Altered Brain Activation during Emotional Face Processing in Relation to Both Diagnosis and Polygenic Risk of Bipolar Disorder

Abstract

ObjectivesBipolar disorder (BD) is a highly heritable disorder with polygenic inheritance. Among the most consistent findings from functional magnetic imaging (fMRI) studies are limbic hyperactivation and dorsal hypoactivation. However, the relation between reported brain functional abnormalities and underlying genetic risk remains elusive. This is the first cross-sectional study applying a whole-brain explorative approach to investigate potential influence of BD case-control status and polygenic risk on brain activation.MethodsA BD polygenic risk score (PGRS) was estimated from the Psychiatric Genomics Consortium BD case-control study, and assigned to each individual in our independent sample (N=85 BD cases and 121 healthy controls (HC)), all of whom participated in an fMRI emotional faces matching paradigm. Potential differences in BOLD response across diagnostic groups were explored at whole-brain level in addition to amygdala as a region of interest. Putative effects of BD PGRS on brain activation were also investigated.ResultsAt whole-brain level, BD cases presented with significantly lower cuneus/precuneus activation than HC during negative face processing (Z-threshold=2.3 as cluster-level correction). The PGRS was associated positively with increased right inferior frontal gyrus (rIFG) activation during negative face processing. For amygdala activation, there were no correlations with diagnostic status or PGRS.ConclusionsThese findings are in line with previous reports of reduced precuneus and altered rIFG activation in BD. While these results demonstrate the ability of PGRS to reveal underlying genetic risk of altered brain activation in BD, the lack of convergence of effects at diagnostic and PGRS level suggests that this relation is a complex one.

The polygenic risk for bipolar disorder influences brain regional function relating to visual and default state processing of emotional information

“…Relatives had intermediate values that were not statistically different from that of either patients or controls. In line with previous reports (Whalley et al, 2012, Tesli et al, 2015), there was a significant effect of PGR-BD on task-related BOLD signal change independent of group but no significant PGR-BD score by group interactions. The current findings therefore support the notion that the polygenic load method is more informative in terms of identifying patterns of neural activation that mediate vulnerability to BD rather than symptom expression.…”

“…The PGR score for BD (PGR-BD) is calculated in each individual by aggregating variation across GWAS loci nominally associated with BD into a quantitative score (International Schizophrenia Consortium et al, 2009, Dima and Breen, 2015), with the current PGR-BD explaining ~ 5% of the variance to BD (Cross-Disorder Group of the Psychiatric Genomics Consortium, 2013). Two previous studies have used the polygenic risk score method to examine the cumulative impact of BD-related risk-conferring SNPs on task-related brain activation during an emotional processing (Tesli et al, 2015) and a word generation task (Whalley et al, 2012). During emotional processing, a positive association was found in patients with BD and in controls between PRG-BD score and brain activation in the right vlPFC.…”

Correlation between Intrinsic Brain Activity and Thyroid-Stimulating Hormone Level in Unmedicated Bipolar II Depression

“…Meda et al[32] found that when compared to slow-5 ALFF, psychotic BD patients showed reduced regional power in the PCC in slow-4 band. Several task-based fMRI studies indicated reduced activation in the PCu/PCC in BD patients during the self-reflection task [33], emotional conflicts, and emotional faces matching tasks [34-36]. Structural MRI studies in BD patients revealed reduced gray matter and cortical thickness in the PCC [37-39].…”

Contribution of oxytocin receptor polymorphisms to amygdala activation in schizophrenia spectrum disorders

“…A widely used and validated paradigm was employed to elicit amygdala activation. [24][25][26][27] In this task, participants selected which of two stimuli (displayed at the bottom of the screen) matched a target stimulus (displayed at the top). In the faces-matching task, the images displayed were human faces expressing either anger or fear ('negative faces'), or happiness ('positive faces').…”

“…In the current study, we focused on participants' amygdala responses to negative stimuli (the negative faces>shapes contrast), because of the implication of altered threat-related facial affect perception in psychotic disorders 33,34 and since this contrast has been reported to reliably engage the amygdala. 27,35,36 Amygdala regions of interest (ROIs) were defined in accordance with the probabilistic Harvard-Oxford subcortical atlas provided with FSL, and were thresholded at 25% probability. Mean BOLD signal changes (parameter estimates) across all ROI voxels were extracted from FSL and used in statistical analyses.…”