All statistical analyses were conducted using IBM SPSS Statistics (Version 27.0). First, common method bias was examined to assess potential biases in the study. Descriptive statistics were then used to evaluate the distribution of variables. Next, normality assumptions were assessed via skewness and kurtosis values; skewness < |3| and kurtosis < |8| indicated acceptable univariate normality for structural equation modeling (67). Subsequently, correlation analyses were performed to examine inter-variable relationships and their strengths. Third, SPSS macro Process Model 6 was utilized to test the chain mediation model involving perceived disease threat and self-efficacy, with health literacy as the independent variable, perceived disease threat and self-efficacy as chain mediators, and treatment adherence as the dependent variable. This model incorporated direct paths from health literacy to treatment adherence, as well as indirect paths through the chain mediators, while controlling for covariates to isolate psychosocial mechanisms. Model fit was evaluated using multiple indices: chi-square (χ²) statistic, comparative fit index (CFI > 0.90), Tucker-Lewis index (TLI > 0.90), root mean square error of approximation (RMSEA < 0.08), and others. Bias-corrected bootstrapping with 5,000 resamples was employed to estimate indirect effects, generating 95% confidence intervals (CIs) to provide robust inferences for mediation without relying on normality assumptions.

Given that all measurement scales in this study employed a 5-point Likert format, with data sourced from the same location and potentially influenced by social desirability, common method bias may be present. Therefore, we safeguarded respondents’ anonymity and data confidentiality to mitigate such bias.

We conducted Harman’s single-factor test by subjecting all variable items to exploratory factor analysis. The results indicated that the first factor accounted for 17.781% of the total variance, below the critical threshold of 40%. Thus, common method bias was not a concern in this study.

Descriptive statistics and correlation analyses were performed for health literacy, perceived disease threat, self-efficacy, and treatment adherence, as shown in Tables 2, 3. Among COPD patients, the mean score for health literacy was 3.384 (SD = 0.601), for perceived disease threat was 2.530 (SD = 0.637), for self-efficacy was 3.435 (SD = 0.607), and for treatment adherence was 3.442 (SD = 0.641). The mean scores for these core variables all exceeded the midpoint (M = 2.5) of the 5-point Likert scale. Skewness values ranged from −0.816 to 0.610, and kurtosis values ranged from 0.845 to 2.003. According to Kline (67) criteria for normality, the core variables approximated a normal distribution.

Correlation analyses revealed a significant negative correlation between health literacy and perceived disease threat (r = −0.512, p < 0.01); a significant positive correlation between health literacy and self-efficacy (r = 0.576, p < 0.01); and a significant positive correlation between health literacy and treatment adherence (r = 0.510, p < 0.01). Self-efficacy showed a significant negative correlation with perceived disease threat (r = −0.578, p < 0.01) and a significant positive correlation with treatment adherence (r = 0.481, p < 0.01). Perceived disease threat exhibited a significant negative correlation with treatment adherence (r = −0.556, p < 0.01).

Health literacy was treated as the independent variable, perceived disease threat and self-efficacy as mediators, and treatment adherence as the dependent variable. We constructed a structural equation model using AMOS 29.0 to test the research hypotheses. Confirmatory factor analysis demonstrated excellent model fit: χ²/df = 1.183; CFI = 0.945; GFI = 0.922; AGFI = 0.912; TLI = 0.941; RMSEA = 0.020.

The results indicated that health literacy exerted a significant negative direct effect on perceived disease threat [β = −0.283, p < 0.001, 95% CI = (−0.377, −0.189)]; health literacy had a positive direct effect on self-efficacy [β = 0.582, p < 0.001, 95% CI = (0.506, 0.658)]. Self-efficacy negatively predicted perceived disease threat [β = −0.445, p < 0.001, 95% CI = (−0.538, −0.352)]; perceived disease threat negatively predicted treatment adherence [β = −0.352, p < 0.001, 95% CI = (−0.445, −0.259)]. Self-efficacy positively predicted treatment adherence [β = 0.139, p < 0.01, 95% CI = (0.037, 0.241)]. Health literacy had a significant positive effect on treatment adherence [β = 0.273, p < 0.001, 95% CI = (0.175, 0.371)]. Detailed results are presented in Table 4.

