Purpose: Conflicting evidence about impairment of attention systems and the absence of a working model of attention has contributed to lack of clarity about comorbidity of attention problems in rolandic epilepsy (RE). Impairments in distributed attention systems may inform a disease model for RE, as well as direct interventions.
Methods: We used a systematic review of the relevant literature published in English from 1990 to 2006 to evaluate impairment in attention in RE. The Mirsky and Posner models of attention were evaluated for applicability, and studies were reviewed for design, instrumentation, and congruence with the Posner model of attention.
Results: Fourteen studies were identified: seven using a cross-sectional design (six active EEG abnormalities; one EEG remission) and seven longitudinal studies (abnormal EEGs and follow-up until normalized). According to the Posner model of attention, 12 studies employed measures that tapped the alerting network, 11 studies the orienting network, and eight the executive network. Nearly all controlled studies demonstrated impairments in all tested attention networks. In contrast, uncontrolled studies uniformly did not demonstrate impairments. Follow-up studies demonstrated complete or near complete resolution of attention impairments.
Discussion: The weight of evidence, defined as the majority of studies evaluated, suggests that all three attention systems are impaired in children with active centrotemporal spikes (CTS), implying a more widespread functional cortical disturbance in RE than previously held. These impairments resolve upon EEG remission, suggesting a common pathological basis to the autosomal dominant CTS trait. Sources of methodological variation are discussed with recommendations for future investigations.