- Email: [email protected]
Negative correlation of P50 peak latencies and reaction times in a simple reaction task
ELSEVIER
Electroencephalographyand clinical Neurophysiology 100 (1996) 74-77
o
Negative correlation of P50 peak latencies and reaction times in a simple reaction task Hideaki Ninomiya a,* Chung-Ho Chen a, Toshiaki Onitsuka b Atsushi Ichimiya a Department ofNeuropsychiatry, Faculty ofMedicine, Kyushu University, 3-1-1 Maidashi, Higashi-ku, Fukuoka, Japan b lpponmatsu Hospital, 142 Natsuyoshi, Tagawa City, Fukuoka, Japan
Accepted for publication: 12 July 1995
Abstract We investigated the influence of stimulus characteristics and tasks (count and simple reaction tasks) on auditory P50. Ten normal volunteers served as subjects. EEGs and auditory evoked potentials were obtained from Fz, Cz, Pz, C3, C4, T3 and T4 referred to linked earlobes (LE) and balanced non-cephalic (BN) electrodes. In the recordings using a BN reference there were significant negative correlations between the reaction times and P50 peak latencies at Fz, Cz, Pz, C3 and C4. In the LE reference recordings there were negative correlations but these did not reach statistical significance. The results are discussed with reference to the activation of the LE reference and the neural sources of P50. The findings suggest the possibility that reaction times are determined at the very early stage of auditory information processing. Keywords: Auditory information processing; Middle latency auditory evoked potential; P50; Reaction time
I. Introduction A positive wave at around 50 msec (P50) elicited by auditory stimuli has been used as an indicator of sensory gating in the field of psychiatric research (Johnson, 1985). Adler et al. (1982) reported that in normal subjects the amplitude of P50 in the second stimulus of a pair was suppressed when the interval between the two stimuli was regular, whereas in schizophrenics the suppression did not occur. They claimed that the deficit of the suppression in schizophrenics indicated a defect in an unconscious and hardwired sensory gating system, which caused sensory flooding in schizophrenics and resulted in thought disorders. On the other hand, Kathmann and Engel (1990) reported that they did not find strong P50 suppression even in normal subjects. Judd et al. (1992) made an objection to the results of Kathmann and Engel and claimed that the main cause of the differing results was the total power of the stimuli used. Linked earlobe electrodes were used as the reference site in these studies.
* Corresponding author. Tel: 92 641 1151, ext. 2537; Fax: 92 632 3558.
Although Picton et al. (1974) defined middle latency auditory evoked potentials ( M L A E P s ) to occur from PO (12 msec) to Nb (36 msec), some studies include P50 (P1) in the definition ( W o o d and Wolpaw, 1982; Erwin and Buchwald, 1986; Cacace et al., 1990). The neural sources of M L A E P s have been uncertain (Cacace et al., 1990), although adoption of chronic subdural electrodes and topographic analysis (Lee et al., 1984; Cacace et al., 1990), selection of an adequate reference (Wolpaw and Wood, 1982), and the recent development of a dipole trace method in magnetoencephalography (Pelizzone et al., 1987) have all been instrumental in revealing the neural sources of the M L A E P components. In previous studies, stimulus variables and reference electrode sites have been examined, but tasks manipulating psychological state have received little attention. Jerger et al. (1992) investigated the influence of alterations in attention on P50 amplitude in the paired-click c o n d i t i o n i n g / testing paradigm, using linked earlobe electrodes as a reference and reported that the manipulations affected N1 but not P50 amplitude. The reference electrode is an important factor in determining the wave shape of M L A E P s ( W o l p a w and Wood, 1982). Boutros et al. (1993) reported that paranoid and non-
0168-5597/96/$15.00 © 1996 Elsevier Science Ireland Ltd. Ail rights reserved SSDI 0013-4694(95)00160-3
EEP 94205
H. Ninomiya et al. / Electroencephalography and clinical Neurophysiology 1 O0 (1996) 74-77
paranoid schizophrenics showed different patterns of P50 suppression, and Schwender et al. (1994) reported that the latency and amplitude of an MLAEP component (Pa) were related to explicit and implicit memory in patients undergoing cardiac surgery. These reports suggest potential applications of MLAEPs in clinical use. We investigated the influence of stimulus characteristics and tasks (count and simple reaction tasks) on P50 using both linked earlobes and balanced non-cephalic electrodes as references. We found a negative correlation between P50 latency and simple reaction time. In this paper we report the findings briefly and discuss their significance.
