Differentiation between depressive disorder and insomnia disorder with depressive symptoms across age groups using machine learning.

OBJECTIVE: To investigate symptom features distinguishing depressive disorder from insomnia disorder with depressive symptoms across different age groups, and to provide evidence for refined clinical differential diagnosis. METHODS: This retrospective cross-sectional study enrolled patients who visited Hangzhou Seventh People's Hospital between January 2016 and October 2024 and completed assessments via the "Good Sleep 365" platform. Based on DSM-5 diagnostic criteria, two psychiatrists classified participants into two diagnostic groups: depressive disorder and insomnia disorder with depressive symptoms. The final sample included 1,873 patients: 190 adolescents, 1,253 adults, and 430 older adults. After an 8:2 training-validation split, feature selection and model training were performed within the training set, and final performance was evaluated on the validation set. Age-specific logistic regression models were constructed using Pittsburgh Sleep Quality Index (PSQI) component scores and item-level data from the 9-Item Patient Health Questionnaire (PHQ-9) and 15-Item Patient Health Questionnaire (PHQ-15), with feature importance analyses performed. RESULTS: The two groups differed in demographic characteristics and total scale scores, although their overall clinical presentations showed substantial overlap. Depressed mood, fatigue, impaired concentration, and sleep-related symptoms consistently emerged across age groups, demonstrating relatively stable discriminative value, while distinct age-specific symptom patterns were also observed. Model performance varied across age groups, with the highest AUC observed in adults (0.8216), followed by older adults (0.7446) and adolescents (0.6229). CONCLUSION: Depressive disorder and insomnia disorder with depressive symptoms exhibit identifiable differences in symptom structure that are age-related. Symptom-level evaluation across age groups may provide complementary information for clinical differential diagnosis between these two conditions.