# Differential directional effects between insomnia symptoms and suicidal ideation across trait depression levels: a cross-lagged network analysis among Chinese medical undergraduates

## TL;DR

This study explores how insomnia and depression traits influence suicidal thoughts in Chinese medical students, finding that specific insomnia symptoms and emotional factors are key.

## Contribution

The study reveals bidirectional relationships and identifies critical nodes in the network linking insomnia, depression traits, and suicidal ideation.

## Key findings

- AIS6 and AIS7 insomnia symptoms strongly predict suicidal ideation factors like despair and suicide.
- Trait anhedonia (TAN) is a central node influenced by insomnia and suicidal ideation.
- High-TAN and low-TAN groups differ in key bridging nodes linking insomnia and emotional factors.

## Abstract

Suicidal ideation (SI) is intricately linked with insomnia and trait depression, yet the directional relationships and the role of trait depression remain unclear. This study sought to investigate the dynamic interplay between insomnia symptoms, SI, and trait depression (including trait anhedonia (TAN) and trait dysthymia (TDY)), aiming to clarify the role of trait depression in the relationship between insomnia and SI.

A longitudinal design was employed to assess 566 undergraduate students (aged 18-25) recruited from a medical university in China. Participants underwent comprehensive assessments with a one-month interval between baseline and follow-up, applying the Athens Insomnia Scale (AIS), the Self-rating Idea of Suicide Scale (SIOSS), and the Trait Depression Scale (TDS). Cross-lagged panel network (CLPN) models were implemented to examine temporal associations, centrality metrics, and network differences between high/low TAN and TDY subgroups. Network stability was evaluated using bootstrap methods.

Insomnia symptoms, particularly AIS6 (sense of well-being during the day) and AIS7 (functioning), emerged as pivotal nodes significantly predicting SI factors, including despair (DSP) and suicide (SUI), with bidirectional feedback observed. TAN emerged as a central node, strongly influenced by insomnia and SI. TDY primarily influenced TAN and optimism (OPT). In the high-TAN group, OPT was a mediator among the nodes, OPT, AIS2 (awakening during the night), and AIS7 were key bridging nodes, whereas AIS3 (final awakening earlier than desired), AIS8 (sleepiness during the day), and DSP bridged in the low-TAN group. High/low TDY networks exhibited structural congruence but significant differences in bridge centrality rankings.

Insomnia symptoms exacerbate SI by impairing daytime functioning and emotional regulation, with trait anhedonia serving as a critical node. Personalized interventions targeting specific insomnia symptoms (e.g., AIS6, AIS7 or AIS8) and suicidal emotional factors (e.g., OPT or DSP) are crucial disrupting feedback loops or critical connections to reduce suicide risk, particularly in individuals with varying levels of trait anhedonia. Although medical undergraduates represent a population commonly affected by mental health problems, the specialized nature of our sample may limit the generalizability of our findings. Future research and validation should be conducted in more diverse populations.