What Increases the Risk of Sleep Problems for Train Drivers? Evidence From Network Analysis

Previous studies have established robust associations between sleep quality in shift workers and factors such as cognition, stressors, mental states, and positive traits. However, the hierarchical relationships among these factors, such as proximal versus distal influences, and their mechanistic interactions in shaping sleep outcomes, remain unclear. In this study, we assessed 769 train drivers at baseline (T1), with 694 participants completing a follow-up sleep assessment 6 months later (T2). Using cross-sectional (T1) and longitudinal (T1-T2) network analyses, we mapped the interrelationships among these variables. Our findings indicate that mental states (e.g., anxiety, somatisation) serve as the most proximal predictors of sleep disturbances, while positive traits (e.g., mindfulness) function as intermediate factors. Cognition and external stressors emerged as the most distal influences. Both cross-sectional and longitudinal networks highlighted anxiety, somatisation, and sleep-related symptoms as key bridge nodes with high centrality. Notably, mindfulness exhibited strong bridging properties in the longitudinal analysis. These results suggest that mental states, particularly anxiety and somatisation, play a critical and immediate role in sleep dysfunction among train drivers. Interventions targeting mindfulness may offer a promising therapeutic avenue for improving sleep in this population.