Causal Machine Learning for Treatment Selection and Outcome Optimization in Mood Disorders

Mood disorders represent a major global burden and are characterized by substantial heterogeneity in symptom profiles, treatment response, and clinical trajectories. Although numerous machine learning models have been developed to predict psychiatric outcomes, most existing approaches remain primarily predictive and provide limited guidance for treatment decision-making in clinical practice. In psychiatry, where treatment choices often involve multiple options and uncertain benefits, models that can estimate individualized treatment effects and support optimal treatment selection are particularly needed. Recent advances in causal inference and causal machine learning offer promising tools to address this challenge by enabling the estimation of heterogeneous treatment effects and the evaluation of intervention strategies using both clinical trial data and real-world evidence. These approaches have the potential to move psychiatric research beyond prediction toward actionable decision support, helping clinicians identify which treatments are most likely to benefit individual patients.The goal of this Research Topic is to advance the development and application of causal machine learning approaches for treatment selection and outcome optimization in mood disorders. While predictive modeling has become increasingly common in psychiatric research, there remains a critical gap between prediction and actionable clinical decision-making. This Research Topic aims to address this gap by promoting methodological and applied research that leverages causal machine learning to estimate heterogeneous treatment effects, identify optimal intervention strategies, and support personalized treatment decisions. We seek to bring together interdisciplinary contributions from psychiatry, data science, epidemiology, and clinical informatics to explore how causal machine learning methods can improve treatment response prediction, evaluate benefit–risk profiles, and inform clinical decision support systems. Particular attention will be given to studies that integrate clinical trial data and real-world evidence, assess model robustness and generalizability, and demonstrate potential pathways toward clinical implementation. By highlighting advances in causal machine learning for psychiatric treatment optimization, this Research Topic aims to facilitate the development of clinically actionable, evidence-based decision support tools for mood disorders.This Research Topic welcomes contributions that explore methodological and applied advances in causal machine learning for treatment selection and outcome optimization in mood disorders. We are particularly interested in studies that develop, evaluate, or apply causal inference and machine learning approaches to support clinically meaningful treatment decisions. Relevant topics include, but are not limited to: estimation of heterogeneous treatment effects and individualized treatment response; causal modeling of treatment outcomes and adverse events; machine learning approaches for treatment recommendation systems and clinical decision support; integration of causal inference with deep learning or language-based models; and the use of real-world data, electronic health records, or multimodal clinical information to evaluate treatment strategies in psychiatric populations. We welcome original research articles, systematic reviews, methodological papers, and perspective articles that contribute to improving evidence-based, personalized treatment strategies for mood disorders. Topic editors: Enzhao Zhu, ZiSheng Ai, Jiaojiao Hou Source: https://www.frontiersin.org/research-topics/79308/causal-machine-learning-for-treatment-selection-and-outcome-optimization-in-mood-disorders