The FMR 1 premutation is associated with a complex clinical phenotype, with increased risk for outcomes across the domains of psychological disorders, motor functioning, and reproductive health. A key gap is understanding intermediate processing that may help to explain how the FMR 1 premutation alters brain functioning to confer increased risk across these domains. The current study begins to address this gap by focusing on two candidate, interrelated processes: cerebellar GABA and executive function. Data were collected from a cohort of 15 adult female FMR 1 premutation carriers and a comparison group of 15 age-matched unaffected controls. Cerebellar GABA concentration was measured using edited magnetic resonance spectroscopy (MEGA-sLASER) at 3 T. Executive function was measured across two units of analysis: the error-related negativity, an EEG marker of automatic error detection, as well as performance-based measures from validated neuropsychological tests. Premutation carriers exhibited an atypical scalp topography of the error-related negativity. In controls but not in premutation carriers, error-related brain activity was associated with cerebellar GABA concentration and performance-based executive function. This pattern of preliminary findings suggests that the FMR 1 premutation may alter the neural network involved in error monitoring. If these findings are confirmed in larger cohorts, they have potential implications for understanding the emergence of the broader clinical phenotype in adult FMR 1 premutation carriers. • In controls, error-related brain activity was moderately associated with cerebellar GABA concentration and performance-based executive functioning. • Notably, in women with the FMR 1 premutation, error related brain activity was unrelated to performance based executive functioning and to cerebellar GABA concentration. • Topography of brain activity during the error-processing task was altered among women with the FMR 1 premutation relative to controls, indicating that error monitoring may operate through a distinct neural network for this population.
Harold et al. (Fri,) studied this question.