Demoralization and depression in breast cancer: A Bayesian network and computer-simulated intervention study

Demoralization frequently co-occurs with depression among breast cancer survivors, yet the symptom-level interactions and causal pathways remain poorly understood. We aimed to map these interrelationships and pinpoint the most effective targets for intervention.

We enrolled 413 breast cancer survivors (42.9 % stage II) and assessed demoralization with the Demoralization Scale II and depressive symptoms with the Depression subscale of the Hospital Anxiety and Depression Scale. A Gaussian graphical model was used to identify core symptoms and bridge nodes linking the two syndromes. We then constructed a Bayesian network to infer potential causal directions. Finally, each symptom was modulated in computer simulation to evaluate its impact on overall network connectivity.

In the Gaussian graphical model, the "optimistic" symptom exhibited the highest bridge centrality (bridge strength = 0.338), serving as the primary conduit between demoralization and depression clusters. The Bayesian network positioned "distressed" as a parent node, initiating a directional cascade from demoralization to depression. Simulation experiments revealed that reducing "distressed" yielded the greatest overall decrease in global network strength (from 4.93 to 2.20).

Demoralization may precipitate depression via key symptom interactions, particularly distress and optimism. Interventions that prioritize alleviating distress and fostering optimism are likely to produce the most efficient dual benefit, mitigating both demoralization and depression in breast cancer patients.