Lithium monotherapy associated clinical improvement effects on amygdala-ventromedial prefrontal cortex resting state connectivity in bipolar disorder

Abstract

Background: This study, for the first time, investigated lithium monotherapy associated effects on amygdala- ventromedial prefrontal cortex (vMPFC) resting-state functional connectivity and correlation with clinical improvement in bipolar disorder (BP) METHODS: Thirty-six medication-free subjects - 24 BP (12 hypomanic BPM) and 12 depressed (BPD)) and 12 closely matched healthy controls (HC), were included. BP subjects were treated with lithium and scanned at baseline, after 2 weeks and 8 weeks. HC were scanned at same time points but were not treated. The effect of lithium was studied for the BP group as a whole using two way (group, time) ANOVA while regressing out effects of state. Next, correlation between changes in amygdala-vMPFC resting-state connectivity and clinical global impression (CGI) of severity and improvement scale scores for overall BP illness was calculated. An exploratory analysis was also conducted for the BPD and BPM subgroups separately.

Results: Group by time interaction revealed that lithium monotherapy in patients was associated with increase in amygdala-medial OFC connectivity after 8 weeks of treatment (p = 0.05 (cluster-wise corrected)) compared to repeat testing in healthy controls. Increased amygdala-vMPFC connectivity correlated with clinical improvement at week 2 and week 8 as measured with the CGI-I scale.

Limitations: The results pertain to open-label treatment and do not account for non-treatment related improvement effects. Only functional connectivity was measured which does not give information regarding one regions effect on the other.

Conclusions: Lithium monotherapy in BP is associated with modulation of amygdala-vMPFC connectivity which correlates with state-independent global clinical improvement.

Fig. 1

Group X time effect between left amygdala and left & right rectus gyrus connectivity in BP & HC between baseline and week-8. The statistical significance was set at p = 0.001 and cluster size = 28, which corresponded to a cluster-wise corrected p = 0.05.

Fig. 2

Group X time effect between right amygdala and right superior OFC connectivity in BP & HC between baseline and week-8. The statistical significance was set at p = 0.001 and cluster size = 28, which corresponded to a cluster-wise corrected p = 0.05.

Fig. 3

Change in connectivity between left amygdala and left superior OFC in BP (Week 2 – Baseline) significantly correlates with Week-2 CGI-I ratings. The statistical significance was set at z > 3 and cluster size = 28, which corresponded to a cluster-wise corrected p = 0.05.

Fig. 4

Change in connectivity between right amygdala and left & right subgACC extending into medial OFC in BP (Week 8 – Baseline) significantly correlates with week-8 CGII ratings, the statistical significance was set at z > 3 and cluster size = 28, which corresponded to a cluster-wise corrected p = 0.05.