The explanation of educational disparities in adiposity by lifestyle, socioeconomic and mental health mediators: a multiple mediation model

Cardiovascular diseases remain the leading cause of death in Czechia, while the prevalence of cardiometabolic drivers (obesity, prediabetes, diabetes) has not substantially changed. Poor cardiometabolic health is associated with lower education, which is a strong determinant of cardiovascular mortality in the Czech population. Prior research in Czech and other high-income populations shows an inverse association between socioeconomic position and obesity. Education likely influences adiposity through downstream factors rather than directly. Lower educational levels are linked to poorer lifestyle, socioeconomic disadvantage, and mental health problems, all of which are associated with increased adiposity. Sex differences are notable: women have higher body fat percentage and tend to have lower education and income, while men are more likely to be obese and differ in dietary patterns and adherence to recommendations. This study investigates educational disparities in adiposity and the role of potential mediators among middle-aged men and women in Brno, Czech Republic, with the hypothesis that lifestyle, socioeconomic, and mental health factors mediate the association, and that women may have a wider set of mediators than men.

Prior studies in Czechia and internationally have documented a strong inverse association between education and obesity. The HAPIEE study and other national data link lower education with higher obesity and cardiometabolic risk. Education is associated with health behaviors (diet, smoking), socioeconomic circumstances (income), and mental health, which in turn are linked to adiposity. Sex differences in adiposity and behavior have been reported, including higher body fat percentage in women, higher obesity prevalence in men, and differing dietary behaviors and adherence. Evidence also suggests higher sedentary behavior in higher socioeconomic groups due to occupational sitting. These strands of literature motivate testing multiple mediators, potentially sex-specific, along the pathway from education to adiposity.

Design and population: Cross-sectional analysis using data from the Kardiovize study, an epidemiological cohort of adult residents of Brno, Czech Republic. A random age- and sex-stratified sample of 2,154 adults aged 25–64 years (54.6% women) was enrolled in January 2013 with assistance from health insurance companies. Data collection: In-person interviews by trained nurses/physicians at the ICRC, St Anne’s University Hospital, Brno. Questionnaires assessed demographics, socioeconomic characteristics, cardiovascular risk behaviors (diet, alcohol, smoking, physical activity/sedentary time), medical history, and mental health. Anthropometrics included measured height, weight, waist circumference; body composition by bioelectrical impedance (InBody 370). Predictor and outcome: Education categorized as high (higher professional or university), middle (high school), and low (elementary/vocational without final graduation). Adiposity was modeled as a unidimensional latent variable indicated by body fat percentage, BMI, waist circumference, and visceral fat; unidimensionality confirmed via confirmatory factor analysis (CFA). Potential mediators: Dietary risk score (0–6) based on six risky dietary patterns identified per Global Burden of Disease methodology from a 43-item FFQ (frequency over past week). Alcohol intake: grams of ethanol consumed in prior 7 days (7-day recall). Smoking: current, ex-, non-smoker categories. Sedentary behaviors: total sedentary time (minutes/week) from the long IPAQ. Income: equivalized household income computed from household income categories and equivalence scale (1.0 first adult; 0.5 for each additional member). Stress: Perceived Stress Scale (PSS-10; 0–40). Depression: PHQ-9 (0–27). Quality of life: self-rated 0–100. Mental health was modeled as a latent factor due to strong intercorrelations among PSS, PHQ-9, and quality of life. Statistical analysis: Conducted separately for men and women using STATA 16 and Mplus 8.6. Descriptive statistics (means/frequencies); group differences across education assessed via one-way ANOVA or chi-square. Bivariate correlations examined mediator collinearity; mental health latent factor constructed. Adiposity latent factor constructed from continuous indicators. Age-adjusted simple mediation analyses tested each mediator individually, estimating total, direct, and indirect effects. Multiple mediation SEMs included all mediators simultaneously, modeling paths from education to each mediator and from mediators to adiposity, estimating direct and indirect effects. Significance of indirect effects assessed via bootstrapping with 5,000 resamples. Model fit indices reported (χ², CFI, RMSEA with 90% CI).

