Abstract
Clinical settings need rapid yet useful methods to screen for diet and activity behaviors for brief interventions and to guide obesity prevention efforts. In an urban pediatric emergency department, these behaviors were screened in children and parents with the 33-item Pediatric-Adapted Liking Survey (PALS) to assess the reliability and validity of a Healthy Behavior Index (HBI) generated from the PALS responses. The PALS was completed by 925 children (average age = 11 ± 4 years, 55% publicly insured, 37% overweight/obese by Body Mass Index Percentile, BMI-P) and 925 parents. Child-parent dyads differed most in liking of vegetables, sweets, sweet drinks, and screen time. Across the sample, child and parent HBIs were variable, normally distributed with adequate internal reliability and construct validity, revealing two dimensions (less healthy-sweet drinks, sweets, sedentary behaviors; healthy-vegetables, fruits, proteins). The HBI showed criterion validity, detecting healthier indexes in parents vs. children, females vs. males, privately- vs. publicly-health insured, and residence in higher- vs. lower-income communities. Parent's HBI explained some variability in child BMI percentile. Greater liking of sweets/carbohydrates partially mediated the association between low family income and higher BMI percentile. These findings support the utility of PALS as a dietary behavior and activity screener for children and their parents in a clinical setting.