Modelling the Longitudinal Dynamics of Paranoia in Psychosis: A Temporal Network Analysis Over 20 Years

Abstract

Paranoia is a highly debilitating, core element of psychosis, although is poorly managed. Theories of paranoia mostly interface with short-scale or cross-sectional data models, leaving the longitudinal course of paranoia underspecified. Here, we develop an empirical characterisation of two aspects of paranoia - persecutory and referential delusions - in individuals with psychosis over 20 years. We examine delusional dynamics by applying a Graphical Vector Autoregression Model to data collected from the Chicago Follow-up Study. We adjusted for age, sex, IQ, and antipsychotic use. We found that referential and persecutory delusions are central themes, supported by other primary delusions, and are strongly autoregressive. In a second analysis we demonstrate that social factors influence the severity of referential, but not persecutory, delusions. We suggest that persecutory delusions represent central, resistant states in the cognitive landscape, whereas referential beliefs are more flexible, offering an important window of opportunity for intervention. Our data models can be collated with prior biological, computational, and social work to contribute toward a more complete theory of paranoia and provide more time-dependent evidence for optimal treatment targets.

Longitudinal evolution of the transdiagnostic prodrome to severe mental disorders: a dynamic temporal network analysis informed by natural language processing and electronic health records

“…For each network (SMD network and three sub-networks), the following steps (pre-processing, network development and stability assessment) were repeated separately in each relevant dataset using a similar step-wise procedure to prior work modelling temporal features in psychopathology. 20…”

“…For each network (SMD network and three sub-networks), the following steps (pre-processing, network development and stability assessment) were repeated separately in each relevant dataset using a similar step-wise procedure to prior work modelling temporal features in psychopathology. 24 Pre-processing and network development methods are detailed in eMethods 4-5. Importantly, during pre-processing, we controlled for a set of demographic, medication and clinical variables through linear regression (see eMethods 4) and used the resultant residuals in our networks to isolate prodromal feature variability over time.…”

“…For each network (SMD network and three sub-networks), the following steps (pre-processing, network development and stability assessment) were repeated separately in each relevant dataset using a similar step-wise procedure to prior work modelling temporal features in psychopathology. 20 Pre-processing and network development methods are detailed in the Supplement (eMethods 3 and 4, respectively). For each network (SMD network and three sub-networks), we extracted the temporal (within individuals), contemporaneous (relationships between nodes averaged over time and averaged across the sample), and between-individuals subject matrices.…”