Global prefrontal and fronto-amygdala dysconnectivity in bipolar I disorder with psychosis history

Abstract

Background: Pathophysiological models of bipolar disorder postulate that mood dysregulation arises from fronto-limbic dysfunction, marked by reduced prefrontal cortex (PFC) inhibitory control. This might occur due to both disruptions within PFC networks and abnormal inhibition over subcortical structures involved in emotional processing. However, no study has examined global PFC dysconnectivity in bipolar disorder and tested whether regions with within-PFC dysconnectivity also exhibit fronto-limbic connectivity deficits. Furthermore, no study has investigated whether such connectivity disruptions differ for bipolar patients with psychosis history, who might exhibit a more severe clinical course.

Methods: We collected resting-state functional magnetic resonance imaging at 3T in 68 remitted bipolar I patients (34 with psychosis history) and 51 demographically matched healthy participants. We employed a recently developed global brain connectivity method, restricted to PFC (rGBC). We also independently tested connectivity between anatomically defined amygdala and PFC.

Results: Bipolar patients exhibited reduced medial prefrontal cortex (mPFC) rGBC, increased amygdala-mPFC connectivity, and reduced connectivity between amygdala and dorsolateral PFC. All effects were driven by psychosis history. Moreover, the magnitude of observed effects was significantly associated with lifetime psychotic symptom severity.

Conclusions: This convergence between rGBC, seed-based amygdala findings, and symptom severity analyses highlights that mPFC, a core emotion regulation region, exhibits both within-PFC dysconnectivity and connectivity abnormalities with limbic structures in bipolar illness. Furthermore, lateral PFC dysconnectivity in patients with psychosis history converges with published work in schizophrenia, indicating possible shared risk factors. Observed dysconnectivity in remitted patients suggests a bipolar trait characteristic and might constitute a risk factor for phasic features of the disorder.

Figure 1. Global Prefrontal Dysconnectivity

(a) Significant between-group differences in prefrontal rGBC between bipolar patients and healthy participants revealed a medial prefrontal cortex region (MPFC) (x=3, y=32, z=1). The red border approximately marks the restricted PFC analysis. (b) rGBC values are shown for the mPFC region across the three groups; healthy participants (white) and bipolar patients without psychosis history (BPW, gray); bipolar patients with history of psychosis (BPP, black). Error bars represent +/− 1 standard error of the mean.

Figure 2. Amygdala Prefrontal Dysconnectivity

Significant group differences in amygdala-prefrontal fcMRI between bipolar disorder subgroups and healthy controls. (a) Yellow/red foci mark regions where bipolar patients with a history of psychosis showed increased amygdala connectivity relative to non-psychotic patients and healthy controls. This pattern was centered on the medial prefrontal cortex (mPFC) (x=1, y=41, z=−3). The red border approximately marks the restricted PFC analysis. (b) A right DLPFC region (x=34, y=43, z=30) is shown in blue for which bipolar patients with a history of psychosis showed decreased amygdala connectivity relative to non-psychotic patients and healthy controls. The rGBC values are shown across both foci for controls (white), bipolar patients without a history of psychosis (BPW, grey) and bipolar patients with history of psychosis (BPP, black). Error bars represent +/− 1 standard error of the mean.

Figure 3. rGBC, Amygdala-Prefrontal Dysconnectivity and Lifetime Psychotic Symptom Severity

(a) Trend-level inverse relationship between mPFC rGBC and lifetime positive psychotic symptom severity across the entire sample of bipolar patients (rho= −.22, p=.07). (b) Significant positive relationship between amygdala-mPFC fcMRI and lifetime positive psychotic symptom severity across the entire sample of bipolar patients (rho=.31, p<.015). (c) Significant inverse relationship between amygdala-DLPFC fcMRI and lifetime positive psychotic symptom severity across the entire sample of bipolar patients (rho= −.44, p<.0001). Direction of all reported individual difference effects show strong convergence with main effects. The scale on the x-axis captures a clinician-rated severity index that ranges from 0 (absent) to 4 (Very severe; Gross or nearly constant effect on function) (67).