capacity for generating alternatives and perceiving control (Dennis and Vander Wal, 2010). Neither framework directly specifies that digital health literacy must precede cognitive flexibility, that ordering is an integrative extension proposed in this study.
Digital health literacy was conceptualized as a domain-specific informational resource, cognitive flexibility as a broader adaptive processing tendency, physical activity self-efficacy as a behavior-specific belief, and sports participation as the behavioral outcome. This ordering is plausible, but alternatives are also reasonable. Cognitively flexible students may navigate digital information more effectively, and physically active students may gain both efficacy and digital health experience (Woo et al., 2024; Yu et al., 2023). Therefore, the present study proposes a sequential “cognition-belief-behavior” pathway to explain how digital health literacy may relate to sports participation among college students.
2. Theoretical framework and hypotheses
Although prior research has examined pairwise associations among digital health literacy, cognitive flexibility, physical activity self-efficacy, and sports participation, evidence integrating all four constructs remains limited. Therefore, this study evaluated hypothesized indirect association models. The association between digital health literacy and sports participation was examined, and specific indirect associations were estimated through cognitive flexibility and physical activity self-efficacy.
Building on the findings of previous research, the following hypotheses are proposed for this study (See Figure 1):
Figure 1.
Hypothetical model.
H1: Digital health literacy is positively associated with sports participation.
H2: Digital health literacy has a positive cross-sectional indirect association with sports participation through cognitive flexibility.
H3: Digital health literacy has a positive cross-sectional indirect association with sports participation through physical activity self-efficacy.
H4: Digital health literacy has a positive ordered cross-sectional indirect association with sports participation through cognitive flexibility and then physical activity self-efficacy.
3. Methods and measures
3.1. Participants
All data were collected in May 2026 through a combination of online and offline methods. Participants were recruited from Guangdong University of Finance and Economics. To improve the representativeness of the sample and ensure coverage of college students at different learning stages, this study employed a stratified random sampling method based on grade level, selecting college students from four grades as research subjects and distributing a total of 650 questionnaires. Eligibility criteria included being a current undergraduate student, aged 18 or above, able to read Chinese, and providing informed consent. Of the 639 returned questionnaires, 39 were excluded. These included nine questionnaires with missing answers to core scale items, 17 questionnaires with response times significantly shorter than reasonable completion times, and 13 questionnaires with obvious patterns or inconsistencies in the responses. The final analysis sample consisted of 600 participants (valid response rate = 93.9%). Participants included 296 males (49.33%) and 304 females (50.67%); 309 from rural areas and 291 from urban areas. The grade distribution was as follows: 149 first-year college students (24.83%), 148 s-year college students (24.67%), 150 third-year college students (25.00%), and 153 fourth-year college students (25.50%). The study was approved by the Guangdong University of Finance and Economics ethics committee (GDUFE202109-01) and conducted in accordance with the Declaration of Helsinki. Before beginning, participants received information about the study purpose, procedures, voluntary participation, anonymity, and right to discontinue. Consent was indicated before questionnaire completion; the ethics approval did not require a signed consent form for this anonymous, minimal-risk survey.
3.2. Measures
This study included demographic components (including subject, gender, region, and grade) and four standardized questionnaires to measure digital health literacy, cognitive flexibility, self-efficacy, and sports participation: the Digital Health Literacy Questionnaire (DHLI), the Cognitive Flexibility Inventory (CFI), the Physical Activity Self-Efficacy Scale (PASE), and the Physical Activity Rating Scale-3 (PARS-3). All questionnaires were previously drafted and validated in English. In this study, all questionnaires were translated from English to Chinese by a bilingual expert in sports psychology and from Chinese to English by a Chinese expert in English linguistics. The two experts then agreed on the final Chinese version of the questionnaire items in a discussion. This Chinese translation process followed the Brislin Standard for forward-backward translation (Brislin, 1970).
3.2.1. Digital Health Literacy Instrument 2.0 Short Form
This survey tool was originally developed by van der Vaart and Drossaert (2017) using the Digital Health Literacy Tool 2.0 Short Form (DHLI 2.0). This study uses an updated version of the Digital Health Literacy Tool (DHLI 2.0) short form. The questionnaire contains 16 items across 8 dimensions: Operational skills (e.g., “Use the buttons and links or hyperlinks on websites?” easy or difficult), Navigation skills (e.g., “You lose track of where you are on a website or in an app?” how often), Information searching (e.g., “Find exactly what you are looking for?” easy or difficult), Evaluating reliability (e.g., “Determine whether the information is reliable or not?” easy or difficult), Determining relevance (e.g., “Determine whether the information found applies to you?” easy or difficult), Generating content (e.g., “Express your opinions, thoughts, or feelings in writing?” easy or difficult), Using security measures (e.g., “Create and remember strong passwords?” easy or difficult) and Protecting privacy (e.g., “do you include sensitive personal information in your message?” How often). Each item is rated on a 4-point scale (“Very difficult” = 1 to “Very easy” = 4), with all responses converted so that higher scores reflect higher levels of digital health literacy. For the present structural model, all 16 items were specified as indicators of one overarching digital health literacy factor because the hypotheses concerned overall digital health literacy. The measurement tool demonstrated good reliability and validity, with a Cronbach’s alpha of 0.867, which meets the standard of greater than 0.7.
