What Predicts Active Parent Participation in a Pediatric Obesity Trial? An AAP Pediatric Research in Office Settings (PROS) Study

Wright M, Shone LP, Delacroix E, Sonneville KR, Steffes J, Harris D, Woo H, Fiks AG, Stockwell MS, Resnicow K

Accepted for 2020 Pediatric Academic Societies Annual Meeting

Background: Understanding factors that contribute to active participation in behavioral trials is an important research priority.

Objective: Examine associations of 1) demographic characteristics and 2) parent perceptions and concerns about their child's weight and health with active parent participation in a pediatric obesity treatment trial.

Methods: Parents of 3 to 12 year-old children with a BMI ≥85^th percentile were recruited for the NIH-funded BMi²⁺ RCT. Parents (n=254) at 9 intervention practices in the AAP PROS network enrolled and completed a baseline survey (English or Spanish) about 1) demographic characteristics and 2) perception of their child's weight and concern about their child's weight and health. During the study, parents have up to 6 motivational interviewing (MI)-based counseling calls (English or Spanish) with a study dietitian (RD). We defined active parent participation as completion of ≥1 MI counseling call. Due to rolling enrollment, parents' time-in-study varies, with maximum follow-up of 24 months. We used Chi square and Wilcoxon rank sum tests for bivariate analyses, and multivariable logistic regression, adjusted for the number of days since enrollment, to determine odds ratios (aOR) and 95% confidence intervals (CI).

Results: 1) Children were 51% female and 40% Latino, 72% had private insurance, and parents were highly educated (33% had a graduate degree). Parents' median time-in-study was 465 days (min=198, max=715). Most parents (82%) completed ≥1 RD call (among those, median=3, range=1-6 calls). In bivariate analyses, parents of Latino children were less likely to actively participate than parents of non-Latino children (75% vs 88%, p<.01), and parents with a graduate degree were more likely to actively participate than those without one (89% vs 79%, p<.05). In multivariable analyses, Latino ethnicity was significantly associated with lower odds of active parent participation (aOR=0.41, 95% CI=0.21-0.81), but parent education was NS. 2) Parent BMI, perception of their child's weight, and concern about their child's weight and health were not associated with active participation.

Conclusion: In this sample of parents enrolled in an obesity treatment trial, there was one independent predictor of active parent participation (child ethnicity). Future analyses will help determine how active participation is associated with outcomes.