Abstracts

Rationale: Rationale: Brain dynamics in patients with epilepsy may be distinct from those in healthy persons and distinguishable on electroencephalography (EEG) even outside interictal epileptic discharges. When the diagnosis is unclear and EEG is without visually-apparent anomalies, a quantitative EEG predictor of epilepsy is desirable. Neuromodulatory and neuroimaging studies often find motor system differences between epilepsy patients and healthy controls. Simple motor tasks robustly drive beta frequency (14-28Hz) EEG power and may provide biomarkers for conditions from Parkinson’s disease to autism. Pharmacologic evidence suggests GABAergic inputs to an area are responsible for beta activity, and modeling indicates inhibitory signaling drives beta dynamics during motor events. We test whether beta “rebound,” an increase in power above baseline after suppression during a motor event, differs between healthy control participants and epilepsy patients.


Methods: Methods: Patients with epilepsy (n=31, ages 5-24) and healthy controls (n=13, ages 23-53) were exposed to 2-second tones and pushed a button after tone offset. Healthy control participants completed μ=366 trials (range=137-400). Epilepsy patients completed μ=164 (45-716) trials. 9 parasagittal and central EEG channels were bandpassed 1-55Hz, transformed by a complex Morlet wavelet (center freq.=5Hz, bandwidth freq.=1Hz), and averaged. The spectrographic region of interest (ROI) comprised a section containing the majority of pixels passing a permutation test of participant condition (2-sided, n=5000; Fig.1). To avoid noise, beta rebound was quantified as the ROI’s 80th percentile power. An electrode was deemed to show beta rebound if the power on button-press trials was greater than in pseudotrials drawn randomly from outside the task passing a permutation test of trial type (α=0.1, n=500).


Results: Results: Epilepsy patients had significantly lower beta rebound power than healthy controls (μ=0.06 vs. 0.30dB; KS=-0.63, p=5.6*10-4; Fig.2) and fewer channels with robust beta rebound (μ=1.26 vs. 4.69, U=-301, p=5.9*10-3). We replicated others’ finding of a correlation between beta rebound power and age among all participants (R=0.57, p=5.8*10-5). In the epilepsy cohort alone, however, there was no significant correlation with age (R=0.2, p=0.28), and adult epilepsy patients (n=8) had reduced beta rebound power compared with healthy controls (t=-2.97, p=7.8*10-3).


Conclusions: Conclusions: Patients with epilepsy have lower beta rebound power than healthy controls. The lack of a correlation with age within the epilepsy cohort suggests an altered maturational trajectory of the beta rebound. The difference in power in adults across conditions confirms that the beta rebound carries information about epileptic brain states. Future studies with age-matched controls will test whether beta rebound identifies epilepsy patients with different morbidities or responses to therapy.


Funding: Funding: NIGMS T32GM007753 and T32GM144273 to Harvard Medical School (DRM); Joseph J. Volpe Chair in Neurology Operating Fund within Boston Children’s Hospital (AR).