- Email: [email protected]
Auditory evoked transient and sustained potentials in the human EEG: I. Effects of expectation of stimuli
Psychiatry Research, 1, 297-306 (1979) @ Elsevier/North-Holland
Biomedical
297 Press
Auditory Evoked Transient and Sustained Potentials in the Human EEG: I. Effects of Expectation of Stimuli Riitta Hari, Mikko Sams, and Timo J&vilehto Received
September
5, 1979; accepted
November
7, 1979.
Abstract. The characteristics of auditory evoked transient and sustained potentials were recorded using trains of four-tone stimuli of l-second duration (interstimulus interval = 1 second) presented once every minute. The subject either attentively expected the stimuli or ignored them while reading. The electroencephalogram was recorded from derivations Cz-Al and Fz-Al. Expectation of the stimuli was associated with increased amplitudes of the transient responses both at the first stimulus of the train and during stimulus repetition. In contrast, the sustained potential at the first stimulus of the train was unchanged or smaller when the subject expected the stimuli. During stimulus repetition, however, the amplitude of the potential was enhanced by expectation of the stimuli. The results support the hypothesis of two sustained potential components and stress the importance of stimulus repetition rate when sustained potentials are studied. Key Words, Audition,
transient
evoked potentials, sustained evoked potentials,
orienting. An auditory stimulus of variable duration elicits both transient and sustained potentials in the human electroencephalogram (EEG). The transient evoked potentials have been extensively studied during the last decades whereas the sustained potentials have attracted less attention. Pi&on et al. (1978a, 1978b) studied the effects of different stimulation parameters on auditory evoked transient and sustained potentials. The authors demonstrated that the amplitude of both transient and sustained potentials increased with increasing stimulus intensity and decreased with increasing stimulus repetition rate, although the decrease was smaller for the sustained potential. On the basis of the differences in scalp distributions and stimulus relationships, the authors suggested different neural generators for auditory evoked transient and sustained potentials. Additionally, two underlying generator mechanisms were suggested for the sustained potential to explain the findings on the frequency specificity of its refractory period. The effect of stimulus repetition on the amplitude and posterior scalp distribution of auditory and visual evoked potentials was studied by Jtirvilehto et al. (1978). Tones and flashes of l-second duration were presented in trains of six stimuli. The first stimulus of the train elicited a large negative sustained potential that was maximal in
Preliminary results of this investigation were presented at the Fourth Biennial International Symposium on Biological Research in Alcoholism, Zurich 1978 (Hari et al., in press). Ritta Hari, M.D., is Assistant Doctor, Department of Physiology, Mikko Sams, M.A., is Assistant, Department of Psychology, and Timo Jarvilehto, Ph.D., is Docent, Department of Psychology, University of Helsinki. (Reprint requests to Dr. Hari at Laboratory of Clinical Neurophysiology, Department of Neurology, University Central Hospital of Helsinki, Haartmaninkatu 4, 00290 Helsinki 29. Finland.)
298 amplitude at vertex for both modalities. Repetition of the stimuli resulted in a marked decrease in the amplitude of the sustained potential. For auditory stimulation a small negative shift was then seen only at vertex and for visual stimulation only at the parietooccipital area. The results indicated that sensory evoked sustained potentials are composed of at least two components. One of these was suggested to reflect neural events associated with the orienting reflex whereas the other component may be related to processing of the stimulus in its specific projection area. The repetition rate of stimuli is thus important in determining the type of potential recorded in different experiments. In the present study, the short-term effect of stimulus repetition on auditory evoked transient and sustained potentials was examined when the subject either expected or ignored short trains of tone stimuli.