3.2.2. The Cognitive Flexibility Inventory
The Cognitive Flexibility Scale (CFI) was originally developed and validated by Dennis and Vander Wal (2010) to measure the type of cognitive flexibility necessary for individuals to successfully challenge and replace maladaptive thoughts with more balanced and adaptive thinking (Dennis and Vander Wal, 2010). The scale consists of 20 items, comprising two dimensions: Alternatives and Control subscales. A 7-point Likert scale was used for scoring, ranging from “Strongly Disagree = 1” to “Strongly Agree = 7”. After back-scoring, the scores of all items were summed to obtain the total score. A higher total score indicates a higher level of cognitive flexibility. Because the structural hypothesis concerned overall cognitive flexibility, the present study specified the 20 items as indicators of one overarching latent factor. The CFI has a reliable two-factor structure, good to excellent internal consistency, and high 7-week test–retest reliability. In this study, the CFI also exhibited good internal consistency, with a Cronbach’s alpha coefficient of 0.959.
3.2.3. Physical Activity Self-Efficacy Scale
This study used the Chinese version of the Physical Activity Self-Efficacy Scale (PASE), developed by Motl et al. (2000) and revised by Chen et al. (2019), to measure college students’ beliefs about physical activity (Chen et al., 2019; Motl et al., 2000). The scale contains eight questions, such as “I can exercise most of the week” and “I can exercise even at home.” It is a one-dimensional scale using a 5-point Likert scale (“Strongly Disagree” = 1 to “Strongly Agree” = 5). The total score reflects the college students’ self-efficacy regarding physical activity. A higher total score indicates better self-efficacy. In this study, the Cronbach’s α coefficient of the scale was 0.798.
3.2.4. Physical Activity Rating Scale-3
Sports participation was measured using the Physical Activity Rating Scale (PARS-3) revised by Liang and Lin (1994). The scale comprises three dimensions: frequency, time, and intensity, all self-rated by the participants. Each item was scored from 1 to 5. For descriptive reporting, the published composite score was calculated as intensity × (duration − 1) × frequency (range 0–100), with scores of 19 or lower, 20–42, and 43 or higher conventionally classified as low, moderate, and high activity, respectively. In CFA and SEM, intensity, duration, and frequency were specified as reflective indicators of a latent sports participation factor. Table 1 reports the composite descriptively, whereas the CFA and structural paths refer to the latent factor. Furthermore, the reflective specification was supported in this sample by standardized loadings of 0.676–0.841, composite reliability of 0.813, and average variance extracted of 0.593. Cronbach’s alpha was 0.725.
Table 1.
Means, standard deviations, and correlations among the main variables.
| Variable | M | SD | DHLI | CFI | PASE | PARS-3 |
|---|---|---|---|---|---|---|
| DHLI | 1.88 | 0.489 | 1 | |||
| CFI | 3.865 | 0.929 | 0.457*** | 1 | ||
| PASE | 3.259 | 0.918 | 0.232*** | 0.206*** | 1 | |
| PARS-3 | 27.190 | 13.397 | 0.247*** | 0.237*** | 0.201*** | 1 |
***p < 0.001.
3.3. Statistical analysis
Analyses were conducted in SPSS 26.0 and AMOS 26.0. After questionnaire screening, the analytic dataset contained complete responses for all modeled items. No item-level missing values were imputed. Responses on Likert-type items were approximated as continuous variables for estimation. Univariate skewness and kurtosis values were inspected and fell within the prespecified acceptable ranges. Confirmatory factor analysis (CFA) evaluated the study-specific four-factor measurement model, with one first-order latent factor representing each focal construct. Models were estimated by maximum likelihood (MLE). Reliability was summarized with Cronbach’s alpha, and convergent validity with standardized loadings, average variance extracted (AVE), and composite reliability (CR). Discriminant validity was evaluated using the Fornell-Larcker criterion. Pearson correlations were reported descriptively and were not treated as prerequisites for testing indirect associations. The structural equation modeling (SEM) included field of study, region, gender, and grade as covariates of sports participation. Specific indirect associations were evaluated with 5,000 bias-corrected bootstrap resamples and 95% confidence intervals. An interval excluding zero was interpreted as statistical evidence of an indirect association.
4. Results
4.1. Common method bias diagnostic
Harman’s single-factor diagnostic extracted a first unrotated factor accounting for 28.38% of total variance, below the prespecified 40% reference value. This result indicates that no single factor dominated the item covariance, but the diagnostic cannot establish the absence of common method bias (Podsakoff et al., 2003). Residual bias associated with self-report and common measurement context therefore cannot be ruled out.
4.2. Confirmatory factor analysis of digital health literacy, cognitive flexibility, physical activity self-efficacy, and sports participation
Confirmatory factor analysis was performed using AMOS, and the results are shown in Table 2. All fit indices of the four-factor model met the recommended criteria (χ2 = 1,584.148, df = 1,028, χ2/df = 1.541, TLI = 0.948, CFI = 0.937, RMSEA = 0.025, SRMR = 0.031), indicating that these four variables have good measurement model fit (See Figure 2). These results support the adequacy of the measurement specification used in this sample. They do not demonstrate that the multidimensional DHLI 2.0 or CFI is inherently unidimensional.
Table 2.
