Selective attention effects on somatosensory evoked potentials

Short and long latency somatosensory evoked potentials (SEPs) to median nerve stimulation were recorded over the contralateral hemisphere. Simultaneously, signals evoked by the same stimulus were monitored at Erb's point. Recordings were made during three conditions which were presented in a different random order to each of the subjects tested. During the control condition the subjects were instructed to attend to and count the number of electrical pulses delivered to the median nerve of their right hand. During the two task conditions, in addition to the pulses, the subjects received stimulation on the dorsal surface of one of their hands. This consisted in drawing circles lightly for the duration of SEP recording using a cotton swab (Q-tip). During the trial, the Q-tip was momentarily withdrawn 15 to 20 times and the subject's task was to ignore the pulses, attend to this cutaneous stimulation and count the number of times the cutaneous stimulation was interrupted. SEPs to the pulse were significantly greater in amplitude when cutaneous stimulation was delivered to the same hand as the pulse (the right hand) than when it was delivered to the left hand. This effect was confined to the long-latency SEPs and did not appear in either the Erb's point response or the short latency SEPs. These data indicate that selective attention to peripheral stimulation is a relatively late process mediated by cortical mechanisms and argue against the notion of early suppression of irrelevant stimulus channels in subcortical centers or in the periphery. Moreover, they indicate that in the specific case of limb stimulation, selection is based on physical channel (the hand) rather than the more subtle nature of the quality of information.