- Sample: 2,154 adults (54.6% women); mean age 46.7 years (men) and 47.8 years (women). Adiposity indicators were worse at lower education in both sexes. - Education–mediator associations: In both sexes, lower education associated with higher dietary risk, higher tobacco use, lower sedentary time, and lower income. In women, higher education associated with higher alcohol intake; in men, alcohol did not differ by education. Higher education associated with better quality of life in both sexes; depressive symptoms were higher with lower education; stress did not differ by education. - Simple mediation (age-adjusted): Direct effect of education on adiposity significant in both sexes (p < 0.001). Men showed a significant indirect effect via sedentary behavior (β ≈ 0.042; 95% CI [0.026, 0.062]; mediation ≈ 24%). Women showed significant indirect effects via dietary risk (β ≈ −0.026), alcohol consumption (β ≈ −0.005), sedentary behavior (β ≈ 0.012), and mental health (β ≈ −0.011). - Multiple mediation SEM model fit: Men: χ²[62] = 319.5, p < 0.001; CFI = 0.959; RMSEA = 0.065 (90% CI 0.058–0.072). Women: χ²[62] = 395.1, p < 0.001; CFI = 0.959; RMSEA = 0.067 (90% CI 0.061–0.074). - Paths from education to mediators: • Men: higher education predicted lower dietary risk (β = −0.14, p < 0.001), higher smoking (β = 0.08, p < 0.05), higher sedentary behavior (β = 0.24, p < 0.001), and higher income (β = 0.30, p < 0.001). • Women: higher education predicted lower dietary risk (β = −0.16, p < 0.001), lower smoking (β = −0.17, p < 0.001), higher alcohol intake (β = 0.10, p < 0.05), higher sedentary behavior (β = 0.08, p < 0.05), higher income (β = 0.34, p < 0.05), and better mental health (β = 0.10, p < 0.05). - Direct effects of education on adiposity (multiple mediation): Men β = −0.21 (95% CI [−0.277, −0.149]); Women β = −0.10 (95% CI [−0.158, −0.044]). - Total indirect effects: Men β = 0.038 (95% CI [0.010, 0.068]); Women β = −0.044 (95% CI [−0.070, −0.019]). - Significant mediators in multiple mediation: • Men: Sedentary behavior (β = 0.041; 95% CI [0.025–0.062]); mediation ratio 23.7%. • Women: Dietary risk (β = −0.023; 95% CI [−0.037, −0.013]); Alcohol intake (β = −0.006; 95% CI [−0.014, −0.001]); Sedentary behavior (β = 0.012; 95% CI [0.004, 0.023]); Income (β = −0.022; 95% CI [−0.041, −0.004]); Mental health (β = −0.007; 95% CI [−0.019, −0.001]). Total mediation ratio in women: 30.5%. - Interpretation: Sedentary behavior exerts an opposite-direction indirect effect (higher education → more sedentary time → higher adiposity), reducing the protective effect of higher education, particularly in men.

The study aimed to explain educational disparities in adiposity via lifestyle, socioeconomic, and mental health pathways, assessing sex-specific mediation. Findings support a robust inverse association between education and adiposity in both sexes. In women, multiple mediators—especially dietary risk and income—partially explain why lower education is linked to higher adiposity, consistent with hypotheses that economic disadvantage and poorer diet quality contribute to adiposity. Mental health and alcohol intake also contributed modestly. In men, however, the expected mediators did not explain the inverse education–adiposity gradient; only sedentary behavior mediated the association, and in the opposite direction, indicating that higher education is associated with more sedentary time which increases adiposity, thereby attenuating the protective impact of education. This aligns with evidence that sedentary jobs are more common among higher-education groups. The results suggest that reducing sedentary time could enhance the protective effect of higher education and should be prioritized in public health interventions, especially among higher socioeconomic groups. Given that only about one-third of the association in women and one-fifth in men was explained by the included mediators, additional unmeasured pathways—such as health literacy, self-management skills, specific eating behaviors, and broader social/physical environmental factors—likely contribute and merit investigation.

Educational disparities in adiposity differ by sex. Sedentary behavior undermines the protective effect of higher education in both sexes, more strongly in men. Among women, risky dietary patterns and lower income substantially mediate the higher adiposity associated with lower education, with additional contributions from mental health and alcohol. Public health strategies should focus on reducing sedentary time across educational strata—particularly in higher socioeconomic groups—and on improving diet quality and addressing economic disadvantage among women. Future research should incorporate broader mediators, including health literacy, self-management, detailed eating behaviors, and social/physical environmental exposures, to more fully explain education-related disparities in adiposity.

- Cross-sectional design precludes causal inference and raises potential for reverse causation. - Urban, city-based sample from Brno limits generalizability beyond urban populations. - Many mediators were self-reported, introducing potential reporting/recall bias; objective measures were not available. - The specific FFQ version used was not previously validated in other studies, although developed following recommended practices.