Cognitive brain potentials in children, young adults, and senior citizens: Homologous components and changes in scalp distribution

Abstract

Event‐related brain potentials (ERPs) were recorded from three groups with mean ages of 7 (n = 22), 25 (n = 25), and 70 (n = 12) years while they performed a pictorial modification of Posner's (1978) letter‐matching task. In separate blocks of trials, participants had to decide whether two sequentially presented pictures were physically identical (PID), shared the same name (NID), or came from the same semantic category (CID). Conditions were always given in the same order: PID, NID, CID. Participants indicated their choice (“same” or “different”) via delayed responses following the presentation of the second pictorial slide. Participants in all three age groups produced ERP waveforms of complex morphology. The data suggested that four components (frontal positive slow wave, E‐wave preceding S2, “P300” to SI and S2, and N400 to SI and S2) might be homologous in all three age groups. However, “P300” showed marked changes in scalp distribution as a function of increasing age, shifting from a sharply focused Pz maximal distribution in children to a more equipotential distribution in senior citizens. Behavioral performance did not differentiate senior citizens and young adults, but young children performed less accurately than either older group. It appears that under conditions tapping primary memory, older adults are well able to perform in these hierarchically organized matching tasks. However, to maintain this level of performance, the senior citizens may have had to prepare more effortfully for S2 onset, as indicated by larger amplitude E waves. Moreover, for E‐wave preceding S2, and N400 elicited by S2, only the senior‐citizen data was ordered CID > NID > PID, suggesting differences in the way this group approached the tasks. Although there were few behavioral differences between groups, the presence of differences in ERP component latency and morphology suggest that similar behavioral performance may have been achieved via different processing modes as a function of chronological age.