Association between health literacy and chronic diseases in Anhui Province: findings from a province-wide cross-sectional survey of 16,067 individuals Authors: Yuansheng Fu, Yuanrui Xia, Zhimin Wang, Jianrong Xie, De Xu Published: 2025 ABSTRACT Chronic diseases represent a major global health burden. Health literacy is a modifiable factor in their prevention and control, yet evidence on its association with multimorbidity remains limited. This study investigated the association between health literacy and chronic diseases, focusing on multimorbidity, among residents of Anhui Province, China. A province-wide cross-sectional survey (July-October 2023) included 16,067 residents aged 15–69 years, selected via multistage stratified probability proportional to size (PPS) sampling. Health literacy was assessed using the Chinese Health Literacy Scale (CHLS). Chronic disease status (none, single, multimorbidity) was self-reported. Multivariate logistic regression analysis was employed to examine the association between health literacy and chronic diseases, adjusting for confounders. The overall adequate health literacy rate was 31.90%. Health literacy levels decreased significantly with increasing chronic disease burden: 35.07% (none), 21.13% (single), 11.61% (multimorbidity; Ptrend < 0.001). After adjustment, individuals with multimorbidity had 24% lower odds of adequate health literacy versus those without chronic diseases (aOR = 0.76; 95% CI 0.60–0.96; P = 0.021). Domain-specific analyses revealed significantly reduced odds of adequate basic knowledge and attitudes (aOR = 0.81; 95% CI 0.67–0.97; P = 0.026) and health-related skills literacy (aOR = 0.77; 95% CI 0.63–0.95; P = 0.014) among individuals with multimorbidity. Multimorbidity was also associated with lower health information literacy (aOR = 0.82; 95% CI 0.69–0.97; P = 0.019). This study observed a significant inverse association between health literacy and chronic diseases, particularly multimorbidity. Targeted health literacy improvement interventions focusing on knowledge, skills, and health information access may represent a crucial, cost-effective strategy for chronic disease prevention and control.Supplementary InformationThe online version contains supplementary material available at 10.1038/s41598-025-34194-x. Introduction Chronic diseases pose a paramount global health challenge, accounting for at least 43 million deaths in 2021—equivalent to 75% of all non-pandemic-related fatalities worldwide1. This burden disproportionately impacts individuals, families, and healthcare systems, straining healthcare resources and compromising global quality of life. In China, the burden of chronic diseases has grown alongside rapid aging and socioeconomic transitions; the proportion of Chinese residents dying from chronic diseases exceeds 80% of all deaths2. Notably, a nationwide survey of 0.5 million Chinese adults revealed that 15.8% of participants had multimorbidity—defined as the co-occurrence of two or more chronic conditions in the same individual3,4. Studies have demonstrated that multimorbidity is associated with functional decline, polypharmacy, depression, increased hospitalization rates, and higher socioeconomic burdens5–8. These conditions not only diminish quality of life but also exacerbate healthcare expenditures, highlighting the urgent need for targeted interventions to address this escalating burden. A growing body of evidence indicates that health literacy has emerged as a key factor in the prevention and control of chronic noncommunicable diseases9. Health literacy is defined as an individual’s ability to access, understand, and utilize basic health information and services to make appropriate decisions that maintain and promote their own health10. As a crucial determinant of health, health literacy influences not only the quality of life and life expectancy of residents but also represents one of the most fundamental, economical, and effective measures to improve population health11,12. However, studies have indicated that nearly 39% of the global population exhibits limited health literacy, with 47.6% of Europeans demonstrating insufficient health literacy13,14. In China, the national health literacy level remains suboptimal (29.70% in 2023), though significant improvements from 6.48% in 2008 highlight incremental progress15,16. Studies have shown that in both cross-sectional and longitudinal investigations, low health literacy is consistently linked to suboptimal comprehension of chronic disease self-management, insufficient treatment adherence, diminished health status, elevated risks of hospitalization, and