Background: It is well recognized that there is great individual variation in response to obesity treatment. However, little is known regarding baseline factors that predict individual response or nonresponse to intervention. This study identified profiles of socio-demographic and behavioral characteristics that distinguish Hispanic youth who responded and who were unresponsive to an established school-based obesity intervention.
Methods: Classification and Regression Trees (CRT) was used to group 77 Hispanic middle school students with BMI percentile ≥85 who received an effective school-based obesity intervention. CRT can detect interactions between predictors and does not have sample size limitations to detect effects. Predictors included baseline age, gender, standardized BMI, health-related quality of life (PedsQL), minutes of moderate-vigorous physical activity (MVPA; accelerometry), energy consumption, and dietary quality (number of foods consumed in each “food zone” taught in the intervention; Block Kids 2004 Food Frequency Questionnaire). Response at the end of the 12-week intervention was defined according to American Academy of Pediatrics guidelines.
Results: According to AAP guidelines, 53.2% of participants responded to the intervention. Seven distinct pathways were identified that predicted 95.1% of responsive and 75.0% of nonresponsive participants. Dietary quality (number of servings of “junk food” and fruits and vegetables) appeared twice in six of the pathways, immediately before and after minutes of MVPA, indicating a potential interaction between minutes of MVPA and dietary consumption.
Conclusion: The high number of distinct pathways and the repeated appearance of the same dietary variable within a pathway illustrates the complex, interactive nature of factors that predict intervention response. More complex modeling to account for this is needed to better understand how to best predict individuals who will be responsive to interventions.