Testing a machine-learning algorithm to predict the persistence and severity of major depressive disorder from...

Area of Science:

• Psychiatry and Mental Health
• Computational Medicine
• Epidemiology

Background:

• Major Depressive Disorder (MDD) illness course heterogeneity complicates clinical decision-making.
• Previous attempts to develop prognostic subtypes using symptom profiles or biomarkers have yielded limited success.
• Machine learning (ML) models show promise in predicting MDD outcomes from self-reported data.

Purpose of the Study:

• To validate machine learning (ML) models for predicting major depressive disorder (MDD) persistence, chronicity, and severity in an independent prospective cohort.
• To compare the predictive accuracy of ML models against conventional logistic regression models.

Main Methods:

• Applied previously developed World Health Organization World Mental Health (WMH) Surveys ML models to baseline self-report data from 1056 respondents with lifetime MDD.
• Validated predictions against observed MDD outcomes assessed 10-12 years post-baseline.
• Compared ML model performance (Area Under the Curve) with logistic regression models.

Main Results:

• ML models demonstrated consistently higher prediction accuracy (AUC 0.63-0.76) than logistic models (AUC 0.62-0.70) for MDD chronicity and severity.
• A significant proportion of individuals with high chronicity (34.6-38.1%) and severity (40.8-55.8%) were correctly identified in the highest risk strata by ML models.
• ML models effectively identified low-risk individuals, with very few subsequent hospitalizations (0.9%) or suicide attempts (1.5%) in the lowest predicted risk group.

Conclusions:

• Clinically useful MDD risk-stratification models can be developed using baseline patient self-reports.
• Machine learning methods significantly improve upon conventional statistical approaches for developing prognostic models in MDD.
• These findings support the use of ML-based risk prediction for personalized clinical decision-making in major depressive disorder.