Confirmatory factor fit indices of digital health literacy, cognitive flexibility, physical activity self-efficacy, and sports participation.
| Common indicators | χ2/df | RMSEA | SRMR | GFI | IFI | TLI | CFI |
|---|---|---|---|---|---|---|---|
| Standard | <5 | <0.08 | <0.08 | >0.8 | >0.9 | >0.9 | >0.9 |
| Numerical values | 1.541 | 0.025 | 0.031 | 0.945 | 0.967 | 0.948 | 0.937 |
The model contains one study-specific first-order latent factor per focal construct; fit does not establish that the source DHLI 2.0 or CFI is unidimensional.
Figure 2.
Confirmatory factor analysis of digital health literacy, cognitive flexibility, physical activity self-efficacy, and sports participation.
4.3. Convergent validity
Standardized item loadings were 0.656–0.863. For each study-specific latent factor, average variance extracted (AVE) was at least 0.50 and composite reliability (CR) exceeded 0.70, supporting convergent validity for the specified measurement model (See Table 3).
Table 3.
Standardized loadings, composite reliability, and average variance extracted for the study-specific measurement model.
| Variable | Items | Standardized loading | CR | AVE |
|---|---|---|---|---|
| Digital health literacy | A1 | 0.821 | 0.958 | 0.589 |
| A2 | 0.746 | |||
| A3 | 0.745 | |||
| A4 | 0.733 | |||
| A5 | 0.774 | |||
| A6 | 0.861 | |||
| A7 | 0.837 | |||
| A8 | 0.716 | |||
| A9 | 0.729 | |||
| A10 | 0.837 | |||
| A11 | 0.863 | |||
| A12 | 0.761 | |||
| A13 | 0.701 | |||
| A14 | 0.735 | |||
| A15 | 0.691 | |||
| A16 | 0.698 | |||
| Cognitive flexibility | B1 | 0.731 | 0.959 | 0.539 |
| B2 | 0.741 | |||
| B3 | 0.744 | |||
| B4 | 0.736 | |||
| B5 | 0.732 | |||
| B6 | 0.74 | |||
| B7 | 0.759 | |||
| B8 | 0.717 | |||
| B9 | 0.751 | |||
| B10 | 0.739 | |||
| B11 | 0.718 | |||
| B12 | 0.747 | |||
| B13 | 0.754 | |||
| B14 | 0.743 | |||
| B15 | 0.722 | |||
| B16 | 0.708 | |||
| B17 | 0.732 | |||
| B18 | 0.719 | |||
| B19 | 0.724 | |||
| B20 | 0.728 | |||
| Physical activity self-efficacy | C1 | 0.736 | 0.909 | 0.557 |
| C2 | 0.754 | |||
| C3 | 0.743 | |||
| C4 | 0.775 | |||
| C5 | 0.656 | |||
| C6 | 0.800 | |||
| C7 | 0.785 | |||
| C8 | 0.714 | |||
| Sports participation | D1 | 0.676 | 0.813 | 0.593 |
| D2 | 0.841 | |||
| D3 | 0.784 |
AVE = average variance extracted; CR = composite reliability. Values refer to the study-specific overarching latent factors used in this analysis.
4.4. Discriminant validity
The discriminant validity analysis indicated that all inter-construct correlation coefficients were lower than the square root of the AVE for their respective constructs. These results demonstrate sufficient differentiation among the latent variables, confirming acceptable discriminant validity of the measurement model (See Table 4).
Table 4.
Fornell-Larcker discriminant-validity assessment.
| Variable | Sports participation | Physical activity self-efficacy | Cognitive flexibility | Digital health literacy |
|---|---|---|---|---|
| Sports participation | 0.770 | |||
| Physical activity self-efficacy | 0.258 | 0.747 | ||
| Cognitive flexibility | 0.285 | 0.237 | 0.734 | |
| Digital health literacy | 0.31 | 0.276 | 0.501 | 0.767 |
The diagonal line corresponds to the square root of the dimension.
4.5. Correlation analysis of variables
Pearson correlations showed positive bivariate associations among all focal variables. Digital health literacy was associated with cognitive flexibility (r = 0.457, p < 0.001), physical activity self-efficacy (r = 0.232, p < 0.001), and sports participation (r = 0.247, p < 0.001). Cognitive flexibility was associated with physical activity self-efficacy (r = 0.206, p < 0.001) and sports participation (r = 0.237, p < 0.001), and physical activity self-efficacy was associated with sports participation (r = 0.201, p < 0.001) (See Table 1).
4.6. Analysis of hypothesized indirect associations
Significant correlations were found between digital health literacy, cognitive flexibility, physical activity self-efficacy, and sports participation, meeting the statistical requirements for further mediation effect analysis of cognitive flexibility and self-efficacy. A structural equation model (SEM) was constructed with digital health literacy as the independent variable, sports participation as the dependent variable, and cognitive flexibility and physical activity self-efficacy as intermediate variables. Table 5 and Figure 3 show the overall fit indices, indicating that the proposed path analysis model demonstrated an acceptable overall fit to the data. χ2/df = 1.984, RMSEA = 0.016, SRMR = 0.033, GFI = 0.930, TLI = 0.993, IFI = 0.982, CFI = 0.986. In summary, these results indicate a high degree of model fit, a reasonable path structure between variables, and data that support the proposed theoretical framework. Gender, grade, region, and subject were entered as covariates in the structural model to account for potential demographic influences on sports participation.
