[Formula: see text] Data-driven profiles of behavior in pediatric medical disorders

Area of Science:

• Child Psychology
• Neurodevelopmental Disorders
• Behavioral Science

Background:

• Behavioral impairments frequently co-occur with pediatric medical conditions, affecting various life domains.
• Prior research identified three distinct behavioral profiles (metacognition, emotion regulation, inhibition) in children with psychiatric disorders using graph-theory analysis.

Purpose of the Study:

• To investigate the presence and characteristics of these behavioral profiles in children with pediatric medical conditions.
• To assess the applicability of a classification model derived from a psychiatric sample to a medical sample.

Main Methods:

• Graph-theory analysis of parent-report behavior measures.
• Application of a support vector machine (SVM) classifier trained on a psychiatric cohort to a large clinical database of children with central nervous system medical conditions.

Main Results:

• Three similar behavioral profiles (metacognition, emotion regulation, inhibition) were identified in children with pediatric medical conditions.
• The SVM classifier achieved high accuracy when applied to the medical sample, indicating significant overlap but also subtle differences, especially in the inhibition profile.
• Findings suggest transdiagnostic behavioral profiles but highlight potential etiological differences.

Conclusions:

• The study supports the existence of three transdiagnostic behavioral profiles applicable to both psychiatric and pediatric medical populations.
• These profiles offer potential targets for personalized interventions.
• The etiology of behavioral problems (psychiatric vs. medical) warrants consideration for tailored therapeutic approaches.