Methods Subjects and Experimental Conditions. Eleven subjects (seven females, four males; age range, 19-29 years) participated in the experiments, which were carried out in an electrically shielded, soundproof chamber. The subject sat with his eyes open while 40 trains of four identical tone stimuli of l-second duration (1 KHz square-wave bursts, intensity 80 dB SPL, l-second interstimulus interval) were delivered through a loudspeaker situated 2 meters in front of the subject’s head. The repetition rate of the trains was one per minute. The subject was instructed either to attend to the stimulus trains or to ignore them while reading. After each train, the subject indicated whether he had been expecting or ignoring the train. The subject was instructed to avoid eye movements and blinks during the stimulus trains. Two experiments were carried out with every subject. Data Recording and Processing. The EEG was recorded from derivation Cz-Al and in nine subjects also from derivation Fz-Al. Ag/ AgCl cup electrodes (I 1 mm diameter) were fixed with collodium on the scalp (electrode resistance below 5k ohms). Eye movements were monitored with an electrode situated just above the left eye. The subject was grounded at the left wrist. The EEG and eye movement recordings were amplified (bandpass 0.08-50 Hz, 3 dB points), stored on magnetic tape and written on paper. All EEG epochs contaminated by visually detectable muscle, electrode, eye movement, and dropout artifacts were omitted from the analysis. The EEG responses and corresponding eye movement recordings were selectively summated by a computer according to the serial position of the stimulus in the stimulus train and according to the indicated subjective state (expecting = E, ignoring = I). Sampling began 200 ms before the stimulus, and 500 samples were taken during an analysis period of 1,920 ms. The baseline for measurements was defined as the mean voltage level during the first 150 ms of the analysis period. For the transient responses to stimulus onset, the peak amplitudes of components N 120, P200, and P300 were measured with reference to this baseline. For the transient off-response, peak-to-peak (N 120-P200) amplitudes were measured. For components N 120 and P200 of the transient on and off responses, the peak
299 latencies were measured. The mean amplitude of the sustained potential was calculated over a period which started at the crossing of the curve with the baseline just after the positive component(s) of the transient on response and ended with the stimulus offset. The mean amplitude of the potential during the last 210 ms of the analysis period was calculated from the averaged responses to the first stimulus of the train. Statistical Analysis. In the statistical analysis of data, averages of the two experiments were first calculated for each subject. The comparison of state (expecting, ignoring) effects was done by testing individual differences against zero with twotailed t tests. In two cases where the distribution was skewed, the Wilcoxon test for pair differences was used. Results Transient Responses to the Stimulus Onset (On Response). Fig. 1 shows the averaged potentials of one subject at the consecutive stimuli of the stimulus train when Fig. 1. Typical subject’s auditory evoked potentials
I
IGNORING
EXPECTING 2opv
-
Average responses to the four consecutive (ordinal numbers) tones of thestimulustrain when the subject either expected or ignored the stimuli. Recording from &-AI. Duration of thestimulus is 1 second. Number of summations 17.
300 the subject either expected or ignored the stimuli. At the first stimulus ofthe train, the mean amplitude of the component N 120 of the transient response to the stimulus onset was at Cz 32.8 + 12.5, PV (mean * standard deviation) when the subject expected the stimuli and 28.3 ? 11.6 PV when the subject ignored the stimuli (significant difference, p < 0.01, t test; see Fig. 2). There was a marked decrease in the amplitude of the component N 120 from the first to the second response, and thereafter the amplitudes remained at a stable level (Fig. 2). The amplitude of N 120 for expected stimuli of the mean response averaged over the second to fourth stimulus was 37 & 14% of the amplitude of the first response and that of ignored stimuli 39 + 16%. The decrease was significant in both cases @ < 0.001, t test) but there was no significant difference in the rate of decrease between N120 for expected and ignored stimuli. As with the first stimulus, the amplitude of N120 for expected stimuli was also larger than that of ignored stimuli during stimulus repetition (mean difference 1.3 pV, p < 0.05, Wilcoxon).
Fig. 2. N120 and P200 amplitude
for each stimulus
position
b"l -3o-
-20 i
0
: 1.
I
2.
3.
stimulus
4.
The mean amplitude of the N120 and P200 components of the transient on response at consecutive stimuli of the stimulus train when subject either expected or ignored the stimuli. All subjects are pooled. The bars show the standard error of the mean.
301 At the first stimulus the amplitude of the component N 120 was larger at Cz than at Fz (mean difference 11.3 pV,p < 0.01, t test; 9 subjects), but during stimulus repetition the amplitude maximum changed to the frontal area (mean difference 1.3 pV, p < 0.05; 2 test). At Cz the amplitude of the component P200 was 14.6 + 10.1 PV for the first stimulus of the train when the subject expected the stimuli and 11.9 k 9.0 PV when the subject ignored the stimuli (difference statistically significant at p < 0.05, Wilcoxon). During stimulus repetition these amplitudes did not differ from each other, nor did they show any significant decrease in amplitude. During the first stimulus, there was no significant difference in the amplitudes of components P200 at Cz and Fz, but during repeated stimuli the amplitudes were larger at Cz than at Fz (mean difference 1.6 I.IV, p < 0.05, t test). During the first stimulus of the train, P200 was followed by a positive deflection P300 in the recordings of eight subjects. P300 appeared only during the first stimulus, and never during repeated stimuli. There were no significant differences in the amplitudes of P300 whether the subject was expecting or ignoring the stimuli, nor was there any significant difference between the amplitude of this component measured at Fz and cz. The latencies of the components N 120 and P200 did not differ from each other when the subject expected or ignored the stimuli (see Fig. 3). In all experiments, stimulus repetition resulted in a marked reduction of both latencies. The mean reduction from the first to the averaged second to fourth response was 23 ms for N120 (expected), 27 ms for N120 (ignored), 39 ms for P200 (expected), and 44 ms for P200 (ignored) (all reductions statistically significant, p < 0.001, t test). Transient Responses to the Stimulus Offset (Off Response). The transient off response was usually a positive deflection, P200, following the negative sustained potential, and it was only in a few cases preceded by a negative deflection N 120. When no negative deflection was seen, the N 120 component was measured at the point where the positive deflection started. The peak-to-peak N120-P200 amplitudes of the off response are shown in Fig. 4. Stimulus repetition did not have any significant effect on the amplitude of the off response which, on the average, was 10.5 IL 2.9 r.lV when the subject expected the stimuli and 9.8 z!z2.8 PV when the subject ignored the stimuli (difference not statistically significant). At the first stimulus, the peak-to-peak amplitude of the off response was larger at Fz than at Cz (mean difference 2.1 pV, p < 0.05, t test). During stimulus repetition, this difference disappeared. Fig. 5 shows the latencies of the N120 and P200 components of the on and off responses. Stimulus repetition did not have any significant effect on the latencies of the off responses, which were similar to the latencies of the N120 and P200 components elicited by repeated stimuli. Sustained Potentials. The mean amplitudes of sustained potentials (SPs) elicited by the consecutive stimuli of the stimulus train are shown in Fig. 6. At the first stimulus there was a trend to a larger amplitude of SP when the subject ignored the stimuli (9.8
302 Fig. 3. N120 and P200 latency for each stimulus
position
expecting O---& ignoring
PP
i:;;_
I
, 1.