increased mortality9,17. For instance, individuals with limited health literacy exhibit worse blood pressure management than those with adequate health literacy, while diabetic patients with higher health literacy show superior glycemic control and lower complication risks18,19. Mechanistically, health literacy operates through the knowledge-attitude-practice (KAP) framework: health knowledge shapes beliefs, which subsequently drive behavioral modifications20. Thus, improving population health literacy, as a potentially effective approach to addressing the health threats posed by chronic diseases, represents a well-founded and cost-effective strategy. Despite these insights, research exploring the association between health literacy and chronic diseases—especially the association between specific dimensions of health literacy and different chronic disease statuses—remains limited. Therefore, the aim of this province-wide cross-sectional survey of 16,067 individuals in Anhui Province is to determine whether health literacy is independently associated with chronic diseases, thereby providing local evidence to optimize health literacy interventions and chronic disease management strategies in Anhui Province, and ultimately contributing to improvements in population health. Methods Study design and participants This cross-sectional study was conducted from July to October 2023 across Anhui Province, China. The inclusion criteria for participants were as follows: (1) aged 15–69 years; (2) permanent residents (living in the local area for a cumulative period of ≥ 6 months within the past 12 months, regardless of household registration status). Exclusion criteria were: (1) residents of military bases, hospitals, prisons, nursing homes, dormitories, or other collective living settings; (2) those with a diagnosis of mental disorders, hearing/speech disabilities, or other conditions that precluded the ability to complete the survey as assessed via pre-screening questionnaires. All participants provided written informed consent after a detailed explanation of the study’s objectives, procedures, and potential risks and benefits. The study received ethical approval from the Anhui Provincial Center for Disease Control and Prevention’s Ethics Review Committee. Data collection Data collection was administered through in-home interviews. Participants primarily completed the questionnaire independently. For those who were unable to fill out the questionnaire independently (e.g., due to lower educational levels or other reasons), the interviewers conducted face-to-face interviews to assist with questionnaire completion. All interviewers had received professional training and passed the assessment, thereby qualifying them to carry out the survey activities. Sampling strategy and sample size A multi-stage stratified probability proportional to size (PPS) sampling method was employed to select participants from 39 monitoring sites in Anhui Province. For further details on the sampling procedure, refer to our previous publication21. The sample size for each monitoring site was determined using the formula \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:N=\frac{{u}_{\alpha\:}^{2}\times\:p\times\:\left(1-p\right)}{{\delta\:}^{2}}\times\:deff$$\end{document}. Based on the health literacy level of Anhui Province in 2022 (29.60%)21, p was set as 0.2960; the allowable relative error was 15%; thus, the allowable absolute error was calculated as δ = 0.2960 × 0.15 = 0.0444. With µα = 1.96 and deff = 1, the minimum sample size for each site was calculated as 406. To account for invalid questionnaires and non-response rates, the sample size was expanded to 450 individuals per monitoring site. Overall, 17,941 questionnaires were collected, of which 16,067 were validated and included in the final analyses. Survey instruments and assessment criteria The Chinese Health Literacy Scale (CHLS)22, developed by the Chinese Center for Health Education, was employed as the survey instrument to collect data on participants’ demographic characteristics, health status, and health literacy. Demographic characteristics included place of residence, age, gender, occupation, education level, and marital status. Health status primarily focused on chronic disease status, assessed based on self-reported medical history. Participants were categorized into three groups: those without self-reported chronic diseases (without chronic diseases group), those with one chronic disease (single chronic disease group), and those with ≥ 2 chronic diseases (multimorbidity group). Chronic diseases included in this study were hypertension, heart