Mediation effect analysis, as shown in Table 5 and Figure 1, revealed that the total effect of health literacy on treatment adherence was 0.544, of which 49.9% was explained by the mediating roles of self-efficacy and perceived disease threat. Further examination showed a significant chain mediation effect via health literacy → self-efficacy → perceived disease threat → treatment adherence [β = 0.091, SE = 0.022, 95% CI = (0.051, 0.137)], accounting for 16.7% of the total effect. For individual mediation paths, health literacy → self-efficacy → treatment adherence exhibited a significant mediating effect [β = 0.081, SE = 0.038, 95% CI = (0.008, 0.157)], representing 14.9% of the total effect; health literacy → perceived disease threat → treatment adherence showed a significant mediating effect [β = 0.100, SE = 0.025, 95% CI = (0.056, 0.150)], comprising 18.3% of the total effect.

This study empirically demonstrates the interrelationships among health literacy, perceived disease threat, self-efficacy, and treatment adherence in patients with chronic obstructive pulmonary disease (COPD). The findings confirm that health literacy indirectly promotes treatment adherence through a chain mediation mechanism involving self-efficacy and perceived disease threat. This revelation elucidates the underlying psychological mechanisms by which health literacy influences patient adherence behaviors, bridging the theoretical gap in life behavior intervention research from knowledge acquisition to actionable cognition. The results align with the core assumptions of the health belief model and protection motivation theory, while extending their applicability, particularly in the domain of respiratory chronic disease management, thereby providing a novel theoretical framework for explaining patient treatment behaviors.

This study supports a significant positive association between health literacy and treatment adherence, further unveiling the mediating roles of perceived disease threat and self-efficacy. Grounded in the health belief model, individual behavioral motivation stems from perceptions of disease risk and beliefs in one’s behavioral efficacy (68), with health literacy serving as a cognitive enhancer. Further analysis reveals that high levels of health literacy not only assist patients in accurately comprehending disease severity but also prevent excessive fear or cognitive distortions, fostering moderate perceived disease threat and thereby bolstering self-efficacy to elicit positive treatment behaviors. Conversely, patients with low health literacy may amplify disease threats due to misinformation, negative beliefs, or barriers in patient-provider communication, undermining self-management beliefs and leading to reduced adherence. This partially explains the paradox observed in prior research of high knowledge yet low adherence.

The study findings further corroborate the mediating role of perceived disease threat in the link between health literacy and adherence. Diverging from traditional views treating it as a singular variable, this research identifies perceived disease threat as a pivotal psychological node in converting cognitive signals into actions, achieving behavioral internalization through the activation of self-efficacy. Specifically, elevated health literacy enables patients to accurately evaluate disease consequences and controllability, transforming external compliance with treatment into internal endorsement. Self-efficacy acts as a bridge in this process, converting rational cognition into the psychological drive for sustained action (69). This model validates the dynamic interplay among environmental cognition, individual beliefs, and behavioral outcomes in social cognitive theory, while enriching the motivational composition framework of protection motivation theory—namely, that individuals can stabilize protective behaviors through confidence-building only after fully understanding threats and resources. The model’s robust stability and fit indices further enhance the credibility and external validity of these results.

This study reveals that health literacy not only directly influences treatment adherence in COPD patients but also indirectly fosters sustained treatment behaviors through a chain mediation pathway involving self-efficacy and perceived disease threat. These findings provide robust empirical support and theoretical foundations for redesigning psychological interventions and health education models in chronic disease management. The core contribution lies in systematically validating, for the first time in human chronic disease psychobehavioral mechanisms, the dynamic transmission pathway of “cognitive resources—subjective beliefs—risk perception—behavioral decision-making,” thereby deepening the psychological connotations of health literacy and expanding the explanatory boundaries of protection motivation theory and social cognitive theory in clinical behavior domains.

First, from the perspective of cognitive-behavioral interventions, the results suggest that merely enhancing patients’ medical knowledge reserves is insufficient for behavioral transformation; rather, the processing of health information and the psychological belief system are the decisive factors for sustained adherence. Health literacy aids individuals in more accurately understanding treatment plans and risk information, thereby alleviating cognitive uncertainty, while self-efficacy translates this cognitive advantage into an “I can do it” behavioral motivation. Moreover, moderate perceived disease threat, under the synergistic regulation of individual cognitive resources and belief structures, can elicit rational defensive actions rather than anxiety-driven avoidance. Therefore, clinical interventions should concurrently address patients’ cognitive enlightenment, self-efficacy reinforcement, and risk perception modulation, promoting adherence internalization through psychological empowerment.

