Idiopathic generalized epilepsies are epileptic syndromes characterized by seizures involving widespread networks without a localized seizure onset on neuroimaging and a presumed genetic etiology. They account for approximately 20% of all epilepsies, according to the study’s background, but remain underrepresented in research. Unlike focal epilepsy, which is associated with discrete seizure foci, idiopathic generalized epilepsies are linked to dysfunction in thalamocortical networks. The thalamus and anterior cingulate cortex, or ACC, are central components of this circuitry and participate in functions including cognitive control, attention and error detection, making them important regions for investigating physiological features of idiopathic generalized epilepsies.

The study was conducted at the University of Alabama at Birmingham using baseline data from two related projects: a blinded randomized controlled trial of combined endurance and resistance exercise training in patients with idiopathic generalized epilepsies and an endurance-training pilot study in healthy controls. The analysis did not test whether exercise changed brain measures. Adults with a clinician-confirmed diagnosis of idiopathic generalized epilepsy and healthy controls underwent the same general procedures, including a physical examination, maximal graded exercise test and magnetic resonance imaging on the same 3T scanner. Most patients were taking one or more antiseizure medications, with leviracetam, lamotrigine and sodium valproate among the most frequently prescribed drugs. This medication heterogeneity was not controlled statistically and could influence exercise capacity, cerebral perfusion or metabolite measurements.

During the cycling test, work began at 25 watts and increased by 25 watts every two minutes while investigators monitored respiratory exchange, blood pressure, heart rate and perceived exertion. Peak oxygen uptake, calculated from the final two 30-second measurements, served as the primary measure of cardiorespiratory fitness. Two participants did not complete the test, leaving 44 people in the fitness analysis: 14 healthy controls, 16 seizure-free patients and 14 patients with active seizures. After adjustment for age and sex, estimated mean peak oxygen uptake was 20.41 millilitres per kilogram per minute in the idiopathic generalized epilepsy group and 22.85 millilitres per kilogram per minute in controls. The difference was not statistically significant, with a p-value of 0.203 and a partial eta-squared of 0.04. Seizure-control status also produced no significant overall difference in fitness, although the estimated mean was lowest among participants with active seizures, at 18.72 millilitres per kilogram per minute, compared with 21.85 among seizure-free patients and 22.88 in controls.

That null result should be interpreted in light of how the comparison group was selected. Both patients and healthy controls had low leisure-time physical activity, with the study requiring activity below the American Heart Association recommendation of 150 minutes per week of moderate-intensity endurance exercise. Their self-reported activity did not significantly differ: estimated marginal means were 2.31 in controls and 2.18 in the idiopathic generalized epilepsy group, with a p-value of 0.696. No objective activity tracker was used. The authors said this similarity may have reduced variation in fitness, potentially masking differences associated with the disorder, although the cross-sectional design cannot establish that explanation. Antiseizure medications may also have influenced exercise capacity. The result therefore indicates no detectable difference in this selected sample, not that cardiorespiratory fitness is unaffected in all people with idiopathic generalized epilepsy.

For cerebral blood flow, researchers used a five-delay pseudo-continuous arterial spin-labeling sequence with post-labeling delays of 500, 1,000, 1,500, 2,000 and 2,500 milliseconds. A multi-delay kinetic model estimated blood flow and arterial transit time, while calibration produced absolute perfusion values in millilitres per 100 grams of tissue per minute. Investigators extracted atlas-based regional values from the ACC and thalamus. Eight participants were excluded because of ghosting artifacts or perfusion variance above the predefined threshold, leaving 38 people for the comparison. After adjustment for age and sex, the idiopathic generalized epilepsy group had significantly higher ACC blood flow than controls: estimated means were 59.61 and 52.72 millilitres per 100 grams per minute, respectively. The difference was 6.89 millilitres per 100 grams per minute, with a p-value of 0.039 and a partial eta-squared of 0.12. Thalamic blood flow did not differ significantly between groups, and seizure-control status was not associated with a significant perfusion difference in either region.