disease, cerebrovascular diseases (e.g., stroke, cerebral infarction, cerebral thrombosis), diabetes mellitus, and cancer. The CHLS comprised 50 items, organized within two conceptual frameworks. Based on the KAP theory, it was divided into three domains: basic knowledge and attitudes, healthy lifestyles and behaviors, and health-related skills. Guided by major public health issues, it was also categorized into six categories: scientific views of health, infectious diseases prevention and treatment, chronic diseases prevention and treatment, safety and first aid, primary medical care, and health information. The scale demonstrated excellent internal consistency, with an overall Cronbach’s alpha of 0.95 and a Spearman-Brown coefficient of 0.9423. The scale had a total of score of 66 points. Participants were classified as having adequate HL if scoring ≥ 53 points (≥ 80% of the total score), and limited health literacy if scoring 0–52 points. Additionally, the criterion for judging adequate HL in each domain or category was defined as achieving a score of ≥ 80% of the total score for that specific domain or category24. Statistical analysis All statistical analyses were performed using R software (version 4.4.1), with a two-sided significance level set at α = 0.05. Categorical variables were presented as frequencies and percentages, while continuous variables with normal distribution were described as means ± standard deviations (SD). The Cochran-Armitage trend test was applied to assess linear trends in health literacy levels across chronic disease status groups. To account for the multistage sampling design, we employed multilevel logistic regression models with monitoring sites as a random effect, using the lme4 package. The multivariable model included a pre-specified set of covariates chosen a priori based on their established theoretical relevance as potential confounders from the literature. Health literacy was specified as the outcome (dependent) variable, and chronic disease status was the primary exposure (independent) variable. To assess potential effect modification, we tested interaction terms between chronic disease status and key covariates (age group, education level, and place of residence) using likelihood ratio tests comparing models with and without interaction terms. Models were used to estimate adjusted odds ratios (aORs) and 95% confidence intervals (CIs), with participants without chronic diseases as reference. Model fit was assessed using deviance (-2 log likelihood), Akaike Information Criterion (AIC), Bayesian Information Criterion (BIC), and McFadden’s pseudo-R². Multicollinearity was assessed using Variance Inflation Factors (VIF). Subgroup analyses by health literacy domains and categories were performed using the same multivariate framework. Results Basic characteristics A total of 17,941 individuals were surveyed, with 16,067 valid questionnaires recovered, for an overall effective rate of 89.55%. Among the respondents, 6952 (43.27%) were from urban areas and 9115 (56.73%) were from rural areas; 7228 (44.99%) were male and 8839 (55.01%) were female. The average age of participants was 43.23 ± 14.82 years. More than half of the respondents (50.44%) were farmers or workers, and 30.86% had a junior high school education level. Over three-quarters of the respondents (75.57%) were married. Notably, 75.25% of the respondents reported without any chronic diseases, while 19.18% and 5.58% reported with a single chronic disease and multimorbidity, respectively. Additionally, 31.90% of the respondents with adequate health literacy. Basic characteristics of the participants are summarized in Table 1. Table 1Basic characteristics.VariableTotal (N = 16,067)Health literacy χ2 valueP valueAdequate (n = 4995)Limited (n = 11072)Place of residence88.03< 0.001 Urban6952 (43.27)2434 (48.73)4518 (40.81) Rural9115 (56.73)2561 (51.27)6554 (59.19)Gender40.14< 0.001 Male7228 (44.99)2432 (48.69)4796 (43.32) Female8839 (55.01)2563 (51.31)6276 (56.68)Age group (years)1494.45< 0.001 15–344464 (27.78)2378 (47.61)2621 (23.67) 35–544999 (31.11)2134 (42.72)4470 (40.37) 55–696604 (41.10)483 (9.67)3981 (35.96)Education level3114.14< 0.001 Primary school or below4423 (27.53)305 (6.11)4118 (37.19) Junior high school4959 (30.86)1256 (25.15)3703 (33.44) Senior high school2933 (18.25)1068 (21.38)1865 (16.84) College or above3752 (23.35)2366 (47.37)1386 (12.52)Occupation1459.79< 0.001 Government/Institutional staff2225 (13.85)1208 (24.18)1017 (9.19) Farmer/worker8104 (50.44)650 (13.01)859 (7.76) Student1509 (9.39)1681 (33.65)2548 (23.01) Other4229 (26.32)1456 (29.15)6648 (60.04)Marital status353.45< 