I
I
I
2.
3.
4.
stimulus The mean peak latencies of the N120 and P200 component of the transient on response at consecutive stimuli of the stimulus train when subjectseitherexpected or ignored the stimuli. The bars show the standard error of the mean.
k 5.6 pV> as compared to that when the subject expected the stimuli (8.5 + 4.1 pV>. This difference was not statistically significant. At the end of the analysis period (last 210 ms), however, the amplitude of the potential was significantly more negative during the I condition than during the E condition (mean difference 2.7 t.~V,p< 0.005, t test). In both conditions the potential remained negative at the end of the analysis period @ < 0.025, t test). There was a marked decrease in the amplitudes of SP from the first to the second response in all but one subject, and thereafter the amplitudes remained at a stable level. The amplitude of the SP (expected) of the averaged second to fourth response was 32 Z!I 17% of that of the first response and that of SP (ignored) 25 f 229& respectively. In both cases the decrease in the amplitude was significant @ < 0.005, t test), but the rate of decrease was not significantly different. The percentage decrease of the SP amplitude was slightly smaller than that of the N120 component of the on response, but this difference was not statistically significant. The amplitudes of both expected and ignored SPs were also significantly negative during stimulus repetition 0, < 0.005, t test). The amplitude of SP (expected) was
303
Fig. 4. Off response N120-P200 peak-to-peak amplitude
e---0 0---c
L
, 1.
I
2.
expecting Ignoring
I
I
3.
4.
stimulus The mean peak-to-peak (N120-P200) amplitude of the oft response at consecutive stimuli of the stimulus train when the subject either expected or ignored the stimuli. The bars show the standard error of the mean.
slightly larger than that of SP (ignored) during repetition of the stimuli (mean difference 0.9 pV, p < 0.05). At the first response no significant difference in the amplitudes of SPs between Fz and Cz was seen. During stimulus repetition, however, the SP amplitude was larger at Fz than at Cz (mean difference 1.2 I.IV, p < 0.05,t test).
Discussion The present work confirms the marked effect of stimulus repetition on both transient and sustained potentials elicited by auditory stimuli (Fruhstorfer et al., 1970; J&ilehto et al., 1978). The effect of expectation of the stimuli on the potentials was, however, more complex. The transient potentials were larger in amplitude when the subject expected the stimuli whereas the sustained potentials were unchanged at the first stimulus of the train or smaller when the subject expected the stimuli. During the stimulus repetition, these potentials also increased in amplitude with expectation. We have suggested earlier (Jgrvilehto et al., 1978) that with infrequent stimulation, both the transient and sustained potentials could reflect activity in neural systems that control phasic and tonic excitability in the motor and sensory system and prepare the organism for immediate action. The dissociation between the effects of expectation on the transient and sustained potentials at the first stimulus of the train indicates that the expectation of the stimuli has different effects on these two postulated systems. The remaining negativity at the end of the analysis period at the first stimulus of the train
304 Fig. 5. N120 and P200 latency for on and off responses
3----o
ON-response OFF-response
p2 p2
150-
100
Nl “1
I).‘r;-,l
t
1 1.
2.
3.
4.
