Complete mediation of depression in the association between sleep disturbance and suicidal ideation among doctoral students: a cross-sectional study of 2300 participants

In the context of ongoing social transformation and educational reforms, mental health issues among doctoral students have attracted considerable attention. This study aims to explore the association of anxiety, depression and sleep disturbance with suicidal ideation among doctoral students, identify the independent influencing factors for suicidal ideation, and examine the mediating effect of depression on the relationship between sleep disturbance and suicidal ideation.

A total of 2300 doctoral students from Central China were selected using cluster sampling between March and April 2026. Generalized Anxiety Disorder Scale (GAD-7), Patient Health Questionnaire Depression Scale-9 (PHQ-9), Insomnia Severity Index (ISI), and Suicidal Behaviors Questionnaire-Revised (SBQ-R) were used for survey. The detection rates of anxiety, depression, sleep disturbance and suicidal ideation and their co-occurrence rates were calculated among the participants. Mann-Whitney U test and Kruskal-Wallis H test were employed to compare the distribution differences of symptom scores across groups. Spearman rank correlation analysis was used to explore the correlations of anxiety, depression and sleep disturbance with suicidal ideation. Binary logistic regression analysis was applied to identify the influencing factors for suicidal ideation. PROCESS Macro Model 4 and Bootstrap method were carried out to test the mediating effect of depression between sleep disturbance and suicidal ideation, whereas PROCESS Macro Model 59 was conducted to examine the moderating effect of gender and grade on the mediating effect.

The detection rates of anxiety, depression, sleep disturbance and suicidal ideation were 25.6%, 27.3%, 22.7% and 7.8%, respectively, among the participants. The co-occurrence rates were 18.7% for anxiety and depression, 14.1% for anxiety and sleep disturbance, and 16.2% for depression and sleep disturbance. Female doctoral students had significantly higher distributions of anxiety (Z=-3.962, P<0.001) and depression (Z=-2.732, P=0.006) scores than male students. Doctoral students in higher grades had significantly higher distributions of anxiety (H=42.732, P<0.001), depression (H=24.093, P<0.001), and sleep disturbance (H=11.187, P=0.011) scores than those in lower grades. Doctoral students aged ≥35 years had significantly higher distributions of suicidal ideation scores than those aged 30—<35 years (Z=2.458, P=0.042). Correlation analyses revealed that anxiety, depression, sleep disturbance and suicidal ideation were positively correlated pairwise (r= 0.247 to 0.744, P<0.001). Binary logistic regression analysis showed that depression was an independent influencing factor for suicidal ideation (OR=1.269, 95%CI: 1.179 to 1.365, P<0.001). Mediation analysis demonstrated that depression played a complete mediating role between sleep disturbance and suicidal ideation (indirect effect size=0.032, 95%CI: 0.024 to 0.040). Grade played a moderating role in the first-stage path from sleep disturbance to depression (β=0.038, P<0.001), whereas gender moderated the second-stage path from depression to suicidal ideation (β=0.015, P<0.05).

Anxiety, depression, and sleep disturbance are significantly positively correlated with suicidal ideation. Depression is a significant predictor of suicidal ideation, and plays a complete mediating role between sleep disturbance and suicidal ideation. Universities may construct a hierarchical intervention system covering sleep health management, depressive symptom screening, and suicide risk prevention and control. Improving the sleep quality may reduce depressive symptoms, weaken the impact of sleep disturbance on suicidal ideation, so as to facilitate the pre-emptive prevention and targeted intervention for psychological crises among doctoral students.