0.001 Unmarried2996 (18.65)1327 (26.57)1669 (15.07) Married12,142 (75.57)3507 (70.21)8635 (77.99) Separated/divorced/widowed929 (5.78)161 (3.22)768 (6.94)Chronic disease status390.84< 0.001 Without chronic diseases12,090 (75.25)4240 (84.88)7850 (70.90) Single chronic disease3081 (19.18)651 (13.03)2430 (21.95) Multimorbidity896 (5.58)104 (2.08)792 (7.15) Table 1Basic characteristics.VariableTotal (N = 16,067)Health literacy χ2 valueP valueAdequate (n = 4995)Limited (n = 11072)Place of residence88.03< 0.001 Urban6952 (43.27)2434 (48.73)4518 (40.81) Rural9115 (56.73)2561 (51.27)6554 (59.19)Gender40.14< 0.001 Male7228 (44.99)2432 (48.69)4796 (43.32) Female8839 (55.01)2563 (51.31)6276 (56.68)Age group (years)1494.45< 0.001 15–344464 (27.78)2378 (47.61)2621 (23.67) 35–544999 (31.11)2134 (42.72)4470 (40.37) 55–696604 (41.10)483 (9.67)3981 (35.96)Education level3114.14< 0.001 Primary school or below4423 (27.53)305 (6.11)4118 (37.19) Junior high school4959 (30.86)1256 (25.15)3703 (33.44) Senior high school2933 (18.25)1068 (21.38)1865 (16.84) College or above3752 (23.35)2366 (47.37)1386 (12.52)Occupation1459.79< 0.001 Government/Institutional staff2225 (13.85)1208 (24.18)1017 (9.19) Farmer/worker8104 (50.44)650 (13.01)859 (7.76) Student1509 (9.39)1681 (33.65)2548 (23.01) Other4229 (26.32)1456 (29.15)6648 (60.04)Marital status353.45< 0.001 Unmarried2996 (18.65)1327 (26.57)1669 (15.07) Married12,142 (75.57)3507 (70.21)8635 (77.99) Separated/divorced/widowed929 (5.78)161 (3.22)768 (6.94)Chronic disease status390.84< 0.001 Without chronic diseases12,090 (75.25)4240 (84.88)7850 (70.90) Single chronic disease3081 (19.18)651 (13.03)2430 (21.95) Multimorbidity896 (5.58)104 (2.08)792 (7.15) Basic characteristics. Health literacy levels across dimensions in different chronic disease statuses In terms of overall health literacy levels, participants without chronic diseases (35.07%) had higher levels than those with single chronic disease (21.13%), who in turn had higher levels than those with multimorbidity (11.61%), with a statistically significant trend (Ptrend<0.001). This pattern was consistent across all three domains—basic knowledge and attitudes, healthy lifestyles and behaviors, and health-related skills—as well as among all six categories of health questions: scientific views of health, infectious diseases prevention and treatment, chronic diseases prevention and treatment, safety and first aid, primary medical care, and health information (all Ptrend < 0.001). Details are shown in Figs. 1 and 2. Fig. 1Overall and domain-specific health literacy levels in different chronic disease statuses. The Cochran–Armitage trend test was used to assess the linear trend across chronic disease status groups. Overall and domain-specific health literacy levels in different chronic disease statuses. The Cochran–Armitage trend test was used to assess the linear trend across chronic disease status groups. Fig. 2Health literacy levels for six categories of health questions in different chronic disease statuses. The Cochran–Armitage trend test was used to assess the linear trend across chronic disease status groups. Health literacy levels for six categories of health questions in different chronic disease statuses. The Cochran–Armitage trend test was used to assess the linear trend across chronic disease status groups. Association between health literacy and chronic diseases Health literacy status (0 = Limited, 1 = Adequate) was set as the dependent variable, and chronic disease status was the primary exposure variable. A multivariable logistic regression model was employed, adjusting for the pre-specified covariates of place of residence, gender, age group, education level, occupation, and marital status, to assess the independent association between chronic disease status and health literacy. After adjusting for confounders, an inverse association was observed between health literacy and chronic diseases. Compared with participants without chronic diseases, those with multimorbidity had lower odds of adequate health literacy, with an aOR of 0.76 (95% CI 0.60–0.96, P = 0.021). Likelihood ratio tests for interaction showed no significant effect modification by age group (P = 0.332), education level (P = 0.065), or place of residence (P = 0.480). The main model demonstrated good fit (AIC = 16,122.000, BIC = 16,245.000, McFadden’s pseudo-R² = 0.182). All variance inflation factors were below 2.0, indicating no substantial multicollinearity concerns. Detailed model fit statistics for all models are provided in Supplementary Table S1. See Fig. 3 for details. Fig. 3Forest plot of univariate logistic regression and