Mental health at risk: Predicting psychological distress in Australian youth through machine learning models.

BACKGROUND: Psychological distress among youth is a growing global health concern and a leading non-communicable disease burden. Early and accurate prediction is vital for effective intervention. This study applied machine learning (ML) techniques to predict psychological distress risk in young Australians. METHODS: This study used data from the most recent two waves, 9C1 and 9C2, of the Longitudinal Study of Australian Children (LSAC). From an initial set of 31 features, relevant predictors were selected based on individual classification accuracy (threshold ≥0.60). Five ML algorithms (Decision Tree, Naive Bayes, Random Forest, Support Vector Machines, eXtreme Gradient Boosting) were developed and rigorously evaluated using 10-fold cross-validation repeated 25 times. Logistic regression was applied for interpretive insights into the top predictors. RESULTS: Random Forest (RF) model consistently demonstrated superior predictive performance across both waves, achieving higher accuracy (0.8168 and 0.8011), F1-scores (0.8276 and 0.8430), AUC values (0.8919 and 0.8833), and Matthews correlation coefficients (0.6321 and 0.5735), along with the lowest Brier scores (0.1348 and 0.1366). Key predictors included loneliness, bullying victimisation, social media addiction, total social support, stressful life events, and coping level. CONCLUSIONS: This study's innovative ML approach uncovers critical social and emotional risk factors for psychological distress in Australian youth. The findings highlight ML's significant role in enhancing early prediction, guiding targeted public health actions, and supporting clinical decisions to improve mental health outcomes for young people. Additionally, they reveal strong real-world potential for integration into school, community, and digital health initiatives, facilitating early detection and personalised mental health assistance.