Event-related potentials in migraine: a meta-analysis

Introduction. The present research is a meta-analytic review which summarizes the results of studies that have assessed information processing through ERP techniques comparing migraine subjects in the interictal period with controls. Method. Studies of potential interest (n = 133) were identified in PsycLit and MedLine databases. Sixty-five (53 conducted on adults and 12 on children and adolescents) presented valid data reporting a total of 332 effects. Different meta-analyses were conducted considering the following different experimental paradigms: Flash, Steady-State, Pattern-Reversal, Intensity dependence, CNV. A further meta-analysis was conducted for evaluating habituation effect across paradigms. Effect size indices (g, d and the equivalent r) were computed through DSTAT software and analysed following Hedges and Olkin procedures. Results. Studies consistently showed, in migraine sufferers compared with control, greater amplitudes of ERP during acoustic stimulation of increasing intensities (d = +0.73; r = 0.34, P < 0.001) and greater CNV amplitudes, particularly in the initial component of CNV (d = +0.58; r = 0.28, P < 0.001). Furthermore, a lack of habituation (d = -0.85; r = -0.39, P < 0.001) in migrainous subjects was clearly demonstrated. No difference between migraine subtypes was found. Discussion. Overall, our meta-analytic findings clearly demonstrate greater cortical reactivity in migraine subjects than in controls, as indicated by cortical hyperexcitability and lack of habituation. Results suggest that ERP recordings are useful for revealing differences in information processing between migrainous subjects and controls, but they do not contribute to differential diagnosis of migraine sub-types.