multivariable multilevel logistic regression results for the association between health literacy and chronic diseases. The multivariable logistic regression model was adjusted for place of residence, gender, age group, education level, occupation, and marital status. aOR = adjusted odds ratio; CI = confidence interval. Forest plot of univariate logistic regression and multivariable multilevel logistic regression results for the association between health literacy and chronic diseases. The multivariable logistic regression model was adjusted for place of residence, gender, age group, education level, occupation, and marital status. aOR = adjusted odds ratio; CI = confidence interval. Association between different dimensions of health literacy and chronic diseases After adjusting for confounders, multivariate logistic regression analysis showed that among the three domains of health literacy, compared with participants without chronic diseases, those with multimorbidity had lower odds of adequate basic knowledge and attitudes literacy and health-related skills literacy, with aOR of 0.81 (95% CI 0.67–0.97, P = 0.026) and 0.77 (95% CI 0.63–0.95, P = 0.014), respectively. For the six categories of health question literacy, compared with those without chronic diseases, participants with single chronic disease had lower odds of adequate infectious diseases prevention and treatment literacy (aOR = 0.85, 95% CI 0.76–0.97, P = 0.012), while those with multimorbidity were less likely to have adequate health information literacy (aOR = 0.82, 95% CI 0.69–0.97, P = 0.019). See Fig. 4 for details. Fig. 4Forest plot of multivariable multilevel logistic regression results for associations between health literacy dimensions and chronic diseases. Analyses were adjusted for place of residence, gender, age group, education level, occupation, and marital status. The three domains and six categories of health literacy were analyzed separately. aOR Adjusted odds ratio, CI Confidence interval. Forest plot of multivariable multilevel logistic regression results for associations between health literacy dimensions and chronic diseases. Analyses were adjusted for place of residence, gender, age group, education level, occupation, and marital status. The three domains and six categories of health literacy were analyzed separately. aOR Adjusted odds ratio, CI Confidence interval. Discussion The results of this study show that the health literacy level among residents aged 15–69 years in Anhui Province was 31.90% in 2023, slightly higher than the national level (29.70%)25 during the same period. However, compared with other provinces in China’s Yangtze River Delta region, such as Zhejiang Province (41.54%)26 and Shanghai City (40.46%)27, there is still a certain gap in Anhui Province, which may be related to differences in regional economic development levels, educational resource allocation, and medical infrastructure. The prevalence of chronic diseases in this study was 24.75%, lower than the national rate of 34.29%28, which may be attributed to the age range of the survey subjects in this study being limited to 15–69 years, while the prevalence of chronic diseases typically increases significantly with age, especially among the elderly. Furthermore, the health literacy levels among individuals with chronic diseases in this study was 18.98%, lower than that of the chronic disease population in Zhejiang Province in 2022 (23.52%)29, but higher than the overall level of health literacy among China’s chronic disease population from 2016 to 2022 (16.91%)30. In this large population-based study, we observed a graded inverse association between health literacy levels and chronic disease status, with the lowest health literacy level occurring in the multimorbidity group. While the strength of this gradient may vary across studies due to population characteristics, the consistent direction of this dose-response pattern aligns with the bulk of previous findings29,31,32. In our sample, this association remained robust after adjusting for key covariates such as age and educational attainment, underscoring the pervasive influence of multimorbidity on health literacy. This pattern reinforces that the number of chronic diseases is a key factor influencing health literacy. It also highlights that individuals with multimorbidity represent a high-risk group for health literacy deficiency33. On the one hand, as a critical competency for accessing and understanding health-related information, inadequate health literacy may directly or indirectly impact the prevention and control of chronic diseases17. On the other hand, multimorbidity