Prediction of antipsychotic associated weight gain in children and adolescents taking second generation...

Second Generation Antipsychotics (SGA) are associated with serious cardiometabolic side effects in children and adolescents, especially antipsychotic-associated weight gain (AAWG). The objective of this study is to develop a prognostic-based machine learning (ML) algorithm that would dynamically predict the real-time risk of AAWG.

This study encompassed children and adolescents who were naïve SGA recipients included in the IQVIA Ambulatory EMR- US database between December 2016 and June 2021. The outcome predicted in this study was the relative change between the baseline BMI z-score and the last BMI z-score categorized as severe ( ≥ 0.5), moderate ( ≥ 0.25 and < 0.5), and minor (<0.25) AAWG. Four ML models (the Multiclass Logistic Regression (MLR) model, the Classification and Regression Trees (CART) model, the Multiclass Random Forest (MRF) model, and the Extreme Gradient Boosting (Xgboost) model were trained using 1) baseline features identified during 12-month period prior to and at SGA initiation and 2) both baseline and time-varying features identified during SGA treatment.

A total of 10,997 patients who met the eligibility criteria were identified. The proportions of patients who experienced minor, moderate, and severe weight gain were 64 %, 10 %, and 26 % respectively. Of the 4 models developed, Xgboost and MRF model trained using both baseline and time varying features demonstrated strong performance, achieving an AUC of 87.07 % and 87.06 % respectively.

The dynamic and individual-level change of AAWG in children and adolescents can be accurately predicted using ML algorithms. These algorithms can be applied in practice to guide personalized monitoring and timely interventions of AAWG.