The mean peak latencies of the negative (N120) and positive (P200) deflections of the transient on and off responses at consecutive stimuli of the stimulus train. Values for expecting and ignoring condition are combined. The bars show the standard error of the mean.
indicated that the “orienting” component of the sustained potential (see Introduction) is not restricted to the duration of the stimulus. The more specific “sensory” component of the sustained potential remaining during repetition of stimuli is smaller in amplitude than the “orienting” component, and it seems to be masked by it. Thus it can be discerned only when the nonspecific component is absent, i.e., when fast stimulus repetition rates are used. When attention is paid to the stimuli throughout their duration (Picton et al., 197&z) or when the stimuli are attentively expected, the amplitude of the specific component increases. The amplitude of the specific component is also larger when stimuli of low rather than of high tonal frequency are used (Picton et al., 1978b). Altogether, the specific component seems to be closely dependent on the stimulation characteristics used (David et al., 1969; Keidel, 1976; Picton et al., 1978b). A novel auditory stimulus may elicit phasic negative shifts in the human EEG that temporally coincide with the N120 component of the transient on response and summate on it (Naatanen and Michie, 1979). The marked amplitude reduction of the N120 component during stimulus repetition may be due to the disappearance of this
305 Fig. 6. Sustained
potential
amplitude
for each stimulus
SUSTAINED
position
POTENTIAL
[@I I
-8
I
7’
1
expecting ignoring
1 -6
I
7
1.
2.
I
I
3.
4.
stimulus The mean amplitude of the sustained potential at consecutive stimuli of the stimulus train when subjects either expected or ignored the stimuli. The bars show the standard error of the mean.
component. The remaining transient potential exhibits many features (frequency and intensity dependence, for example) similar to the specific sustained potential (Picton et al., 19786). Thus it seems that the transient evoked potential may be composed of two functionally different components, just as appears to be true of the sustained potential. The decrease of the amplitude of the N120 component of the on response with stimulus repetition was similar to that earlier reported with click stimuli (Fruhstorfer et al., 1970). However, only a trend toward smaller amplitudes of P200 with stimulus repetition was seen, whereas Fruhstorfer et al. (1970) found that P200 also showed a clear decrease in amplitude with stimulus repetition. This difference may be due to the longer stimuli and longer time constant recordings used in the present study. According to Picton et al. (1978~) the auditory evoked sustained potential begins 150 ms after stimulus onset and thus during the P200 component. A slow surface negative shift may then summate on P200 and shift it to a negative direction. This effect will be most prominent in cases where the sustained potential is large-at the first stimulus of the train in the present study. Variable opinions exist about the relationship between the on and off responses (Johannsen et al., 1972; Klingenberg-Schweitzer and Tepas, 1974; Onishi and Davis, 1968). Both the interstimulus interval and the duration of the stimulus are important for this relationship, because the interstimulus interval determines the degree of decrease of the amplitude of the on response, and the duration of the stimulus
306 determines the additional decrease of the off response in relation to the on response. In the present study where the interstimulus interval within the train was the same as the duration of the stimuli, the off responses were in many respects (form, amplitude, latency) similar to the P200 components of the on responses recorded during stimulus repetition. This indicates that the off responses reflect similar neural events as the P200 component of the on response. References David, E., Finkenzellx, P., Kallert, S., and Keidel, W.D. Akustischen Reizen zugeordnete GleichspannungGnderungen am intakten Schadel des Menschen. Pfliigers Archiv; European
Journal
Fruhstorfer,
of Physiology,
309, 362 (1969).
of the auditory evoked response in man. Electroencephalography and Clinical Neurophysiology, 28, 153 (1970). Hari, R., Sams, M., and Jlrvilehto, T. Effects of small ethanol doses on the auditory evoked transient and sustained potentials in the human EEG. In: Begleiter, H., ed. Biological Effects of Alcohol. Plenum Press, New York (in press). JPrvilehto, T., Hari, R., and Sams, M. Effect of stimulus repetition on negative sustained potentials elicited by auditory and visual stimuli in the human EEG. Biological Psychology, 7,
H., Soveri, P., and Jarvilehto, T. Short-term habituation
1 (1978).
Johannsen, H.S., Keidel, W.D., and Spreng, M. Der Einfluss von Intensitat und Dauer der Beschallung auf den Off-Effect des akustisch evozierten Potentials. Archivfur Klinische und Experimentelle Ohren-, Nasen- und Kehlkopfheilkunde, 201, 208 ( 1972). Keidel, W.D. The physiological background of the electric response audiometry. In: Keidel, W.D., and Neff, W.D., eds. Handbook of Sensory Physiology: Auditory System. Vol. V. Springer, Berlin (1976). Klingenberg-Schweitzer, P., and Tepas, D.1. Intensity effects of auditory evoked brain response to the stimulus onset and cessation. Perception and Psychophysics, 16, 396 (1974). Nsi’riMnen, R., and Michie, P.T. Different variants of endogenous negative brain potentials in performance situations: A review and classification. In: Lehman, D., and Callaway, E., eds. Human Evoked Potentials. Plenum Publishing Corporation, New York (1979). Onishi, S., and Davis, H. Effects of duration and rise time of tone burst on evoked V potentials. Journal
of the Acoustical
Society
of America,
44, 582 (1968).
Picton, T.W., Woods, D.L., and Proulx, G.B. Human auditory sustained potentials: I. The nature of the response. Electroencephalography and Clinical Neurophysiologv, 45, 186 (1978a).
Picton, T.W., Woods, D.L., and Proulx, G.B. Human auditory sustained potentials: Stimulus relationships. Electroencephalography and Clinical Neurophysiology, 45, (19786).
