Tracing burnout in motion: a temporal network perspective on job demands and job resources

Background: Burnout remains a pressing workplace issue with far-reaching consequences for employees, organizations, and society at large. While extensive research has established the importance of job demands and job resources in shaping burnout, most studies conceptualize these factors as static or unidirectional. As a result, we know relatively little about how job demands, job resources, and burnout symptoms interact dynamically over time. This gap limits our ability to understand how burnout unfolds in real-world settings, where work characteristics and psychological states fluctuate continuously. To address this, we adopt a dynamic systems perspective that captures the temporal interplay between job characteristics and burnout dimensions.

Methods: We collected weekly self-report data from over 250 employees across five consecutive weeks, focusing on key job demands (e.g., workload), job resources (e.g., social support), and burnout (e.g., emotional exhaustion). To model the temporal and contemporaneous relationships among these variables, we employed Gaussian Graphical Models (GGMs) and Multilevel Vector Autoregression (mlVAR). This analytic approach allows us to examine both within-person fluctuations and between-person differences, while identifying potential feedback loops and time-lagged effects.

Result: Our analyses reveal that burnout is not a static condition triggered by isolated job demands and/or job resources, but rather a dynamic phenomenon emerging from a shifting network of job demands, job resources, and burnout symptoms. Emotional exhaustion, in particular, surfaced as a central node in both contemporaneous and temporal networks. Rather than being solely driven by high workload, exhaustion appears to be shaped by complex, reciprocal patterns of strain and support. For example, a temporary dip in some job resources (e.g., less supervisor support) predicted increases in exhaustion the following week, which in turn dampened perceptions of social support. These findings suggest that burnout symptoms can both result from and contribute to changes in the work environment, reinforcing their persistence over time.

Conclusion: By capturing burnout as a dynamic process, this study offers a more nuanced and ecologically valid understanding of how it develops and persists. Our findings challenge linear stress models and highlight the importance of timing in occupational health research. From a practical standpoint, the identification of central nodes and time-lagged effects opens up new possibilities for targeted, timely interventions. Rather than focusing solely on reducing job demands or increasing job resources in general, organizations may benefit from monitoring dynamic patterns and intervening at critical moments when employees are most vulnerable to cascading effects. This dynamic systems approach thus holds promise for both theory and practice in the prevention and management of burnout.