Second, at the level of medical practice and patient education, the model’s evidence-based support informs precision chronic disease management. Healthcare providers should incorporate multi-tiered health literacy intervention strategies during treatment: at the cognitive level, establish health literacy profiles based on stratified assessments, utilizing multimodal, accessible educational materials (e.g., interactive videos, visual diagrams, concise health guides) to improve information comprehension; at the belief level, enhance patient self-efficacy through individualized goal management, behavioral feedback, and peer modeling support, encouraging proactive participation in disease management within controllable scopes; at the emotional level, facilitate appropriate disease threat cognition via risk communication training, avoiding excessive fear or complacency to maintain vigilant yet rational health motivations.

Health literacy challenges are not confined to COPD populations or to a single country; rather, wide disparities in health literacy exist globally, with many regions reporting limited public awareness, inadequate comprehension of disease information, and substantial gaps in self-management capabilities (70). International studies have shown that low health literacy is common in both high-income and low- to middle-income countries, often contributing to delayed diagnosis, poor treatment adherence, and suboptimal chronic disease outcomes (71). These disparities highlight a pressing global need to strengthen health literacy through coordinated public health strategies, culturally adapted education materials, and improved health communication systems. The present study contributes to this global agenda by demonstrating how psychological mechanisms link health literacy to behavior, offering insights that may be applicable to diverse populations. By clarifying how self-efficacy and perceived disease threat influence adherence, our findings underscore the importance of developing globally relevant interventions that go beyond information dissemination and instead empower individuals to interpret, internalize, and act on health information in meaningful ways.

Traditional IEC materials often focus predominantly on transmitting factual information; however, our results suggest that such approaches may be insufficient unless they simultaneously reinforce patients’ confidence in managing treatment tasks and help them form accurate, moderate perceptions of disease threat. By integrating explicit self-efficacy enhancement strategies and calibrated risk communication into IEC programs, health systems can shift from knowledge provision alone toward psychologically informed interventions that more reliably induce adherence. This refined model offers a pathway to strengthen community-level disease management efforts and thereby supports the overarching goals of IEC: reducing disease burden through sustained behavioral engagement.

While IEC interventions were originally designed to enhance public awareness and promote treatment adherence through improved health literacy, their real-world effectiveness has remained limited. This is largely due to inadequate resource allocation, insufficiently trained communication personnel, and the lack of innovative, population-tailored strategies for building social awareness. Consequently, IEC functions are often underdeveloped across the continuum of care—from preventive and promotive services to curative, rehabilitative, and palliative care—and across primary, secondary, and tertiary levels of the health system. Our findings underscore the urgent need to redesign IEC components in a more evidence-based and psychologically informed manner. In addition to disseminating information, IEC initiatives must incorporate structured approaches to enhance patients’ self-efficacy and calibrate perceived disease threat, as these mechanisms play a critical role in transforming knowledge into behavior. Furthermore, the importance of health literacy for self-management is increasingly recognized in a wide range of non-communicable diseases, such as hypertension, diabetes, and cancer screening. Strengthening IEC capacity therefore requires expanding investment in trained health educators, developing engaging and culturally sensitive communication materials, and integrating behavioral science principles into public health messaging. Accordingly, the results of this study provide a theoretical and empirical foundation for policymakers to reinforce and modernize IEC strategies in order to improve long-term chronic disease management and reduce population-level disease burden.

Furthermore, the theoretical model holds significant implications for public health policy and chronic disease management system development. Health literacy can be regarded as a foundational health competency in national chronic disease prevention frameworks, warranting continuous enhancement through community health education, digital health platforms, and “clinician-patient-family” tripartite collaboration mechanisms. Integrating health literacy assessments into COPD stratified care systems can optimize patient risk stratification and individualized management, shifting paradigms from “medically oriented” to “patient cognition-oriented.” Additionally, reinforcing self-efficacy and optimizing threat perception can serve as intermediate psychological indicators in health promotion programs, offering novel operational pathways for integrating precision medicine with psychological health interventions.