II. 198
Biomedical
297 Press
Auditory Evoked Transient and Sustained Potentials in the Human EEG: I. Effects of Expectation of Stimuli Riitta Hari, Mikko Sams, and Timo J&vilehto Received
September
5, 1979; accepted
November
7, 1979.
Abstract. The characteristics of auditory evoked transient and sustained potentials were recorded using trains of four-tone stimuli of l-second duration (interstimulus interval = 1 second) presented once every minute. The subject either attentively expected the stimuli or ignored them while reading. The electroencephalogram was recorded from derivations Cz-Al and Fz-Al. Expectation of the stimuli was associated with increased amplitudes of the transient responses both at the first stimulus of the train and during stimulus repetition. In contrast, the sustained potential at the first stimulus of the train was unchanged or smaller when the subject expected the stimuli. During stimulus repetition, however, the amplitude of the potential was enhanced by expectation of the stimuli. The results support the hypothesis of two sustained potential components and stress the importance of stimulus repetition rate when sustained potentials are studied. Key Words, Audition,
transient
evoked potentials, sustained evoked potentials,
orienting. An auditory stimulus of variable duration elicits both transient and sustained potentials in the human electroencephalogram (EEG). The transient evoked potentials have been extensively studied during the last decades whereas the sustained potentials have attracted less attention. Pi&on et al. (1978a, 1978b) studied the effects of different stimulation parameters on auditory evoked transient and sustained potentials. The authors demonstrated that the amplitude of both transient and sustained potentials increased with increasing stimulus intensity and decreased with increasing stimulus repetition rate, although the decrease was smaller for the sustained potential. On the basis of the differences in scalp distributions and stimulus relationships, the authors suggested different neural generators for auditory evoked transient and sustained potentials. Additionally, two underlying generator mechanisms were suggested for the sustained potential to explain the findings on the frequency specificity of its refractory period. The effect of stimulus repetition on the amplitude and posterior scalp distribution of auditory and visual evoked potentials was studied by Jtirvilehto et al. (1978). Tones and flashes of l-second duration were presented in trains of six stimuli. The first stimulus of the train elicited a large negative sustained potential that was maximal in
Preliminary results of this investigation were presented at the Fourth Biennial International Symposium on Biological Research in Alcoholism, Zurich 1978 (Hari et al., in press). Ritta Hari, M.D., is Assistant Doctor, Department of Physiology, Mikko Sams, M.A., is Assistant, Department of Psychology, and Timo Jarvilehto, Ph.D., is Docent, Department of Psychology, University of Helsinki. (Reprint requests to Dr. Hari at Laboratory of Clinical Neurophysiology, Department of Neurology, University Central Hospital of Helsinki, Haartmaninkatu 4, 00290 Helsinki 29. Finland.)
298 amplitude at vertex for both modalities. Repetition of the stimuli resulted in a marked decrease in the amplitude of the sustained potential. For auditory stimulation a small negative shift was then seen only at vertex and for visual stimulation only at the parietooccipital area. The results indicated that sensory evoked sustained potentials are composed of at least two components. One of these was suggested to reflect neural events associated with the orienting reflex whereas the other component may be related to processing of the stimulus in its specific projection area. The repetition rate of stimuli is thus important in determining the type of potential recorded in different experiments. In the present study, the short-term effect of stimulus repetition on auditory evoked transient and sustained potentials was examined when the subject either expected or ignored short trains of tone stimuli.
