Tobacco use continues to be the leading preventable cause of death worldwide, with its health burden disproportionately affecting populations along social and economic gradients (1). According to the World Health Organization Global Tobacco Epidemic Report (2021), the prevalence of smoking among males in low- and middle-income countries (36.7%) significantly exceeds that in high-income countries (22.3%) (2). These disparities reflect the systemic inequalities embedded in the social determinants of health (SDoH)—a framework that emphasizes the influence of education, income, geographic location, and access to information on individual health behaviors (3).

As the world’s largest tobacco consumer, China presents complex challenges for tobacco control, with notable variations in smoking prevalence across regions, age groups, and socioeconomic strata (4). In this context, within the broader framework of SDoH, Health literacy (HL) has gained recognition as a modifiable dimension of increasing relevance to public health (5). Defined as the ability to access, understand, evaluate, and apply health information, HL is increasingly recognized as a key determinant of health behavior, including tobacco use (6).

Recent studies have demonstrated that low HL may impair an individual’s understanding of the health risks associated with smoking and reduce their capacity to engage with prevention and cessation strategies (5). Nutbeam’s conceptual model of HL outlines three dimensions—functional, interactive, and critical literacy—that may differentially influence health behaviors depending on context and population (7).

Quanzhou was selected as the research site due to its significant provincial representativeness. Its geographic environment exhibits a strong correspondence with the key characteristics of Fujian Province, encompassing typical landforms such as coastal plains, inland hills, and mountainous regions. This diverse terrain provides a regional reference for examining the impact of environmental factors on health behaviors (8, 9). As the city with the largest resident population in Fujian, Quanzhou’s large population base helps reduce sampling error and enhances the statistical robustness of research findings (10, 11). From an economic perspective, Quanzhou has maintained the highest GDP in the province for more than 20 years. Its diversified industrial structure, led by the private sector, includes key sectors such as manufacturing, cross-border e-commerce, and cultural tourism, which reflect the economic development characteristics of Fujian Province (12). The World Health Organization (WHO) emphasizes the importance of considering internal regional differences in SDoH research (13). Quanzhou, as the cultural core of the Minnan region and a significant diasporic hub, is characterized by a high mobility of its resident population, providing a valuable empirical context for exploring the interactive mechanisms between HL and SDoH (14).

This study addresses this gap by examining the association between smoking status and HL among residents aged 15 to <70 years in Quanzhou, Fujian Province. Using data from a large-scale, population-based survey and applying both descriptive and multivariable statistical analyses, we aim to: (1) describe the current smoking prevalence among residents; (2) assess the independent associations between smoking behavior, HL levels, and key social determinants such as age, education, and residence.

This study aims to explore the associations between HL and the SDoH in relation to smoking behavior in a coastal city of southeastern China. It further seeks to provide evidence-based insights for precision tobacco control strategies—particularly in the context of the “Healthy China 2030” national policy framework.

This cross-sectional study examined the prevalence of smoking among residents aged 15–70 years in Quanzhou, a coastal city in southeastern China, and explored its associations with various SDoH, including HL as a key modifiable component. Owing to the negligible prevalence of smoking among female participants, the analyses in this study are predominantly based on male data. Specifically, the overall smoking prevalence was 25.68%, with a remarkably low rate of only 0.66% among females, indicating a substantial gender-based disparity in smoking behaviors. This finding is consistent with the results reported by Xia et al. (17), and may be attributed to the social norms embedded in traditional Chinese culture that discourage smoking among women (18–20). These observations suggest that gender should be considered in tobacco control policies, and that future qualitative studies are needed to further explore the sociocultural mechanisms underlying the extremely low smoking prevalence among women in China (21).

In relation to SDoH, univariate analysis showed that smoking prevalence was significantly higher among rural residents than urban residents, which may reflect disparities in health education, access to cessation resources, or cultural norms that normalize smoking in rural contexts (22). Smoking rates increased with age, peaking in the 55–<65 age group, potentially due to cohort effects, prolonged exposure to smoking environments, and lower cessation success among older adults (23). In addition, multivariable logistic regression analysis revealed that educational attainment was inversely associated with smoking; individuals with only junior high school education had the highest prevalence, while those with a bachelor’s degree or higher had significantly lower odds of current smoking, with a risk ratio of only 0.187 compared to illiterate or semi-literate individuals. This trend is in agreement with prior studies by Jin et al. (24) and Deng et al. (25), both of which identified education as a key predictor of smoking behavior. These findings indicate the importance of prioritizing rural, lower-educated, and older populations in targeted tobacco control interventions.

Multivariable logistic regression analysis further confirmed the significant effects of age on smoking behavior among men. Male participants aged 25–<35 had 5.763 times the odds of smoking compared to those aged 15–<25. This age group may be particularly vulnerable due to being in a critical stage of career development, facing elevated work stress, more frequent social interactions, and greater financial independence, all of which may facilitate the use of nicotine as a coping mechanism (26). Furthermore, this age group is more susceptible to peer influence, and in certain occupations, workplace culture may reinforce smoking behavior (27). These findings suggest the necessity of designing context-specific tobacco control strategies that address occupational and psychosocial risk environments.

Regarding HL, univariate analysis revealed that male participants with adequate HL—specifically in the domains of Basic Knowledge & Concepts, Lifestyle & Behaviors, and Practical Skills—had lower smoking prevalence compared to those with inadequate HL. This is consistent with findings from Li et al. (28) and Guo et al. (29), who reported that higher HL levels are associated with healthier decision-making behaviors. However, after adjusting for sociodemographic variables such as age and education, multivariable logistic regression showed that overall HL, as well as the Basic Knowledge & Concepts and Lifestyle & Behaviors dimensions, were not significantly associated with smoking behavior. This suggests that the influence of these aspects of HL diminishes when other structural factors, such as age and education, are taken into account (30). These findings highlight the need to consider HL in conjunction with other SDoH when developing public health interventions.