Table 5.
Fit indices for the hypothesized cross-sectional indirect-association model.
| Common indicators | χ2/df | RMSEA | SRMR | GFI | IFI | TLI | CFI |
|---|---|---|---|---|---|---|---|
| Standard | <5 | <0.08 | <0.08 | >0.8 | >0.9 | >0.9 | >0.9 |
| Numerical values | 1.984 | 0.016 | 0.033 | 0.930 | 0.982 | 0.993 | 0.986 |
Figure 3.
Indirect-association model of cognitive flexibility and physical activity self-efficacy between digital health literacy and sports participation.
4.7. Model path coefficient test
Table 6 presents the structural estimates. Digital health literacy was positively associated with cognitive flexibility (β = 0.504, p < 0.001), physical activity self-efficacy (β = 0.213, p < 0.001), and sports participation (β = 0.191, p = 0.002). Cognitive flexibility was associated with physical activity self-efficacy (β = 0.134, p = 0.012) and sports participation (β = 0.150, p = 0.003). Physical activity self-efficacy was also associated with sports participation (β = 0.173, p = 0.002). Except for the digital health literacy-cognitive flexibility association, these standardized coefficients were small. The standardized focal-variable solution corresponded to approximate explained-variance values of R2 = 0.254 for cognitive flexibility, 0.091 for physical activity self-efficacy, and 0.147 for sports participation. None of the four demographic covariates was statistically associated with sports participation (all p > 0.05). The cross-sectional association specified in Hypothesis 1 was supported.
Table 6.
Structural associations in the hypothesized model.
| Path | β | B | S.E. | C.R. | p |
|---|---|---|---|---|---|
| Digital health literacy → Cognitive flexibility | 0.504 | 1.015 | 0.117 | 8.982 | *** |
| Cognitive flexibility → Self-efficacy | 0.134 | 0.160 | 0.049 | 2.438 | 0.012 |
| Digital health literacy → Self-efficacy | 0.213 | 0.406 | 0.113 | 3.53 | *** |
| Self-efficacy → Sports participation | 0.173 | 0.215 | 0.069 | 3.083 | 0.002 |
| Digital health literacy → Sports participation | 0.191 | 0.447 | 0.146 | 3.063 | 0.002 |
| Cognitive flexibility → Sports participation | 0.150 | 0.184 | 0.064 | 2.697 | 0.003 |
| Subject → Sports participation | −0.067 | −0.097 | 0.059 | −1.350 | 0.147 |
| Region → Sports participation | −0.034 | −0.075 | 0.094 | −0.514 | 0.513 |
| Gender → Sports participation | −0.041 | −0.064 | 0.094 | −0.091 | 0.889 |
| Grade → Sports participation | 0.036 | 0.053 | 0.042 | 0.763 | 0.421 |
β = standardized estimate; B = unstandardized estimate; C.R. = critical ratio. ***p < 0.001.
4.8. Bootstrap indirect-association estimates
Table 7 reports standardized specific indirect associations based on 5,000 bias-corrected bootstrap resamples and 95% confidence intervals. The estimates quantify how contemporaneous covariance was partitioned under the specified model, and they do not establish a process unfolding over time.
Table 7.
Bias-corrected bootstrap estimates of cross-sectional indirect associations.
| Path | Effect | Lower | Upper | p |
|---|---|---|---|---|
| Digital health literacy → Cognitive flexibility → Sports participation | 0.076 | 0.018 | 0.145 | 0.01 |
| Digital health literacy → Physical activity self-efficacy → Sports participation | 0.036 | 0.008 | 0.084 | 0.006 |
| Digital health literacy → Cognitive flexibility → Physical activity self-efficacy → Sports participation | 0.011 | 0.002 | 0.031 | 0.012 |
| Total effect | 0.312 | 0.191 | 0.412 | 0.001 |
| Direct effect | 0.189 | 0.038 | 0.322 | 0.014 |
| Indirect effect | 0.123 | 0.059 | 0.214 | 0.001 |
The specific indirect association between digital health literacy and sports participation through cognitive flexibility was 0.076 (95% CI [0.018, 0.145], p = 0.010). Because the interval excluded zero, the cross-sectional pattern was consistent with Hypothesis 2.
The specific indirect association through physical activity self-efficacy was 0.036 (95% CI [0.008, 0.084], p = 0.006). The interval excluded zero, providing cross-sectional statistical support for Hypothesis 3.
The theory-ordered sequential indirect association through cognitive flexibility and physical activity self-efficacy was 0.011 (95% CI [0.002, 0.031], p = 0.012). Although statistically different from zero, its magnitude was small. The covariance pattern was consistent with Hypothesis 4.
5. Discussion
Digital health literacy represents an important psychological resource in digitally mediated health environments. This study examined whether digital health literacy was associated with sports participation and whether the covariance pattern was consistent with prespecified indirect links through cognitive flexibility and physical activity self-efficacy. Digital health literacy showed a direct standardized association of 0.191 with sports participation, and the total indirect association was 0.123. The sequential component was only 0.011. Thus, the results support a statistically coherent cross-sectional association model, not evidence that a psychological process unfolded in the proposed order. A noteworthy contribution of the present study is that it extends the understanding of physical activity behavior from traditional motivational explanations toward a digital cognitive perspective. While previous studies have primarily focused on motivational determinants of sports participation, the current findings suggest that adaptive information-processing abilities may represent an additional psychological mechanism through which digital competencies are associated with health behavior.