Methods Subjects and Experimental Conditions. Eleven subjects (seven females, four males; age range, 19-29 years) participated in the experiments, which were carried out in an electrically shielded, soundproof chamber. The subject sat with his eyes open while 40 trains of four identical tone stimuli of l-second duration (1 KHz square-wave bursts, intensity 80 dB SPL, l-second interstimulus interval) were delivered through a loudspeaker situated 2 meters in front of the subject’s head. The repetition rate of the trains was one per minute. The subject was instructed either to attend to the stimulus trains or to ignore them while reading. After each train, the subject indicated whether he had been expecting or ignoring the train. The subject was instructed to avoid eye movements and blinks during the stimulus trains. Two experiments were carried out with every subject. Data Recording and Processing. The EEG was recorded from derivation Cz-Al and in nine subjects also from derivation Fz-Al. Ag/ AgCl cup electrodes (I 1 mm diameter) were fixed with collodium on the scalp (electrode resistance below 5k ohms). Eye movements were monitored with an electrode situated just above the left eye. The subject was grounded at the left wrist. The EEG and eye movement recordings were amplified (bandpass 0.08-50 Hz, 3 dB points), stored on magnetic tape and written on paper. All EEG epochs contaminated by visually detectable muscle, electrode, eye movement, and dropout artifacts were omitted from the analysis. The EEG responses and corresponding eye movement recordings were selectively summated by a computer according to the serial position of the stimulus in the stimulus train and according to the indicated subjective state (expecting = E, ignoring = I). Sampling began 200 ms before the stimulus, and 500 samples were taken during an analysis period of 1,920 ms. The baseline for measurements was defined as the mean voltage level during the first 150 ms of the analysis period. For the transient responses to stimulus onset, the peak amplitudes of components N 120, P200, and P300 were measured with reference to this baseline. For the transient off-response, peak-to-peak (N 120-P200) amplitudes were measured. For components N 120 and P200 of the transient on and off responses, the peak
299 latencies were measured. The mean amplitude of the sustained potential was calculated over a period which started at the crossing of the curve with the baseline just after the positive component(s) of the transient on response and ended with the stimulus offset. The mean amplitude of the potential during the last 210 ms of the analysis period was calculated from the averaged responses to the first stimulus of the train. Statistical Analysis. In the statistical analysis of data, averages of the two experiments were first calculated for each subject. The comparison of state (expecting, ignoring) effects was done by testing individual differences against zero with twotailed t tests. In two cases where the distribution was skewed, the Wilcoxon test for pair differences was used. Results Transient Responses to the Stimulus Onset (On Response). Fig. 1 shows the averaged potentials of one subject at the consecutive stimuli of the stimulus train when Fig. 1. Typical subject’s auditory evoked potentials
I
IGNORING
EXPECTING 2opv
-
Average responses to the four consecutive (ordinal numbers) tones of thestimulustrain when the subject either expected or ignored the stimuli. Recording from &-AI. Duration of thestimulus is 1 second. Number of summations 17.
300 the subject either expected or ignored the stimuli. At the first stimulus ofthe train, the mean amplitude of the component N 120 of the transient response to the stimulus onset was at Cz 32.8 + 12.5, PV (mean * standard deviation) when the subject expected the stimuli and 28.3 ? 11.6 PV when the subject ignored the stimuli (significant difference, p < 0.01, t test; see Fig. 2). There was a marked decrease in the amplitude of the component N 120 from the first to the second response, and thereafter the amplitudes remained at a stable level (Fig. 2). The amplitude of N 120 for expected stimuli of the mean response averaged over the second to fourth stimulus was 37 & 14% of the amplitude of the first response and that of ignored stimuli 39 + 16%. The decrease was significant in both cases @ < 0.001, t test) but there was no significant difference in the rate of decrease between N120 for expected and ignored stimuli. As with the first stimulus, the amplitude of N120 for expected stimuli was also larger than that of ignored stimuli during stimulus repetition (mean difference 1.3 pV, p < 0.05, Wilcoxon).
Fig. 2. N120 and P200 amplitude
for each stimulus
position
b"l -3o-
-20 i
0
: 1.
I
2.
3.
stimulus
4.
The mean amplitude of the N120 and P200 components of the transient on response at consecutive stimuli of the stimulus train when subject either expected or ignored the stimuli. All subjects are pooled. The bars show the standard error of the mean.
301 At the first stimulus the amplitude of the component N 120 was larger at Cz than at Fz (mean difference 11.3 pV,p < 0.01, t test; 9 subjects), but during stimulus repetition the amplitude maximum changed to the frontal area (mean difference 1.3 pV, p < 0.05; 2 test). At Cz the amplitude of the component P200 was 14.6 + 10.1 PV for the first stimulus of the train when the subject expected the stimuli and 11.9 k 9.0 PV when the subject ignored the stimuli (difference statistically significant at p < 0.05, Wilcoxon). During stimulus repetition these amplitudes did not differ from each other, nor did they show any significant decrease in amplitude. During the first stimulus, there was no significant difference in the amplitudes of components P200 at Cz and Fz, but during repeated stimuli the amplitudes were larger at Cz than at Fz (mean difference 1.6 I.IV, p < 0.05, t test). During the first stimulus of the train, P200 was followed by a positive deflection P300 in the recordings of eight subjects. P300 appeared only during the first stimulus, and never during repeated stimuli. There were no significant differences in the amplitudes of P300 whether the subject was expecting or ignoring the stimuli, nor was there any significant difference between the amplitude of this component measured at Fz and cz. The latencies of the components N 120 and P200 did not differ from each other when the subject expected or ignored the stimuli (see Fig. 3). In all experiments, stimulus repetition resulted in a marked reduction of both latencies. The mean reduction from the first to the averaged second to fourth response was 23 ms for N120 (expected), 27 ms for N120 (ignored), 39 ms for P200 (expected), and 44 ms for P200 (ignored) (all reductions statistically significant, p < 0.001, t test). Transient Responses to the Stimulus Offset (Off Response). The transient off response was usually a positive deflection, P200, following the negative sustained potential, and it was only in a few cases preceded by a negative deflection N 120. When no negative deflection was seen, the N 120 component was measured at the point where the positive deflection started. The peak-to-peak N120-P200 amplitudes of the off response are shown in Fig. 4. Stimulus repetition did not have any significant effect on the amplitude of the off response which, on the average, was 10.5 IL 2.9 r.lV when the subject expected the stimuli and 9.8 z!z2.8 PV when the subject ignored the stimuli (difference not statistically significant). At the first stimulus, the peak-to-peak amplitude of the off response was larger at Fz than at Cz (mean difference 2.1 pV, p < 0.05, t test). During stimulus repetition, this difference disappeared. Fig. 5 shows the latencies of the N120 and P200 components of the on and off responses. Stimulus repetition did not have any significant effect on the latencies of the off responses, which were similar to the latencies of the N120 and P200 components elicited by repeated stimuli. Sustained Potentials. The mean amplitudes of sustained potentials (SPs) elicited by the consecutive stimuli of the stimulus train are shown in Fig. 6. At the first stimulus there was a trend to a larger amplitude of SP when the subject ignored the stimuli (9.8
302 Fig. 3. N120 and P200 latency for each stimulus
position
expecting O---& ignoring
PP
i:;;_
I
, 1.
