Sleep Spindle Measures in Japanese Patients With Drug‐Naïve Schizophrenia Using Standard Polysomnography Data: A Retrospective Pilot Study

BACKGROUND: Sleep spindle abnormalities in schizophrenia have recently attracted attention as candidate potential biomarkers. Using polysomnography (PSG) data previously recorded at Kurume University Hospital from untreated Japanese patients with schizophrenia, this exploratory pilot study evaluated whether PSG-derived spindle parameters provide discriminatory information between groups. METHODS: We analysed 7 antipsychotic-naïve, untreated patients with schizophrenia (DSM-IV) and 7 healthy controls (mean age 27.6 ± 5.0 years; sex ratio 4:3). Spindle indices included frequency, detection time, duration, and mean amplitude. Spindles were further classified as total (9-15 Hz), fast (12-15 Hz), and slow (9-12 Hz), and discriminative performance was assessed using receiver operating characteristic (ROC) curves and the area under the curve (AUC). RESULTS: Compared with healthy controls, patients with schizophrenia showed lower total spindle frequency and detection time, with the largest between-group differences at C3. At C3, total spindle frequency was 45.2/h in schizophrenia and 112.96/h in controls (p = 0.0041). Total spindle detection time was 289.0 and 835.8 s, respectively (p = 0.0041). In ROC analyses, the AUC for fast spindle frequency and fast spindle detection time at C3 was 0.939 for both measures (95% CI 0.755-1.000). CONCLUSIONS: Spindle indices derived from standard, clinically acquired PSG recordings may provide informative features for group discrimination in untreated schizophrenia. However, this study is exploratory with a small sample size, and larger prospective studies are required for validation.