Validity of the energy-restricted Mediterranean Diet Adherence Screener

Abstract

Background: Short dietary assessment tools can be useful to estimate food intake and diet quality in large-scale epidemiological studies with time constraints. Objective: To determine the concurrent validity of the 17-item energy-restricted Mediterranean Adherence Screener (er-MEDAS) used in the PREDIMED (PREvención con DIeta MEDiterránea)-Plus trial and to analyse its capacity to detect 1-year changes in diet and cardiometabolic risk factors. Methods: Validation study nested in the PREDIMED-Plus (n = 6760, 55–75 years). Dietary data were collected by the 17-item er-MEDAS and a 143-item validated semiquantitative food frequency questionnaire (FFQ) at baseline and after 1-year intervention. Cardiometabolic risk markers were measured at both time points. A Mediterranean diet (MedDiet) score was derived from both instruments. Concurrent validity was evaluated by Pearson and intra-class correlation coefficients (ICC) and Bland and Altman limits of agreement. Construct validity was evaluated by assessing 1-year changes in FFQ-reported dietary intake and cardiometabolic profile changes in relation to changes in er-MEDAS. Results: A moderate to good correlation between the MedDiet score calculated by both measurement instruments was found: r = 0.61 and ICC = 0.60 (both p < 0.001). Agreement of each of the er-MEDAS items ranged from 55.4% to 85.0% with a moderate mean concordance (kappa = 0.41). Between baseline and 1-year follow-up, energy intake measured by the FFQ decreased by 242 kcal, while Mediterranean food consumption increased in participants with the highest increase in the er-MEDAS MedDiet score. An increase in the er-MEDAS MedDiet score ratings was associated with a decrease in BMI, waist circumference, triglycerides, fasting glucose, diastolic blood pressure, and triglycerides/HDL-cholesterol ratio (p < 0.001 for all), and with an increase in HDL-cholesterol (p = 0.006). Conclusion: The er-MEDAS shows a modest to good concurrent validity compared with FFQ data. It shows acceptable construct validity, as a greater er-MEDAS score was associated with more favourable dietary and cardiometabolic profiles over time.