In addition, These findings align with the knowledge-action gap, a well-established concept in public health, wherein possessing health knowledge does not necessarily translate into behavior change, particularly in addictive behaviors like smoking (6, 31). In the context of smoking, while individuals may be aware of the risks, psychological dependence, addiction, and social norms are significant barriers to adopting healthier behaviors (6, 32). Furthermore, the findings suggest that smoking behavior is strongly influenced by cultural factors and social influences, which often override individual health knowledge (33). The lack of significance for Basic Knowledge & Concepts and Lifestyle & Behaviors may be explained by the cultural normalcy of smoking in certain environments, particularly in rural settings, where smoking may be ingrained as part of the social fabric (34). In such contexts, individuals may continue smoking despite knowing its health risks, as the behavior may be normalized in their social and work environments, thus reducing the impact of basic health knowledge (27). In addition, smoking is often a coping mechanism for stress and emotional issues, further complicating the role of knowledge in smoking cessation (35, 36).

Practical Skills, which encompass the ability to apply health knowledge effectively in real-life situations, were significantly associated with smoking behavior. This suggests that self-efficacy, the belief in one’s ability to make decisions and take actions that influence health outcomes, is a key determinant in smoking cessation (37). Individuals with better practical skills are more likely to resist smoking triggers and adopt effective cessation strategies, underscoring the importance of not just providing health information but also equipping individuals with the skills to act on that information (38). This aligns with findings from other studies, which suggest that skills-based interventions, such as coping strategies for managing cravings and stress, can significantly improve smoking cessation rates (39, 40).

Additionally, the findings suggest that while HL in its functional and interactive forms may have a limited impact, Critical HL, as conceptualized by Nutbeam in 2000, may be more relevant for influencing smoking behavior (7). Critical HL, which involves the ability to critically analyze health information and understand the social and environmental factors influencing health decisions, could provide individuals with the tools to resist social pressures and make informed, autonomous decisions about smoking cessation (41, 42). Critical HL may enable individuals to question societal norms, including the normalization of smoking, and to adopt behaviors that align with their long-term health goals (28, 41). As such, interventions aimed at enhancing Critical HL could be more effective than those focused solely on increasing functional health knowledge, as they equip individuals to engage with health information in a more critical, contextual manner, ultimately leading to more sustained behavior change (41–43). Accordingly, future tobacco control efforts should not rely solely on knowledge dissemination, but should emphasize the development of critical HL competencies, including risk interpretation, structural awareness, and participatory decision-making, in order to enhance behavior change outcomes.

This study is subject to several limitations that constrain the interpretation and generalizability of the findings. First, the exclusive focus on Quanzhou limits the ability to extrapolate the results to regions with divergent developmental contexts, such as inland or western China. Additionally, the cross-sectional design precludes causal inference regarding the relationships between HL, SDoH, and smoking behavior. Critically, the underrepresentation of key demographic strata impedes comprehensive analysis: the exclusion of adults aged ≥70 years limits the life-course perspective, and the extreme scarcity of female smokers (n = 11, 0.66% of the sample) fundamentally precludes gender-stratified exploration, limiting the study’s ability to further analyze female smoking behavior and its associated factors, while also indicating that the conclusions are primarily applicable to male populations. Furthermore, the exclusive reliance on self-reported data introduces dual measurement biases, including probable underreporting of smoking behavior due to social desirability pressures and potential overestimation of HL levels due to recall inaccuracies. Although core sociodemographic confounders were adjusted for in the analyses, several key structural SDoH were not included. Due to data sensitivity and the design limitations of the standardized questionnaire, data on critical upstream SDoH such as economic status (e.g., household income), occupational exposures (e.g., job strain), social capital (e.g., support networks), and exposure to tobacco control policies were not collected, as these domains were not explicitly incorporated into the survey instrument. As a result, the structural dimension of SDoH may not have been accurately captured, potentially introducing bias in the estimated effects of education, age, and residence, and limiting the ability to reflect broader macro-level influences on smoking behavior. Future research should address these constraints by employing longitudinal designs to establish temporal precedence, purposive sampling of vulnerable subgroups for a more comprehensive analysis, and integrating biochemical verification alongside self-reports to mitigate biases. Additionally, multilevel modeling incorporating community-level SDoH indicators is needed to account for the broader contextual factors influencing smoking behavior. These steps will help refine our understanding of the complex relationships between HL, SDoH, and smoking, and improve the generalizability and accuracy of future findings.

This study highlights a significant prevalence of smoking among residents in Quanzhou, with HL identified as a critical and modifiable factor within the broader framework of SDoH. Smoking prevalence is notably associated with factors including geographic location, age, and education, especially affecting rural populations, young men, and individuals with lower educational attainment. To effectively address smoking in these high-risk groups, interventions should prioritize HL enhancement through tailored educational strategies. For rural communities, accessible smoking cessation support via township health centers and mobile health technologies can address barriers like limited healthcare access. Among young men, early educational initiatives in schools and workplaces should emphasize the risks of smoking and benefits of cessation, employing targeted communication reflecting their social contexts. For those with lower educational levels, simplified visual aids and interactive health information are essential to bridge HL gaps effectively. By integrating HL improvements with culturally sensitive and context-specific approaches, these recommendations align with the goals of Healthy China 2030 and global tobacco control initiatives outlined by WHO.