Objectives: Previous studies indicate substantial correlations between low health literacy and poor health outcomes. However, empirical findings remain inconsistent and are theoretically challenging. In this study, we conceptually place health literacy within an established model of health inequality. Studying multiple pathways, we estimate the associations between health literacy and six health outcomes and decompose these associations with health literacy's covariates.
Methods: Cross-sectional data from the Young Adult Survey Switzerland was used for the analyses (n = 5959, age = 18-25). Logistic regression and KHB decomposition analyses were applied to estimate health literacy's coefficients and confounding percentages.
Results: Eleven covariates were associated with health literacy (p < 0.001). Ten covariates reduced the naïve health literacy coefficient when included in the regression models (confounding percentages: 36.7-86.9%). In three out of six models, the confounding effects led to non-significant health literacy coefficients.
Conclusions: We found that health literacy's associations with health outcomes are confounded by socioeconomic, material, psychosocial, and health-related factors. More investigations on the causal importance of health literacy, respectively, on its potential to health promotion are required.
Keywords: Decomposition analysis; Determinants of health; Health behavior; Health literacy; Health status; Young adults.