Heterogeneity in the association between sleep disturbances and suicide‑related symptoms in 15,396 adult outpatients

Sleep disturbances are established risk factors for suicidal thoughts and behaviors, but it remains unclear whether distinct sleep profiles differentially associate with specific suicide‑related symptoms after adjustment for psychiatric symptom dimensions. In this cross-sectional, person-centered analysis of 15,396 adult outpatients with sleep complaints from a psychiatric specialist hospital in China, sleep profiles were identified via latent profile analysis (LPA) using seven Pittsburgh Sleep Quality Index (PSQI) dimensions. Suicide-related symptoms were assessed with three Symptom Checklist-90 (SCL-90) items: active suicidal ideation, death-related thoughts, and guilt. Logistic regression adjusted for age, sex, and other SCL-90 dimensions examined associations between sleep profiles and each outcome, with false discovery rate (FDR) correction. After adjustment for age, sex, and other psychiatric symptom dimensions, three profiles (Mild Disturbance, Severe Insomnia, and Moderate Non‑Medicated) showed elevated odds of active suicidal ideation (aORs 1.22–1.25). In exploratory adjusted analyses, death-related thoughts showed suggested associations with the Medication Use but Functional and Moderate Non-Medicated profiles, but these did not remain statistically significant after FDR correction and the overall sleep profile effect was marginal. Guilt showed no significant associations after adjustment. Notably, the Moderate Non‑Medicated profile had the highest symptom rates despite milder sleep impairment. These findings indicate that suicide‑related symptoms are not driven solely by sleep severity but follow profile‑specific patterns, supporting refined clinical risk stratification. Not applicable.