Digital health literacy showed a significant positive association with sports participation. This finding aligns with previous studies indicating that individuals with higher levels of digital health literacy are more likely to engage in health-promoting behaviors, including regular physical activity (Jiang et al., 2024; Shudayfat et al., 2023). In the current digital environment, college students increasingly rely on social media, health applications, wearable devices, and online exercise platforms to obtain exercise-related information. Individuals with stronger digital health literacy are generally more capable of identifying reliable health information, filtering misinformation, and transforming fragmented digital resources into actionable health behaviors. Unlike traditional health literacy, digital health literacy emphasizes not only comprehension ability but also interactive information processing and decision-making capacity within technology-based environments (Norman and Skinner, 2006). Therefore, students with higher digital health literacy may be more willing to seek exercise knowledge and maintain participation in sports activities actively.
The specific indirect association through cognitive flexibility (0.076) was larger than the indirect association through physical activity self-efficacy alone (0.036), although both were small in absolute terms. In digitally saturated environments, cognitive resources may represent a potentially important psychological factor associated with sports participation in digital environments. Unlike traditional exercise behavior models that primarily emphasize motivation or behavioral intention, the current findings highlight the importance of adaptive cognitive processing under conditions characterized by information fragmentation, multitasking, and continuous digital stimulation. This may help explain why students with stronger cognitive flexibility were more likely to maintain higher levels of sports participation despite the unstable and distraction-rich nature of contemporary digital environments (Dennis and Vander Wal, 2010). Previous studies have shown that cognitively flexible individuals demonstrate better emotional regulation, behavioral adaptation, and problem-solving ability under stress (Nakhostin-Khayyat et al., 2024; Yurt and Hayli, 2025).
Physical activity self-efficacy was positively associated with sports participation (β = 0.173), consistent with social cognitive theory’s emphasis on efficacy beliefs (Bandura, 2013). Whether college students maintain a regular exercise routine depends not only on knowledge acquisition but also on their belief that they can effectively maintain exercise amidst their busy daily lives. In today’s digital environment, exercise-related apps, wearable devices, and online fitness communities can continuously provide behavioral feedback, social comparisons, and short-term achievement experiences. These experiences can gradually enhance students’ confidence in managing exercise behavior under pressure. Therefore, digital health literacy can indirectly promote participation in physical activity by improving individuals’ ability to organize and maintain physical activity in daily life (Bebeley et al., 2017; Dishman et al., 2005). The current findings further extend this literature by demonstrating that digital health literacy may serve as an upstream psychological resource that contributes to the formation of self-efficacy.
Students who possess stronger digital health literacy may first become more cognitively adaptive when dealing with exercise-related information and lifestyle challenges. This increased flexibility may subsequently strengthen their confidence in maintaining physical activity, ultimately contributing to more stable sports participation. Compared with traditional linear explanations of exercise behavior, the present findings emphasize that sports participation in digital environments may involve multiple interconnected psychological adjustment processes rather than a single motivational factor alone (Mohammadkhani et al., 2022; Woo et al., 2024). The indirect association between digital health literacy, cognitive flexibility, physical activity self-efficacy, and sports participation was statistically significant, but its value was only 0.011, indicating a weak association with potentially limited practical significance. Theoretically, students’ ability to access and use digital health information may be related to their ability to flexibly adjust their thinking, and cognitive flexibility may be related to confidence in physical activity self-efficacy, which in turn is related to sports participation. This interpretation is consistent with the overall association pattern of the research data. However, since all variables were measured at the same time, this study cannot prove that these variables occurred in the above order, nor can it rule out the possibility of inverse relationships or mutual influences. Future studies should employ longitudinal research with longer time intervals to further examine whether changes in digital health literacy precede changes in cognitive flexibility, physical activity self-efficacy, and sports participation.
In conclusion, digital health literacy may play a broader role in college students’ health behaviors than previously thought. Digital health literacy is not merely the technical ability to search for health information online. It may also influence how individuals cognitively adapt to complex digital environments and maintain confidence in long-term behavioral regulation. Against the backdrop of the rapid development of digital fitness culture and technology-driven lifestyles, understanding these psychological pathways may become increasingly important for future strategies to promote physical activity among young people.
6. Conclusion
Among 600 college students from one Chinese university, digital health literacy was positively associated with cognitive flexibility, physical activity self-efficacy, and sports participation. The prespecified indirect-association model fit the data, but most coefficients were small, and the ordered three-path estimate was only 0.011. The study therefore identifies a theory-consistent covariance pattern rather than a verified causal mechanism. Its main contribution is to connect digital-health-information competence with adaptive cognition and exercise-related beliefs while making clear that temporal ordering and practical effects require longitudinal, experimental, and multi-institutional confirmation.
