Background: Feasible, valid, and reliable tools are needed to assess dietary quality among preschoolers. We aimed to test construct-/criterion-related validity and reliability of a dietary quality index generated from a liking survey and novel dietary patterns for characterizing food hedonism and parental feeding practices.
Methods: Participants included 416 economically disadvantaged, diverse preschoolers (41% overweight/obese) from educational centers and their parents. Parents completed liking and frequency surveys; the liking survey took half as long to complete. Preschoolers' skin carotenoid status (measured by Resonance Raman spectroscopy) and BMI percentile (measured weight/height) were assessed. The Healthy Eating Preference Index (HEPI) was constructed from weighted averages of liking scores for food groups and healthy variety score and Healthy Eating Index 2010 (HEI) from nutrient analysis of the frequency survey.
Results: The HEPI was normally distributed and showed construct validity and adequate internal reliability. In hierarchical regression analyses, the HEPI explained carotenoid status and adiposity as alternative or value-added predictors to HEI, supporting criterion-related validity of HEPI. Parental reporting of children's liking of high-fat/sweet/salty foods in excess of pleasurable activities (food hedonism) predicted heavier preschoolers, as did discord between HEI and HEPI (potential parental pressure toward healthy eating). HEPI alone or with HEI explained variability in carotenoid status and adiposity in path models with adequate to good fits.
Conclusions: With simple collection and processing, the liking survey can generate a valid/reliable dietary quality index in child care settings to identify preschoolers at risk for lower nutritional status. Using liking and frequency-based screeners could improve understanding of parental feeding behaviors and precision of predicting nutritional status.