When explaining the psychological mechanism in this study, it’s important to also consider another key external factor that can affect adherence behavior: the structure of medical cost payment. Our sample comes from top-tier hospitals in Beijing, where patients generally benefit from a well-developed public funding and health insurance system. As a result, economic burdens have less direct impact on treatment adherence. However, in many countries or regions where out-of-pocket payments dominate or drug co-pays are high, the financial cost itself becomes a critical determinant of adherence. Previous research has shown that even patients with high health literacy and clear awareness of disease risks may significantly reduce their adherence due to heavy medication expenses, making it difficult to sustain standardized treatment long-term (72, 73).

Therefore, the psychological mechanism model developed in this study is more applicable to contexts with solid healthcare coverage and relatively low patient cost-sharing. In countries or populations facing greater economic burdens, the behavioral benefits gained from improved health literacy may be partly offset by high expenses. Future research should incorporate systemic factors such as “medical cost burden,” “payment methods,” and “reimbursement rates” into expanded models and examine their interactions or moderating effects with psychological variables. This would help build a more comprehensive framework of treatment adherence determinants and enhance the cross-cultural generalizability of the findings.

While the present findings are derived from COPD patients, the mechanisms identified—linking health literacy to treatment adherence through self-efficacy and calibrated disease threat—offer broader policy implications that extend well beyond traditional IEC efforts. The results provide policymakers with a theoretically grounded framework for strengthening health literacy at the population level through more targeted and psychologically informed programmatic interventions.

A major challenge in public health is achieving high levels of health literacy across diverse communities using scalable and resource-sensitive strategies. Although resource constraints are well known, certain intervention models can substantially enhance the effectiveness of existing IEC systems without requiring major structural expansion. Three illustrative examples demonstrate how policy makers can translate the mechanisms identified in this study into actionable health literacy strategies.

Although this study provides novel empirical evidence for understanding the psychological mechanisms of health literacy in COPD patients, several limitations and avenues for improvement remain. Given the cross-sectional design of this study, causal relationships among health literacy, self-efficacy, perceived disease threat, and treatment adherence cannot be inferred. The associations identified reflect correlational patterns at a single time point, and the design does not permit any directional or temporal conclusions. For instance, patients with high adherence may develop elevated health literacy and self-efficacy through long-term treatment experiences. Future research could employ longitudinal designs or experimental interventions to verify the dynamic effects of health literacy interventions on self-efficacy, perceived disease threat, and adherence over time, thereby strengthening causal explanations.

Second, the sample was drawn from patients at tertiary Grade A hospitals in Beijing, potentially exhibiting higher healthcare accessibility and cognitive levels, limiting generalizability to primary care or rural COPD populations. Subsequent studies should expand sample sources to include patients from diverse regions and cultural backgrounds, utilizing multicenter stratified sampling to enhance sample heterogeneity and the external validity of results.

All measurement instruments were self-report scales, susceptible to social desirability bias and common method variance. Although this study mitigated risks through anonymous responses and statistical tests, future research could incorporate objective behavioral data (e.g., medication usage records, follow-up attendance data, or wearable monitoring metrics) to validate adherence measurements, establishing composite models based on objective behavioral indicators.

Another limitation of this study is the absence of a control group. The inclusion of age- and gender-matched normal controls and diseased controls would have enabled direct comparison of health literacy levels across different population groups in Beijing. Such comparisons could help determine whether the observed patterns are unique to COPD patients or reflect broader population trends. Without these control groups, the generalizability of the findings—particularly regarding the relative impact of health literacy—remains restricted. Future research should incorporate matched control samples to strengthen comparative validity and to more precisely evaluate the specificity of health literacy and psychological mechanisms in COPD.

Finally, this study did not fully account for potential moderating variables. For example, social support, cultural orientations, and patient-provider trust may influence the formation pathways of self-efficacy and threat perception. Future work could build multilayer moderated mediation frameworks on the chain mediation model to explore the impact of socio-contextual variables on model stability. Additionally, potential interactive effects of demographic variables such as gender, age, education level, and disease duration should be considered to uncover psychological mechanism differences across subgroups.