Dissociating decision-making deficits in temporal and frontal lobe epilepsy using Bayesian methods.

PURPOSE: People with epilepsy exhibit various cognitive impairments, including deficits in executive functioning that could impact their decision-making. This study examined differences in objective risk-based decision-making among individuals with focal epilepsy, specifically comparing right temporal lobe epilepsy (RTLE), left temporal lobe epilepsy (LTLE), and frontal lobe epilepsy (FLE) against healthy controls. METHODS: Decision-making was assessed in a case-control study design with 43 adults with pharmacoresistant focal epilepsy (15 RTLE, 18 LTLE, 10 FLE) and 15 controls using the Cambridge Gambling Task (CGT). Data were analyzed using Bayesian regression models, employing likelihood functions (Beta, Lognormal, and Gaussian) tailored to the distributional properties of specific outcome measures. RESULTS: Decision quality and risk adjustment tended to be lower across epilepsy groups compared to controls, with the strength of evidence varying substantially by group and task. A suggestive divergence in behavioral measures emerged based on focus laterality: LTLE patients exhibited significant deliberative slowing, taking 1.73 times longer than controls to make decisions ($PP = 1.00$). Conversely, RTLE patients showed a pattern toward heightened delay aversion ($PP =.95$) paired with a conservative, risk-averse betting profile ($PP =.97$). The FLE group presented an intermediate profile with modest evidence of reduced decision quality. CONCLUSION: Focal epilepsy impacts decision-making efficiency in a manner contingent upon seizure focus. The observed distinctions between temporal groups tentatively suggest potential avenues for future longitudinal research into tailored neuropsychological interventions. These findings highlight the utility of Bayesian frameworks in providing a nuanced, probabilistic characterization of cognitive deficits in small sample sizes.