Neuronal Correlates of Social Cognition in BPD
Neuronal Correlates of Social Cognition in BPD
There was no main effect of group or interaction by group and condition, neither for performance nor for reaction times (all ps > 0.05; Supplementary FigureS1). Correlation analyses revealed a significant correlation between affective ToM and emotion recognition performance in the healthy controls (r = 0.618; P = 0.024), replicating earlier findings (Mier et al., 2010a,b), and on a trend level in the BPD patients (r = 0.506; P = 0.078), too.
Main Effect of Group HCs showed a significantly stronger activation of the right inferior prefrontal gyrus and thalamus, and in the left visual association cortex and cerebellum compared to BPD patients ( Table 2 ). ROI-analyses revealed significantly higher activation in the HCs than in the BPD patients for BA 44 bilaterally (left: peak voxel −57 12 6, cluster size 19, T = 3.37, P = 0.041, small volume corrected; right: peak voxel 48 12 15, cluster size 45, T = 3.60, P = 0.019, small volume corrected).
BPD patients, on the other hand, showed stronger activations than the HCs in the left amygdala and somatosensory, and primary motor cortex, and in the right visual association cortex ( Table 2 ; Figures 2 and 3A). ROI-analyses confirmed the hyperactivation of the left amygdala in the BPD patients (peak voxel: −33 −3 −18, cluster size = 8, T = 3.45, P = 0.019, small volume corrected).
(Enlarge Image)
Figure 2.
Main effect of group. Displayed is the significantly increased left amygdala activation in the borderline patients compared to the healthy controls. Threshold for display: P < 0.005 uncorrected.
(Enlarge Image)
Figure 3.
Mean signal change for both groups in the three conditions. (A) Signal change in the left amygdala. (B) Signal change in the right BA 44. (C) Signal change in the right STS. Note: BP = Borderline Patients, HC = Healthy Controls, ToM = affective ToM, Emo = Emotion recognition, Neutral = Neutral facial processing.
Group × Condition Interactions Whole-brain analysis revealed no regions showing a significant interaction between groups and task condition. However, a group difference that was modulated by the to be solved task could be observed in ROI-analyses for the left BA 44 (peak voxel −51 15 9, cluster size 32, T = 3.57, P = 0.025, small volume corrected) and left STS (peak voxel −51 60 21, cluster size 47, T = 3.64, P = 0.027, small volume corrected), resulting from increasing activation from the neutral task to the affective ToM task in the healthy control group, but not in the BPD group (Figure 3B and C).
We additionally tested for group differences at the single social cognitive stages. These results are reported in the Supplementary Materials (Supplementary Text and Tables S5–S8).
Relationship to Depressive Symptomatology To explore potential influences of the comorbid depression in the BPD-group, simple regression analyses between BDI sum score and brain activation in the ROIs were conducted, using SPM. In none of the tasks a significant relationship was found.
Results
Behavioral Results
There was no main effect of group or interaction by group and condition, neither for performance nor for reaction times (all ps > 0.05; Supplementary FigureS1). Correlation analyses revealed a significant correlation between affective ToM and emotion recognition performance in the healthy controls (r = 0.618; P = 0.024), replicating earlier findings (Mier et al., 2010a,b), and on a trend level in the BPD patients (r = 0.506; P = 0.078), too.
Functional Brain-imaging Data
Main Effect of Group HCs showed a significantly stronger activation of the right inferior prefrontal gyrus and thalamus, and in the left visual association cortex and cerebellum compared to BPD patients ( Table 2 ). ROI-analyses revealed significantly higher activation in the HCs than in the BPD patients for BA 44 bilaterally (left: peak voxel −57 12 6, cluster size 19, T = 3.37, P = 0.041, small volume corrected; right: peak voxel 48 12 15, cluster size 45, T = 3.60, P = 0.019, small volume corrected).
BPD patients, on the other hand, showed stronger activations than the HCs in the left amygdala and somatosensory, and primary motor cortex, and in the right visual association cortex ( Table 2 ; Figures 2 and 3A). ROI-analyses confirmed the hyperactivation of the left amygdala in the BPD patients (peak voxel: −33 −3 −18, cluster size = 8, T = 3.45, P = 0.019, small volume corrected).
(Enlarge Image)
Figure 2.
Main effect of group. Displayed is the significantly increased left amygdala activation in the borderline patients compared to the healthy controls. Threshold for display: P < 0.005 uncorrected.
(Enlarge Image)
Figure 3.
Mean signal change for both groups in the three conditions. (A) Signal change in the left amygdala. (B) Signal change in the right BA 44. (C) Signal change in the right STS. Note: BP = Borderline Patients, HC = Healthy Controls, ToM = affective ToM, Emo = Emotion recognition, Neutral = Neutral facial processing.
Group × Condition Interactions Whole-brain analysis revealed no regions showing a significant interaction between groups and task condition. However, a group difference that was modulated by the to be solved task could be observed in ROI-analyses for the left BA 44 (peak voxel −51 15 9, cluster size 32, T = 3.57, P = 0.025, small volume corrected) and left STS (peak voxel −51 60 21, cluster size 47, T = 3.64, P = 0.027, small volume corrected), resulting from increasing activation from the neutral task to the affective ToM task in the healthy control group, but not in the BPD group (Figure 3B and C).
We additionally tested for group differences at the single social cognitive stages. These results are reported in the Supplementary Materials (Supplementary Text and Tables S5–S8).
Relationship to Depressive Symptomatology To explore potential influences of the comorbid depression in the BPD-group, simple regression analyses between BDI sum score and brain activation in the ROIs were conducted, using SPM. In none of the tasks a significant relationship was found.
Source...