I
I
I
2.
3.
4.
stimulus The mean peak latencies of the N120 and P200 component of the transient on response at consecutive stimuli of the stimulus train when subjectseitherexpected or ignored the stimuli. The bars show the standard error of the mean.
k 5.6 pV> as compared to that when the subject expected the stimuli (8.5 + 4.1 pV>. This difference was not statistically significant. At the end of the analysis period (last 210 ms), however, the amplitude of the potential was significantly more negative during the I condition than during the E condition (mean difference 2.7 t.~V,p< 0.005, t test). In both conditions the potential remained negative at the end of the analysis period @ < 0.025, t test). There was a marked decrease in the amplitudes of SP from the first to the second response in all but one subject, and thereafter the amplitudes remained at a stable level. The amplitude of the SP (expected) of the averaged second to fourth response was 32 Z!I 17% of that of the first response and that of SP (ignored) 25 f 229& respectively. In both cases the decrease in the amplitude was significant @ < 0.005, t test), but the rate of decrease was not significantly different. The percentage decrease of the SP amplitude was slightly smaller than that of the N120 component of the on response, but this difference was not statistically significant. The amplitudes of both expected and ignored SPs were also significantly negative during stimulus repetition 0, < 0.005, t test). The amplitude of SP (expected) was
303
Fig. 4. Off response N120-P200 peak-to-peak amplitude
e---0 0---c
L
, 1.
I
2.
expecting Ignoring
I
I
3.
4.
stimulus The mean peak-to-peak (N120-P200) amplitude of the oft response at consecutive stimuli of the stimulus train when the subject either expected or ignored the stimuli. The bars show the standard error of the mean.
slightly larger than that of SP (ignored) during repetition of the stimuli (mean difference 0.9 pV, p < 0.05). At the first response no significant difference in the amplitudes of SPs between Fz and Cz was seen. During stimulus repetition, however, the SP amplitude was larger at Fz than at Cz (mean difference 1.2 I.IV, p < 0.05,t test).
Discussion The present work confirms the marked effect of stimulus repetition on both transient and sustained potentials elicited by auditory stimuli (Fruhstorfer et al., 1970; J&ilehto et al., 1978). The effect of expectation of the stimuli on the potentials was, however, more complex. The transient potentials were larger in amplitude when the subject expected the stimuli whereas the sustained potentials were unchanged at the first stimulus of the train or smaller when the subject expected the stimuli. During the stimulus repetition, these potentials also increased in amplitude with expectation. We have suggested earlier (Jgrvilehto et al., 1978) that with infrequent stimulation, both the transient and sustained potentials could reflect activity in neural systems that control phasic and tonic excitability in the motor and sensory system and prepare the organism for immediate action. The dissociation between the effects of expectation on the transient and sustained potentials at the first stimulus of the train indicates that the expectation of the stimuli has different effects on these two postulated systems. The remaining negativity at the end of the analysis period at the first stimulus of the train
304 Fig. 5. N120 and P200 latency for on and off responses
3----o
ON-response OFF-response
p2 p2
150-
100
Nl “1
I).‘r;-,l
t
1 1.
2.
3.
4.
The mean peak latencies of the negative (N120) and positive (P200) deflections of the transient on and off responses at consecutive stimuli of the stimulus train. Values for expecting and ignoring condition are combined. The bars show the standard error of the mean.
indicated that the “orienting” component of the sustained potential (see Introduction) is not restricted to the duration of the stimulus. The more specific “sensory” component of the sustained potential remaining during repetition of stimuli is smaller in amplitude than the “orienting” component, and it seems to be masked by it. Thus it can be discerned only when the nonspecific component is absent, i.e., when fast stimulus repetition rates are used. When attention is paid to the stimuli throughout their duration (Picton et al., 197&z) or when the stimuli are attentively expected, the amplitude of the specific component increases. The amplitude of the specific component is also larger when stimuli of low rather than of high tonal frequency are used (Picton et al., 1978b). Altogether, the specific component seems to be closely dependent on the stimulation characteristics used (David et al., 1969; Keidel, 1976; Picton et al., 1978b). A novel auditory stimulus may elicit phasic negative shifts in the human EEG that temporally coincide with the N120 component of the transient on response and summate on it (Naatanen and Michie, 1979). The marked amplitude reduction of the N120 component during stimulus repetition may be due to the disappearance of this
305 Fig. 6. Sustained
potential
amplitude
for each stimulus
SUSTAINED
position
POTENTIAL
[@I I
-8
I
7’
1
expecting ignoring
1 -6
I
7
1.
