The impact of motor activity on intracerebral ERPs: P3 latency variability in modified auditory odd-ball paradigms involving a motor task.

The P3 wave of event-related potentials was recorded with intracranial electrodes in 24 epileptic patients during the pre-surgical evaluation of epilepsy surgery. Three different cognitive auditory paradigms were used: (1) odd-ball paradigm with no output required (PGI) where patients had simply to recognize target tones, (2) odd-ball with motor response (PGII), where patients had to press a button in response to target tones, and (3) odd-ball with both counting task and motor response (PGIII), where patients had to recognize target tones, press a button in response to them, and count their number. The occurrence of P3 potential, its latency and amplitude, and the dependence of P3 latency on the task complexity were calculated. Identifiable P3 potentials in all the three paradigms were recorded from locations in mesial cortex (18 locations mesial temporal, eight locations mesial frontal, two locations mesial parietal) and lateral sites (eight sites lateral temporal, five lateral frontal, and two lateral parietal). P3 latency values ranged from 257 to 320 ms in all explored cortical areas when PGI was used; they significantly increased or decreased during PGII and PGIII, depending on the task and structure explored. In the mesial temporal cortex, the changes of P3 latency between paradigms were minimal. In the mesial parietal cortex, there was significant P3 delay in both PGII and PGIII relative to PGI. In the mesial frontal cortex, there was a significant latency decrease in PGII, and practically identical mean latency in PGI and PGIII. In all lateral cortices (temporal, frontal and parietal), there was always a P3 latency increase in PGII and PGIII relative to PGI, the most significant results being observed in the parietal and frontal lateral areas. The results support the multi-generator theory of P3. Prolongation of the mean P3 latency in lateral frontal and parietal cortices when the paradigm involved the execution of a motor task might reflect specific gating in this area during active movements, while the absence of modification in the temporal lobe may reflect minimal involvement of this region in motor planning or processing. The prolongation of mean P3 latency in practically all lateral structures in PGIII suggests that most cortical areas were involved in the cognitive functions needed for this test. The finding of reduction and subsequent prolongation of P3 latency in the mesial frontal cortex might reflect the unique specialization of this area and its specific involvement in motor processing.