Abstracts

Rationale: Psychiatric comorbidities, such as depression and anxiety, affect approximately half of individuals with epilepsy, impacting treatment outcomes and quality of life. Such comorbidities could be explained by the effect the disease may have on cognitive processes such as emotion. Examining brain network deficits based on depression and anxiety work may help target treatments for epilepsy comorbidities. Here, we examine effective connectivity (EC) among relevant brain regions during an emotional conflict task in adolescents with focal or generalized epilepsy compared to controls. We hypothesize that adolescents with epilepsy have lower EC within the emotion and cognitive processing networks, which differ between epilepsy type.


Methods: Fifty-three adolescents (28 epilepsy [14 focal, 14 generalized] and 25 controls; 10-19y) underwent magnetoencephalography recordings while completing the emotional Face-Word Stroop Task (Fig. 1A). Accuracy and response times were computed and compared between conditions and groups. Participants also completed self-report measures of anxiety (GAD-7) and depression (PHQ-9) which were compared between groups (Fig. 1B-C). Contrast (incongruent–congruent) source activity maps time-locked to stimulus presentation (-500—1000ms) were computed using realistic head models and dynamic statistical parametric mapping (Fig. 1D-E). Time-series from 37 regions-of-interest (ROIs) related to depression and emotional conflict processing were extracted and EC values were computed on ten subsequent 100-ms epochs. Average inward (to) and outward (from) EC values for each ROI were computed and compared between groups for each epoch using logistic regression models, followed by post-hoc t-tests (Bonferroni corr. p< .05, Fig. 1F)


Results: While controls exhibited normative behavioral patterns, responding slower and less accurately to emotional conflicting stimuli, those with epilepsy did not show this pattern except for the focal group with their response times. Unexpectedly, anxiety and depression scores did not significantly differ between groups. Relative to controls, adolescents with generalized epilepsy had lower inward EC across all epochs and across all included networks, but predominantly in the Fusiform Gyrus and in ROIs of the Salience and Somatosensory networks (all p< .05). Conversely, those with focal epilepsy showed lower inward and outward EC than controls in the first 600 ms post-stimulus, predominantly in the Fusiform Gyrus, and ROIs of the Default Mode and Salience networks (all p< .05). Those with generalized epilepsy had lower inward EC than those with focal epilepsy, predominantly in ROIs in the Salience network (p< .05). See Fig. 2.


Conclusions: Relatively lower inward and outward EC in adolescents with epilepsy alongside behavioral differences during an emotional conflict task may suggest deficient network connections relevant for emotion identification and regulatory processes. Further, connectivity differences between focal and generalized epilepsy may represent network deficits unique to each epilepsy subtype. These network deficits may be valuable targets for treatments of epilepsy, its subtypes, and its psychiatric comorbidities.


Funding: JEH Foundation