7. Limitations and future directions
Digital health literacy may simultaneously function as both a health resource and a potential cognitive burden. First, the cross-sectional design cannot establish temporal precedence, causal direction, or longitudinal mediation; reverse and reciprocal relations are plausible (Maxwell and Cole, 2007). Future studies should use at least three temporally separated assessments or experimental manipulation when testing the proposed sequence. Second, while higher digital health literacy may facilitate sports participation through improved information processing and behavioral regulation, excessive exposure to digital information may also contribute to information overload, screen fatigue, attentional fragmentation, and digital dependence. These negative aspects of digital environments have received comparatively limited attention in existing physical activity research. Therefore, future studies should adopt a more balanced perspective by examining both the beneficial and potentially adverse psychological consequences of digital health environments. Third, all variables were measured using self-report questionnaires, which may increase the possibility of response bias and social desirability effects. Although Harman’s single-factor test suggested that common method bias was not severe, the use of a single data collection approach may still influence the stability of the findings. Future studies may benefit from combining objective indicators, such as wearable device data, exercise records, or digital behavior tracking, to improve measurement accuracy and ecological validity. Finally, the current sample was limited to college students, which may reduce the generalizability of the findings to other populations. Digital health literacy and exercise behavior may differ substantially across age groups, educational backgrounds, and sociocultural contexts. Future research should examine whether the proposed psychological mechanisms remain stable among adolescents, older adults, athletes, or clinical populations.
Acknowledgments
The authors gratefully acknowledge the support provided by South China University of Technology, Ningbo University, and City University of Macau during the conduct of this research.
Funding Statement
The author(s) declared that financial support was received for this work and/or its publication. This work was supported by the 2026 Guangzhou Philosophy and Social Sciences Planning Project (Grant No. 2026GZQN69); supported by “the Fundamental Research Funds for the Central Universities” (Grant No. QNMS2610); and the Student Research Program of South China University of Technology (Grant No. X202610561714).
Footnotes
Edited by: António Miguel Monteiro, Instituto Politécnico de Bragança, Portugal
Reviewed by: Artur Jorge Baptista Santos, Polytechnic Institute of Bragança (IPB), Portugal
Alberto Rocha, Higher Institute of Educational Sciences of the Douro, Portugal
Data availability statement
The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding author/s.
Ethics statement
The studies involving humans were approved by the Guangdong University of Finance and Economics. The studies were conducted in accordance with the local legislation and institutional requirements. Written informed consent for participation was not required from the participants or the participants’ legal guardians/next of kin in accordance with the national legislation and institutional requirements.
Author contributions
YumX: Formal analysis, Data curation, Writing – original draft. JP: Data curation, Writing – original draft, Formal analysis. YutX: Validation, Supervision, Writing – review & editing. NW: Methodology, Writing – review & editing, Supervision.
Conflict of interest
The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Generative AI statement
The author(s) declared that Generative AI was not used in the creation of this manuscript.
Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.
Publisher’s note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
References
- Acar A. K., Savci S., Kahraman B. O., Tanriverdi A. (2024). Comparison of e-health literacy, digital health and physical activity levels of university students in different fields. J. Basic Clin. Health Sci. 8, 380–389. doi: 10.30621/jbachs.1339191 [DOI] [Google Scholar]
- Ban S., Kim Y., Seomun G. (2024). Digital health literacy: a concept analysis. Digit. Health 10:20552076241287894. doi: 10.1177/20552076241287894, [DOI] [PMC free article] [PubMed] [Google Scholar]
- Bandura A. (2013). “Self-efficacy: the foundation of agency,” in Control of Human Behavior, Mental Processes, and Consciousness, (New York, NY: Psychology Press; ), 16–30. [Google Scholar]
- Bayati T., Dehghan A., Bonyadi F., Bazrafkan L. (2018). Investigating the effect of education on health literacy and its relation to health-promoting behaviors in health center. J. Educ. Health Promot. 7:127. doi: 10.4103/jehp.jehp_65_18, [DOI] [PMC free article] [PubMed] [Google Scholar]
- Bebeley S. J., Liu Y., Yi-Gang W. (2017). Physical exercise self-efficacy for college students' level of motivation in physical activity. Int. J. Sci. Res. 6, 2274–2278. doi: 10.21275/ART20175799 [DOI] [Google Scholar]
