Assessing Sustainable and Healthy Diets in Large-Scale Surveys: Validity and Applicability of a Dietary Index Based on a Brief Food Group Propensity Questionnaire Representing the EAT-Lancet Planetary Health Diet
Miranda A., Murante AM., Manca F., Consalez F., Jani A., DeClerck F., Maillot M., Verger E.
Background: Ensuring healthy diets within planetary boundaries is essential. However, current instruments measuring adherence to the EAT-Lancet planetary health diet are unsuitable for large-scale surveys. Simplified tools assessing consumption frequency can improve response rates, lower costs, and facilitate administration. Objectives: This study aimed to develop a practical and concise index for evaluating relative adherence to the EAT-Lancet diet across large-scale multicountry surveys. Methods: First, the EAT-Lancet Consumption Frequency Index (ELFI) was developed using a brief food propensity questionnaire of 14 food groups representing the planetary health diet from the Food systems that support transitions to hEalthy And Sustainable dieTs survey, which encompassed 27 European countries (n = 27,417). Subsequently, ELFI was further validated using 24-h dietary recalls from the Third French Individual and National Food Consumption Survey (n = 1645), correlating it with the valid EAT-Lancet Index (ELI), which evaluates absolute adherence, as well as with food group consumption, measures of nutritional health (nutrient adequacy and diet quality), and environmental impact. Analyses included assessment of reliability, structural validity, concurrent validity, and nomological validity. Results: ELFI showed strong reliability (α > 0.80) and factor analysis revealed a 2-factor solution: “foods to encourage” and “foods to balance and to limit.” Confirmatory factor analysis demonstrated that ELFI is structurally valid. Concurrent validity was confirmed as it was associated with sex, age, education, income, household size, physical activity, and smoking habit (P < 0.05). ELFI correlated with ELI (0.44, P < 0.0001) and food group consumptions. Regarding nomological validity, the ELFI subscores for “foods to encourage” and “foods to balance and to limit” were associated with better nutritional health (β = 0.62 and 0.23, respectively; P < 0.0001) and a lower environmental impact (β = –0.16 and –0.36, respectively; P < 0.0001). Conclusions: ELFI approach represents a valuable and easy-to-implement index for evaluating relative adherence to sustainable and healthy diets in large-scale multicountry studies.