3.4 CHANGES IN AMYGDALA AND HIPPOCAMPAL FUNCTIONAL CONNECTIVITY IN SUBCLINICAL PSYCHOSIS: RELATIONSHIP TO SYMPTOM PERSISTENCE, PARANOIA AND ABERRANT SALIENCE

Prior studies have found evidence for abnormalities of medial temporal lobe structures in patients with psychotic disorders. It has been hypothesized that these abnormalities, which have been linked to dysregulation of dopaminergic neurotransmission in animal models of psychosis, may be associated with a cognitive bias to mislabel neutral events or objects in the environment as salient, which may, in turn, give rise to delusional beliefs and other psychotic symptoms. One persistent challenge for this line of research arises from the confounds associated with testing such models in psychotic disorder patients who have been treated with antipsychotic medication. In light of recent evidence for a neurobiological continuum between clinical and subclinical psychosis, in the current study, medial temporal lobe connectivity was measured in non-help-seeking young people (college students) with a range of severity of subclinical delusional beliefs (n = 131). We sought to determine whether changes in medial temporal lobe connectivity are associated with such beliefs, in particular persecutory ideas, as well as with a bias to mislabel neutral information as negatively-valenced. We also sought to better understand the phenomenology and antecedents of delusional beliefs in this young adult population, by examining relationships to other subthreshold psychotic experiences and childhood adversity in a larger cohort of subjects (n = 399). Participants with a range of severity of delusional beliefs were identified and recruited from college campuses in the Boston area. Associations among symptoms and childhood trauma were examined using hierarchical regression modeling. Resting-state functional magnetic resonance imaging data were collected in a subset of these subjects, and seed-based functional connectivity analyses were conducted using atlas-based amygdala and hippocampus seeds. Also, biases to mislabel neutral information were measured using a validated word classification task. Lastly, a portion of the cohort was followed longitudinally, via on-line assessments. A history of childhood trauma was highly associated with both delusional beliefs and hallucinatory experiences, which were highly correlated with each other. The severity of delusional beliefs in this cohort strongly correlated with the strength of the connectivity of the amygdala to early visual cortical areas. This effect remained significant after accounting for symptoms of depression, anxiety and hallucinatory experiences. Moreover, the effect was stronger when the analyses were limited to those participants with delusional beliefs that had persisted over a period of one year. Further analyses revealed that these effects were mainly accounted for by the presence of persecutory beliefs, rather than other delusion types. The hippocampal seed showed a similar, but less robust, pattern of effects. Lastly, a significant association between the tendency to mislabel neutral information as negatively-valenced and delusional belief severity was mediated by the strength of amygdala-visual cortex connectivity. Subclinical delusional beliefs, particularly persecutory ideas, are associated with a history of childhood trauma and increased connectivity between the amygdala and visual cortex, which was linked to a tendency to mislabel neutral information as negatively-valenced. These findings suggest a mechanistic path by which early stress-induced abnormalities in medial temporal lobe function could give rise to misperceptions and misinterpretations of incoming sensory stimuli.