is often accompanied by progressive cognitive decline, which directly impairs patients’ capacity to process health information, thereby contributing to inadequate health literacy34. Furthermore, due to the complex composition of multimorbidity, compared with those with single chronic disease, they need to cope with exponentially increased health-related information and face more complex healthcare service needs and self-management activities, such as long-term medication use, regular self-monitoring, clinical tests, repeated medical visits, and lifestyle modifications35. Notably, improving health literacy can empower people to adopt healthier lifestyles and behaviors, and to better manage their own health, thus preventing the onset and progression of chronic diseases. Therefore, enhancing health literacy represents a fundamental and cost-effective strategy for the prevention and management of chronic diseases. Our analysis reveals that individuals with multimorbidity exhibit significantly reduced odds of adequate health literacy in the domains of “basic knowledge and attitudes” (aOR = 0.81; P = 0.026) and “health-related skills” (aOR = 0.77; P = 0.014). This aligns with prior evidence showing that patients with ≥ 2 chronic conditions struggle with disease-specific knowledge and self-management skills36,37. For instance, multimorbid patients often exhibit deficits in medication regimen comprehension, self-monitoring techniques (e.g., blood glucose or blood pressure tracking), and lifestyle modification adherence38. These findings highlight the need for targeted educational interventions addressing foundational knowledge gaps and skill-building for multimorbid populations. In category-level analysis, multimorbidity was associated with reduced “health information” literacy (aOR = 0.82, P = 0.019), consistent with studies showing individuals with multiple physical conditions face disproportionate challenges in accessing, interpreting, and applying health information32. These barriers include difficulty navigating complex healthcare systems, understanding prescription instructions, and discerning credible health resources39. It was noteworthy that community-based health literacy interventions have efficacy in improving information-seeking behaviors among multimorbid patients40. Therefore, in the future, strategies should prioritize health information literacy for multimorbidity patients through tailored interventions (e.g., digital platforms, peer education, provider training) to enhance their ability to acquire, interpret, and apply health information. The study has several limitations. First, due to its cross-sectional design, we cannot establish a causal association between health literacy and chronic diseases. Second, the self-reported nature of chronic diseases in our study may lead to recall bias. Some participants may not accurately report their disease status, especially for those with mild symptoms or who are not fully aware of their conditions. Third, the sample was only from Anhui Province, limiting the generalizability of the results to other regions with different population characteristics and health environments. Fourth, the study primarily explored the overall association between chronic disease status and health literacy, without conducting a detailed analysis of specific chronic disease types (e.g., hypertension, diabetes, or cardiovascular diseases). Fifth, we lacked data on certain behavioral and clinical factors (e.g., BMI, tobacco and alcohol use, physical activity) and detailed socioeconomic indicators, which may lead to residual confounding. Conclusion In conclusion, this large cross-sectional study of 16,067 individuals in Anhui Province, China, establishes a significant inverse association between health literacy and chronic diseases. Domain-specific analyses further identify “basic knowledge and attitudes” and “health-related skills” as core competencies disproportionately compromised in multimorbidity individuals. The convergence of these findings with international evidence underscores health literacy as a critical modifiable factor in chronic disease outcomes. Investing in improving population health literacy, particularly among high-risk groups like those with or at risk of multimorbidity, is not only a moral imperative but also a cost-effective approach to alleviate the burgeoning burden of chronic diseases in China and globally. Future longitudinal research is needed to elucidate causal pathways and evaluate the effectiveness of specific health literacy interventions on chronic disease outcomes. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 1