Biological determinants of P300: the effects of a barbiturate on latency and amplitude

Polich & Kok (1995) (Biological Psychology, 41, 103-146) have recently argued that P300 is not only sensitive to specific 'cognitive' variables, but also to non-specific biological processes such as arousal. Fluctuations in arousal are said to be indexed by an inverse relationship between latency and amplitude. We tested this hypothesis with a drug that decreases arousal-the barbiturate secobarbital sodium. Twelve subjects performed a visual 80-20% oddball task at two levels of stimulus quality and after ingesting the drug (2.9 mg/kg body weight) or a placebo. Reaction time (RT) and P300 were collected simultaneously and the latter was analyzed on both a single trial and average basis. The RT results confirmed that secobarbital interacts with stimulus quality. Secobarbital slowed single trial P300 by about half as much as RT, and this slowing was additive with stimulus quality. Thus the two measures dissociated. Secobarbital did not influence P300 amplitude. Average P300 revealed the same pattern of results, although the size of the latency effects was somewhat attenuated. RT and P300 latency were more strongly correlated than P300 latency and amplitude. We propose that P300 latency reflected the slowing of stimulus evaluation produced by the depressant properties of the drug, and that fluctuations in arousal are not necessarily associated with a simple inverse relationship between P300 latency and amplitude.