Mental disorders, network models, and dynamical systems

Denny Borsboom

DOI: https://doi.org/10.1093/med/9780198796022.003.0011

2017-04-01

Abstract:Network approaches have been proposed as an alternative way of thinking about relations between symptoms of mental disorders. Unlike traditional psychometric approaches, network models view these associations as the result of direct interactions between symptoms. Disorders are defined as alternative stable states of a network due to increased connectivity between symptoms. This increased connectivity creates a pattern of reinforcement, so the system can get stuck in a state of prolonged activation. Mental health is defined as the stable state of a weakly connected network. Although symptomatology may be temporarily increased in a healthy network (e.g., due to adverse life events), as the influence of a shock wanes the network will spontaneously return to its healthy state. Strongly connected networks, however, may transition into disordered states upon similar external shocks, and may not naturally recover. Thus, the proposed definitions yield plausible conceptualizations of resilience and vulnerability.