- Berkman N. D., Davis T. C., McCormack L. (2010). Health literacy: what is it? J. Health Commun. 15, 9–19. doi: 10.1080/10810730.2010.499985, [DOI] [PubMed] [Google Scholar]
- Brislin R. W. (1970). Back-translation for cross-cultural research. J. Cross-Cult. Psychol. 1, 185–216. doi: 10.1177/135910457000100301 [DOI] [Google Scholar]
- Bull F. C., Al-Ansari S. S., Biddle S., Borodulin K., Buman M. P., Cardon G., et al. (2020). World Health Organization 2020 guidelines on physical activity and sedentary behaviour. Br. J. Sports Med. 54, 1451–1462. doi: 10.1136/bjsports-2020-102955 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Chen H., Dai J., Gao Y. (2019). Measurement invariance and latent mean differences of the Chinese version physical activity self-efficacy scale across gender and education levels. J. Sport Health Sci. 8, 46–54. doi: 10.1016/j.jshs.2017.01.004, [DOI] [PMC free article] [PubMed] [Google Scholar]
- Cho J., Park D., Lee H. E. (2014). Cognitive factors of using health apps: systematic analysis of relationships among health consciousness, health information orientation, eHealth literacy, and health app use efficacy. J. Med. Internet Res. 16:e125. doi: 10.2196/jmir.3283, [DOI] [PMC free article] [PubMed] [Google Scholar]
- Dadaczynski K., Okan O., Messer M., Leung A. Y. M., Rosário R., Darlington E., et al. (2021). Digital health literacy and web-based information-seeking behaviors of university students in Germany during the COVID-19 pandemic: cross-sectional survey study. J. Med. Internet Res. 23:e24097. doi: 10.2196/24097, [DOI] [PMC free article] [PubMed] [Google Scholar]
- Dennis J. P., Vander Wal J. S. (2010). The cognitive flexibility inventory: instrument development and estimates of reliability and validity. Cogn. Ther. Res. 34, 241–253. doi: 10.1007/s10608-009-9276-4 [DOI] [Google Scholar]
- Dishman R. K., Motl R. W., Sallis J. F., Dunn A. L., Birnbaum A. S., Welk G. J., et al. (2005). Self-management strategies mediate self-efficacy and physical activity. Am. J. Prev. Med. 29, 10–18. doi: 10.1016/j.amepre.2005.03.012, [DOI] [PMC free article] [PubMed] [Google Scholar]
- Fornell C., Larcker D. F. (1981). Evaluating structural equation models with unobservable variables and measurement error. J. Mark. Res. 18, 39–50. doi: 10.1177/002224378101800104 [DOI] [Google Scholar]
- Hohl K., Dolcos S. (2024). Measuring cognitive flexibility: a brief review of neuropsychological, self-report, and neuroscientific approaches. Front. Hum. Neurosci. 18:1331960. doi: 10.3389/fnhum.2024.1331960, [DOI] [PMC free article] [PubMed] [Google Scholar]
- Huang M., Wang S., Yao W., Tan Z., Zhao H., Chen W., et al. (2026). The impact of digital teaching on learning burnout in college students: the moderating role of self-efficacy and the mediating effect of learning adaptability. BMC Med Educ 26:194. doi: 10.1186/S12909-025-08330-0, [DOI] [PMC free article] [PubMed] [Google Scholar]
- Jiang S., Ng J. Y. Y., Choi S. M., Ha A. S. (2024). Relationships among eHealth literacy, physical literacy, and physical activity in Chinese university students: cross-sectional study. J. Med. Internet Res. 26:e56386. doi: 10.2196/56386, [DOI] [PMC free article] [PubMed] [Google Scholar]
- Kühn L., Bachert P., Hildebrand C., Kunkel J., Reitermayer J., Wäsche H., et al. (2022). Health literacy among university students: a systematic review of cross-sectional studies. Front. Public Health 9:680999. doi: 10.3389/fpubh.2021.680999, [DOI] [PMC free article] [PubMed] [Google Scholar]
- Li H., Yang Z. (2024). The impact of digital literacy on personal health: evidence from the China family panel study. Int. J. Health Plann. Manag. 39, 1411–1433. doi: 10.1002/hpm.3816, [DOI] [PubMed] [Google Scholar]
- Liang D. Q., Lin S. J. (1994). The relationship between stress level and physical exercise for college students. Chin. Ment. Health J. 8, 5–6. [Google Scholar]
- Liu H. X., Chow B. C., Hassel H., Huang Y. J. W., Liang W., Wang R. B. (2024). Prospective association of eHealth literacy and health literacy with physical activity among Chinese college students: a multiple mediation analysis. Front. Public Health 12:1275691. doi: 10.3389/fpubh.2024.1275691, [DOI] [PMC free article] [PubMed] [Google Scholar]
- Liu L., Fu M., Wu J., Wang H., Zhao J., Chen P., et al. (2024). Digital health literacy among undergraduate nursing students in China: associations with health lifestyles and psychological resilience. BMC Med. Educ. 24:1139. doi: 10.1186/s12909-024-06075-w, [DOI] [PMC free article] [PubMed] [Google Scholar]
- Martin M. M., Rubin R. B. (1995). A new measure of cognitive flexibility. Psychol. Rep. 76, 623–626. doi: 10.2466/pr0.1995.76.2.623 [DOI] [Google Scholar]
- Maxwell S. E., Cole D. A. (2007). Bias in cross-sectional analyses of longitudinal mediation. Psychol. Methods 12, 23–44. doi: 10.1037/1082-989X.12.1.23, [DOI] [PubMed] [Google Scholar]
- Mohammadkhani S., Foroutan A., Akbari M., Shahbahrami M. (2022). Emotional schemas and psychological distress: mediating role of resilience and cognitive flexibility. Iran. J. Psychiatry 17, 284–293. doi: 10.18502/ijps.v17i3.9728, [DOI] [PMC free article] [PubMed] [Google Scholar]
