Biofeedback in epileptics: equivocal relationship of reinforced EEG frequency to seizure reduction

It has been reported that biofeedback training of 12- to 14-Hz activity recorded over Rolandic cortex was accompanied by a reduction in seizure incidence in four human epileptics (Sterman et al., 1974). Biofeedback training of 12- to 14-Hz activity was provided for two epileptics and had no effect on clinical EEGs, seizure incidence, or proportion of EEG spectral power in the frequency range being trained. Subsequently, biofeedback training of 6- to 12-Hz Rolandic activity was provided for three epileptics. Two patients experienced reductions in seizure not accompanied by medication changes. Since no learning of 6- to 12-Hz activity was detected, the changes in seizure incidence are not attributed to EEG biofeedback. It is suggested that the experience in the feedback setting provided these two patients with new techniques of relaxation. In view of the lack of statistical evidence of EEG changes following EEG biofeedback and the small number of patients trained to date, it appears wise to maintain a cautious attitude until the issue of causality is clear.