Mapping mental health complexity in esports: an exploratory psychosocial network approach

The rapid rise in popularity of esports has brought increased attention to the unique mental health challenges faced by esports players, yet research in this area remains in its infancy. Early research suggests a substantial mental health burden and has examined associations between esports-specific stressors, mental health, and sleep disturbances. However, much of this work has relied on traditional methodologies, providing limited insight into how affective, psychosocial, sleep-related, and esports-specific indicators are organised within this competitive performance environment.

To address this gap, we developed a bespoke 17-item questionnaire tailored for Directed Acyclic Graph (DAG) network analysis, an exploratory approach for mapping the conditional dependency structure of complex multivariate systems such as mental health.

A total of 1025 esports players from over 60 countries completed the survey. Game-related thought preoccupation, a core feature of gaming disorder, emerged as the most robust structural sink node, downstream of several psychosocial indicators, raising the hypothesis that it represents a compound indicator of broader upstream psychosocial strain. Depressed mood occupied an upstream position within the network, an observation that held more robustly for the psychopathological subnetwork as a whole than for depressed mood as the single source. Mental fatigue occupied an intermediate, connecting position between affective variables and the sleep and fatigue indicators. Its high prevalence, together with a positioning upstream of the sleep indicators, suggests that it may warrant closer examination in future research on training load, recovery, and sleep in esports, though this ordering rested on edges with moderate directional support. Emotion control and a sense of autonomy appeared to connect upstream affective distress with downstream esports-specific indicators such as esports-related stress, though was the least stable of the structural features identified.

Collectively, these findings provide an exploratory map of how mental health-relevant indicators were organised in this performance environment, and generate concrete, testable hypotheses for future longitudinal, experimental, and intervention research.