The researchers also used proton magnetic resonance spectroscopy to measure concentrations of N-acetylaspartate, or NAA, and the combined glutamate-glutamine signal, Glx, relative to total creatine in the ACC and thalamus. NAA is considered a marker of neuronal integrity, while Glx reflects aspects of excitatory neurotransmission and cognitive processing. Point-resolved spectroscopy used separate voxels over the bilateral ACC and thalamus, and the resulting spectra were corrected for tissue composition. Seven participants were excluded under predefined spectral-quality criteria, leaving 39 participants in the metabolite analysis. In the thalamus, NAA relative to total creatine was lower in patients with idiopathic generalized epilepsies than in controls, with estimated means of 1.35 and 1.49, respectively. The difference was 0.14, with a p-value of 0.012 and a partial eta-squared of 0.16. Neither ACC metabolite ratio nor the overall thalamic Glx ratio differed significantly between the main study groups.

Seizure-control status was associated with specific thalamic differences. Patients with active seizures had a lower thalamic NAA-to-total-creatine ratio than healthy controls, with an estimated mean of 1.29 compared with 1.49 in controls and a post-hoc p-value of 0.017. Patients who had been seizure-free for the preceding six months did not differ significantly from controls in post-hoc comparisons. The active-seizure group also had a lower thalamic Glx-to-total-creatine ratio than the seizure-free group, with estimated means of 0.57 and 0.81 and a difference of 0.24. The post-hoc p-value was 0.004, while the overall seizure-status effect had a p-value of 0.005 and a partial eta-squared of 0.26. The authors described this pattern as consistent with altered thalamocortical network physiology, but the measurement is not a direct readout of extracellular glutamate or acute seizure activity. Single-voxel spectroscopy has limited spatial and temporal resolution, cannot separate glutamate from glutamine, and could not detect GABA peaks. Medication effects are also plausible because the participants’ drug regimens varied substantially, including both monotherapy and polytherapy.

Several limitations constrain how far the results can be generalized. The sample was relatively small, particularly for the spectroscopy subgroup comparisons, and the investigators used atlas-based regions that averaged across distinct thalamic nuclei and ACC subregions. Such averaging may obscure localized perfusion effects and could contribute to the absence of a significant thalamic blood-flow difference. Seizure control was reduced to a binary classification based on whether any seizures had occurred during six months, without incorporating seizure frequency, severity or type, and the investigators noted that definitions of seizure freedom vary across epilepsy research. Medication effects were not analyzed directly, physical activity was not measured objectively, and the deliberately inactive healthy controls could not establish whether more active people would show different results. The findings are observational associations from a selected cohort, not evidence that altered blood flow or metabolite ratios cause seizures, that exercise can correct them, or that these measurements can predict clinical outcomes. Nevertheless, the combined measurements show that idiopathic generalized epilepsies are associated in this sample with region-specific differences in thalamocortical perfusion and metabolism, while the role of cardiorespiratory fitness remains unresolved.

Subject of Research: Cardiorespiratory fitness, cerebral perfusion and thalamic metabolism in adults with idiopathic generalized epilepsies

Article Title: Investigating cardiorespiratory fitness, cerebral blood flow, and brain metabolism in epilepsy

Article References: Lien, Y., Mitchell, B. S., Szaflarski, J. P., Ver Hoef, L., Buford, T. W., Goodman, A. M., Willoughby, W. R., Morgan, C. J., Moyana, A., Pilkington, J., August, A. R., Bennett, J. T., & Allendorfer, J. B. (2026). Investigating cardiorespiratory fitness, cerebral blood flow, and brain metabolism in epilepsy. Frontiers in Network Physiology, 6, Article 1894335. https://doi.org/10.3389/fnetp.2026.1894335

Image Credits: AI Generated

DOI: 10.3389/fnetp.2026.1894335

Keywords: idiopathic generalized epilepsies, cerebral blood flow, brain metabolism, cardiorespiratory fitness, thalamocortical network, magnetic resonance spectroscopy, arterial spin labeling, seizure control, anterior cingulate cortex, thalamic metabolism, Investigating, cardiorespiratory