2.
I
I
3.
4.
stimulus The mean amplitude of the sustained potential at consecutive stimuli of the stimulus train when subjects either expected or ignored the stimuli. The bars show the standard error of the mean.
component. The remaining transient potential exhibits many features (frequency and intensity dependence, for example) similar to the specific sustained potential (Picton et al., 19786). Thus it seems that the transient evoked potential may be composed of two functionally different components, just as appears to be true of the sustained potential. The decrease of the amplitude of the N120 component of the on response with stimulus repetition was similar to that earlier reported with click stimuli (Fruhstorfer et al., 1970). However, only a trend toward smaller amplitudes of P200 with stimulus repetition was seen, whereas Fruhstorfer et al. (1970) found that P200 also showed a clear decrease in amplitude with stimulus repetition. This difference may be due to the longer stimuli and longer time constant recordings used in the present study. According to Picton et al. (1978~) the auditory evoked sustained potential begins 150 ms after stimulus onset and thus during the P200 component. A slow surface negative shift may then summate on P200 and shift it to a negative direction. This effect will be most prominent in cases where the sustained potential is large-at the first stimulus of the train in the present study. Variable opinions exist about the relationship between the on and off responses (Johannsen et al., 1972; Klingenberg-Schweitzer and Tepas, 1974; Onishi and Davis, 1968). Both the interstimulus interval and the duration of the stimulus are important for this relationship, because the interstimulus interval determines the degree of decrease of the amplitude of the on response, and the duration of the stimulus
306 determines the additional decrease of the off response in relation to the on response. In the present study where the interstimulus interval within the train was the same as the duration of the stimuli, the off responses were in many respects (form, amplitude, latency) similar to the P200 components of the on responses recorded during stimulus repetition. This indicates that the off responses reflect similar neural events as the P200 component of the on response. References David, E., Finkenzellx, P., Kallert, S., and Keidel, W.D. Akustischen Reizen zugeordnete GleichspannungGnderungen am intakten Schadel des Menschen. Pfliigers Archiv; European
Journal
Fruhstorfer,
of Physiology,
309, 362 (1969).
of the auditory evoked response in man. Electroencephalography and Clinical Neurophysiology, 28, 153 (1970). Hari, R., Sams, M., and Jlrvilehto, T. Effects of small ethanol doses on the auditory evoked transient and sustained potentials in the human EEG. In: Begleiter, H., ed. Biological Effects of Alcohol. Plenum Press, New York (in press). JPrvilehto, T., Hari, R., and Sams, M. Effect of stimulus repetition on negative sustained potentials elicited by auditory and visual stimuli in the human EEG. Biological Psychology, 7,
H., Soveri, P., and Jarvilehto, T. Short-term habituation
1 (1978).
Johannsen, H.S., Keidel, W.D., and Spreng, M. Der Einfluss von Intensitat und Dauer der Beschallung auf den Off-Effect des akustisch evozierten Potentials. Archivfur Klinische und Experimentelle Ohren-, Nasen- und Kehlkopfheilkunde, 201, 208 ( 1972). Keidel, W.D. The physiological background of the electric response audiometry. In: Keidel, W.D., and Neff, W.D., eds. Handbook of Sensory Physiology: Auditory System. Vol. V. Springer, Berlin (1976). Klingenberg-Schweitzer, P., and Tepas, D.1. Intensity effects of auditory evoked brain response to the stimulus onset and cessation. Perception and Psychophysics, 16, 396 (1974). Nsi’riMnen, R., and Michie, P.T. Different variants of endogenous negative brain potentials in performance situations: A review and classification. In: Lehman, D., and Callaway, E., eds. Human Evoked Potentials. Plenum Publishing Corporation, New York (1979). Onishi, S., and Davis, H. Effects of duration and rise time of tone burst on evoked V potentials. Journal
of the Acoustical
Society
of America,
44, 582 (1968).
Picton, T.W., Woods, D.L., and Proulx, G.B. Human auditory sustained potentials: I. The nature of the response. Electroencephalography and Clinical Neurophysiologv, 45, 186 (1978a).
Picton, T.W., Woods, D.L., and Proulx, G.B. Human auditory sustained potentials: Stimulus relationships. Electroencephalography and Clinical Neurophysiology, 45, (19786).
II. 198