Pediatric Adapted Liking Survey: A Novel, Feasible and Reliable Dietary Screening in Clinical Practice

Abstract

Purpose: Rapid yet useful methods are needed to screen for dietary behaviors in clinical settings. We tested the feasibility and reliability of a pediatric adapted liking survey (PALS) to screen for dietary behaviors and suggest tailored caries and obesity prevention messages. Methods: In an observational study, children admitted to a pediatric emergency department (PED) for nonurgent care were approached to complete the PALS (33 foods, 4 nonfoods including brushing teeth). Measured height/weight were used for body mass index (BMI) percentile determination. Feasibility was assessed by response rate and PALS completion time. Reliability was assessed by internal consistency of food groups and test-retest reliability for PED-home reported PALS. Results: PALS was completed by 144 children (96% of approached) - 54% male (average age = 11 ± 3 years) with diversity in family income (43% publicly insured), race/ethnicity (15% African American, 33% Hispanic, 44% Caucasian) and adiposity (3% underweight, 50% normal, 31% overweight, 17% obese, 8% extremely obese). The average completion time was 3: 52 min, and conceptual food groups had reasonable internal reliability. From 57% (n = 82) with PED-home completion, PALS had a good/excellent test-retest reliability. Relative preferences for sweets versus brushing teeth identified unique groups of children for tailored prevention messages (high sweet/brushing preference, sweets > brushing, brushing > sweets). Females with higher adiposity reported significantly greater preference for sweet/high-fat foods, independently of demographic variables; the relationship was nonsignificant in males and with the other food groups. Conclusion: The PALS appears to be a fast, feasible and reliable dietary screener in a clinical setting to assist in forming tailored diet-related messages for dental caries and obesity prevention.