The Potential Role of Previous Physical Exercise Program to Reduce Seizure Susceptibility: A Systematic Review and Meta-Analysis of Animal Studies

Background: Clinical and pre-clinical studies indicate a reduction in seizure frequency as well as a decrease in susceptibility to subsequently evoked seizures after physical exercise programs. In contrast to the influence of exercise after epilepsy previously established, various studies have been conducted attempting to investigate whether physical activity reduces brain susceptibility to seizures or prevents epilepsy. We report a systematic review and meta-analysis of different animal models that addressed the impact of previous physical exercise programs to reduce seizure susceptibility. Methods: We included animal model (rats and mice) studies before brain insult that reported physical exercise programs compared with other interventions (sham, control, or naïve). We excluded studies that investigated animal models after brain insult, associated with supplement nutrition or drugs, that did not address epilepsy or seizure susceptibility, ex vivo studies, in vitro studies, studies in humans, or in silico studies. Electronic searches were performed in the MEDLINE (PubMed), Web of Science (WOS), Scopus, PsycINFO, Scientific Electronic Library Online (SciELO) databases, and gray literature, without restrictions to the year or language of publication. We used SYRCLE's risk of bias tool and CAMARADES checklist for study quality. We performed a synthesis of results for different types of exercise and susceptibility to seizures by random-effects meta-analysis. Results: Fifteen studies were included in the final analysis (543 animals), 13 of them used male animals, and Wistar rats were the most commonly studied species used in the studies (355 animals). The chemoconvulsants used in the selected studies were pentylenetetrazol, penicillin, kainic acid, pilocarpine, and homocysteine. We assessed the impact of study design characteristics and the reporting of mitigations to reduce the risk of bias. We calculated a standardized mean difference effect size for each comparison and performed a random-effects meta-analysis. The meta-analysis included behavioral analysis (latency to seizure onset, n = 6 and intensity of motor signals, n = 3) and electrophysiological analysis (spikes/min, n = 4, and amplitude, n = 6). The overall effect size observed in physical exercise compared to controls for latency to seizure onset was −130.98 [95% CI: −203.47, −58.49] (seconds) and the intensity of motor signals was −0.40 [95% CI: −1.19, 0.40] (on a scale from 0 to 5). The largest effects were observed in electrophysiological analysis for spikes/min with −26.96 [95% CI: −39.56, −14.36], and for spike amplitude (μV) with −282.64 [95% CI: −466.81, −98.47]. Discussion: Limitations of evidence. A higher number of animal models should be employed for analyzing the influence of exerciseon seizure susceptibility. The high heterogeneity in our meta-analysis is attributable to various factors, including the number of animals used in each study and the limited number of similar studies. Interpretation. Studies selected in this systematic review and meta-analysis suggest that previous physical exercise programs can reduce some of the main features related to seizure susceptibility [latency seizure onset, spikes/min, and spike amplitude (μV)] induced by the administration of different chemoconvulsants. Systematic Review Registration: PROSPERO, identifier CRD42021251949; https://www.crd.york.ac.uk/prospero/display_record.php?RecordID=251949.