- Motl R. W., Dishman R. K., Trost S. G., Saunders R. (2000). Factorial validity and invariance of questionnaires measuring social-cognitive determinants of physical activity among adolescent girls. Prev. Med. 31, 584–594. doi: 10.1006/pmed.2000.0735, [DOI] [PubMed] [Google Scholar]
- Nakhostin-Khayyat M., Borjali M., Zeinali M., Fardi D., Montazeri A. (2024). The relationship between self-regulation, cognitive flexibility, and resilience among students: a structural equation modeling study. BMC Psychol. 12:337. doi: 10.1186/s40359-024-01843-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Norman C. D., Skinner H. A. (2006). eHEALS: the eHealth literacy scale. J. Med. Internet Res. 8:e27. doi: 10.2196/jmir.8.4.e27, [DOI] [PMC free article] [PubMed] [Google Scholar]
- Paige S. R., Stellefson M., Krieger J. L., Anderson-Lewis C., Cheong J. W., Stopka C. (2018). Proposing a transactional model of eHealth literacy: concept analysis. J. Med. Internet Res. 20:e10175. doi: 10.2196/10175, [DOI] [PMC free article] [PubMed] [Google Scholar]
- Podsakoff P. M., MacKenzie S. B., Lee J.-Y., Podsakoff N. P. (2003). Common method biases in behavioral research: a critical review of the literature and recommended remedies. J. Appl. Psychol. 88, 879–903. doi: 10.1037/0021-9010.88.5.879, [DOI] [PubMed] [Google Scholar]
- Ridwan M., Mar'atus A., Acd S., Budi B., Abd P. (2022). Health literacy and health-related behavior in sport among university students in East Java, Indonesia: a cross-sectional study. J. Sport Area 7, 104–116. doi: 10.25299/sportarea.2022.vol7(1).8098 [DOI] [Google Scholar]
- Shokrkon A., Nicoladis E. (2022). The directionality of the relationship between executive functions and language skills: a literature review. Front. Psychol. 13:848696. doi: 10.3389/fpsyg.2022.848696, [DOI] [PMC free article] [PubMed] [Google Scholar]
- Shudayfat T., Hani S. B., Shdaifat E., Al-Mugheed K., Alsenany S. A., Farghaly Abdelaliem S. M. (2023). Electronic health literacy and its association with lifestyle behavior among undergraduate students: a cross-sectional survey. Digit. Health 9:20552076231185429. doi: 10.1177/20552076231185429, [DOI] [PMC free article] [PubMed] [Google Scholar]
- Singh B., Ahmed M., Staiano A., Vasiloglou M. (2025). Lifestyle eHealth and mHealth interventions for children and adolescents: systematic umbrella review and meta-meta-analysis. J. Med. Internet Res. 27:e69065. doi: 10.2196/69065, [DOI] [PMC free article] [PubMed] [Google Scholar]
- Tong K., Chan Y. N., Cheng X., Cheon B., Ellefson M., Fauziana R., et al. (2023). Study protocol: how does cognitive flexibility relate to other executive functions and learning in healthy young adults? PLoS One 18:e0286208. doi: 10.1371/journal.pone.0286208, [DOI] [PMC free article] [PubMed] [Google Scholar]
- van der Vaart R., Drossaert C. (2017). Development of the digital health literacy instrument: measuring a broad spectrum of health 1.0 and health 2.0 skills. J. Med. Internet Res. 19:e27. doi: 10.2196/jmir.6709, [DOI] [PMC free article] [PubMed] [Google Scholar]
- Wang Y., Yang Y., Xiao W. T., Su Q. (2016). Validity and reliability of the Chinese version of the cognitive flexibility inventory in college students. Chin. Ment. Health J. 30, 58–62. doi: 10.3969/j.issn.1000-6729.2016.01.002 [DOI] [Google Scholar]
- Woo S. H., Seo J. P., Kim H. R., So W. Y., Sim Y. K. (2024). Health-promoting behaviors, physical self-efficacy, exercise adherence, and sports commitment among older adults who participate in sports activities. Healthcare 12:2135. doi: 10.3390/healthcare12212135, [DOI] [PMC free article] [PubMed] [Google Scholar]
- Xie L., Hu H., Lin J., Mo P. K. H. (2024). Psychometric validation of the Chinese digital health literacy instrument among Chinese older adults who have internet use experience. Int. J. Older People Nursing 19:e12568. doi: 10.1111/opn.12568, [DOI] [PubMed] [Google Scholar]
- Yu M., Han X., Wang X., Guan R. (2023). Effects of physical exercise on executive functions among college students in China: exploring the influence of exercise intensity and duration. Behav. Sci. 13:987. doi: 10.3390/bs13120987, [DOI] [PMC free article] [PubMed] [Google Scholar]
- Yurt E., Hayli Ç. M. (2025). Mediating role of self-efficacy and cognitive flexibility in the relationship between critical thinking and positive mental health in Turkish nursing students: a cross-sectional study. BMJ Open 15:e097631. doi: 10.1136/bmjopen-2024-097631, [DOI] [PMC free article] [PubMed] [Google Scholar]
- Zhang Z., Rao Z., Liu M., Han J., Zheng L. (2026). The relationship between physical exercise and learning engagement among chinese college students: the chain mediating roles of cognitive flexibility and psychological resilience. Front. Psychol. 17:1731426. doi: 10.3389/fpsyg.2026.1731426, [DOI] [PMC free article] [PubMed] [Google Scholar]
- Zhu W., Liu J., Lou H., Mu F., Li B. (2024). The impact of electronic health literacy on emotional management ability among college students: the mediating roles of peer relationships and exercise self-efficacy. BMC Psychol. 12:747. doi: 10.1186/s40359-024-02276-6, [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding author/s.