A review and integration of models on delusion maintenance

Contemporary definitions of delusions highlight their resistance to conflicting evidence as the core feature. However, most etiological models of delusions have focused on delusion formation rather than maintenance and we lack a coherent understanding of why delusions persist. We conducted a systematic literature search of models on delusion maintenance, extracted their core postulates, point to explanatory gaps, and derive an integrated framework. We identified 74 published accounts that include postulated mechanisms of delusion maintenance. We classified the models into six core perspectives that informed them: Bayesian inference (17 models), associative learning theory (6 models), neurobiological (11 models), cognitive-behavioral (23 models), motivational (7 models), and social (6 models). Most models highlight a mechanistic role of avoidance and operant learning, converging on the idea that a delusional explanation is reinforced. Another repeatedly suggested mechanism is that the delusional belief, once formed, influences the way further information is processed. In addition, most models propose a key role of individual deficits and biases. The proposed factors can be combined in temporal progression, including early risk factors and resulting vulnerability, the common proposed mechanism of formation (i.e. search for explanation of ambiguous experiences), and the short- and long-term consequences of the delusional explanation along with feedback loops. By considering numerous factors and their interactions, the integrative model provides a considerably more compelling account of why delusions persist than any single perspective alone. It can help to identify novel directions for research and intervention, such as addressing short-term benefits of delusion maintenance.