Improving Behavioral Nutrition Interventions: A Structural Equation Modeling Study to Model Pathways Among Consumer Food Characteristics and Health

BACKGROUND: Consumer food characteristics have been correlated with dietary behaviors. However, the pathways in which these factors relate to one another and impact health outcomes have yet to be elucidated. OBJECTIVE: We used structural equation modeling to identify pathways between consumer food characteristics and self-reported health outcomes. We constructed a conceptual model and hypothesized 10 pathways in the model based on previous literature. METHODS: Eight-hundred and eighty adults completed a series of measures for different food consumer characteristics and their health. We used Mplus to estimate and validate our hypothesized model of pathways between food consumer characteristics. We then applied multiple group analyses to determine the magnitude of associations found in the final structural model across demographic and food consumer characteristics by conducting measurement invariance testing. RESULTS: Our final structural model yielded excellent fit (comparative fit index = 0.960, Tucker-Lewis index = 0.954, root mean square error of approximation = 0.047, standardized root mean square residual = 0.050) with all hypothesized paths being statistically significant. Results also revealed differences in the strength of some of the pathway relations based on age, education, income, grocery spending, snack spending, and active control of eating. CONCLUSIONS: Findings have implications for how to improve consumer food characteristics to enhance health and well-being, with particular attention to nutrition self-efficacy.