2. Materiai and methods Ten right-handed healthy volunteers (age: 26-36 years, mean age: 27.6) served as subjects. Informed consent was obtained from each subject. A count task and a simple reaction task were included. Subjects were instructed to count silently the number of auditory stimuli in the count task, and to push the button in the right hand with their thumb as soon as possible to every auditory stimulus in the simple reaction task. In each task EEGs were recorded referred to linked earlobe (LE) and balanced non-cephalic (BN) electrodes separately. Thus 4 runs in all were carried out for each subject (count/BN, count/LE, simple reaction/BN, simple reaction/ LE), and the order of the runs was counterbalanced. An auditory stimulus was a t o n e burst (duration: 50 msec, fise and fall time: 5 msec, 70 dB SPL), and the stimuli were presented through a plastic tube (5 cm) and ear pieces to the ears with 1-3 sec interstimulus intervals (mean interval: 2 sec). Ag/AgC1 cup electrodes were placed at Fz, Cz, Pz, C3, C4, T3 and T4, according to the international 10-20 system. Electrode resistances were all below 5 k/2. Electrodes were fixed above and below the right eye, and on the lateral canthi to monitor vertical and horizontal eye movements. The bandpass filter was 0.5-500 Hz. For each condition, 100 artefact-free trials were averaged with respect to the onset of the auditory stimulus. The analysis window of 512 msec (1024 sampling points) included 32 msec pre stimulus and 480 msec post stimlus. The latency of the most positive voltage between 25 and 75 msec was defined as P50 peak latency.
3. Results The group mean reaction time was 221 + 34 msec in the LE reference and 215 _ 27 msec in the BN reference conditions. Pearson's correlation coefficient (CC) between reaction times in the LE and BN reference conditions were 0.851 (n = 10, P < 0.01). The number of the pairs is 10 for ail CCs reported in the text. In the BN reference condition there were significant
••
70.
60 . ~ , ~
~lDe
C3/LE
so.l "%~=;~" , 70. °
so
~"--...
•
75
,
C4/LE
•
Q
~ I
=- ,
C3/BN =- .
~
•
,
C4/BN =- .
o C
40
~
30 20 70.
60.l 50.
~ o • r=-0.026 Cz/LE __ ~ , ~
ee• •
40
•
~
"-.~
•
3O 20 100
200
CzJBN r=-0.704
300 4;0100 200 Reaction time(msec)
300
4;0
Fig. 1. Scattergrams of simple reaction times and P50 peak latencies at C3, C4 and Cz. BN signifies balanced non-cephalic electrodes used as reference and LE indicates linked earlobe references. Small r identifies Pearson's correlation coefficient between the reaction times and P50 peak latencies. Solid lines in the scattergram are regression lines.
negative correlations at Fz ( r = - 0 . 7 8 3 , P < 0.01), Cz (r = - 0 . 7 0 4 , P < 0.05), Pz (r = -0.645, P < 0.05), C3 ( r = - 0 . 6 5 1 , P <0.05), C4 ( r = -0.675, P <0.05). In the LE reference condition there were negative correlations but the CCs did not reach statistical significance. Fig. 1 shows scattergrams of P50 peak latencies and reaction times at Cz, C3 and C4. No significant correlation was found between the reaction times and the subjects' ages. Fig. 2 shows grand average auditory evoked potentials (AEPs) in both BN and LE reference conditions at Cz, C3, C4, T3 and T4 in the simple reaction task, and those at Cz
[ Cz
:* T3
c3,,
c~~ \ \
i~~v
100msec
P50
Fig. 2. Grand average auditory evoked potentials (AEPs). Thick line signifies AEPs in the BN reference condition and thin line shows AEPs in the LE condition. The left column and the upper and middle rows of the right column show AEPs in the simple reaction task. The bottom row of the right column shows AEPs in the count task. Dotted lines indicate the onset of the auditory stimulus. Asterisks indicate a negative wave at around 45 msec. Negativity is plotted upward.
H. Ninomiya et al. / Electroencephalography and clinical Neurophysiology 100 (1996) 74-77
76
fell on the border ( 2 5 / 7 5 msec) of the criterion except a few cases in the count task, i.e., positive peaks could be identified within the period (e.g. Fig. 3).
4.2. Influence of the reference electrodes on P50 latency
e:
50msec
Fig. 3. Average AEPs of each subject at C3 in the simple reaction task with BN reference. Arrows show the onset of the auditory stimuli and filled circles signify the P50 peak. Negativity is plotted upward. If the wave ( * ) is taken as the P50 peak, the correlation coefficient (CC) between the reaction times and the P50 peak latencies at C3 is - 0 . 6 4 8 . The CC is still statistically significant.
(bottom of right colunm) in the count task. Smali negative waves at around 45 msec (indicated by asterisks) were seen in AEPs of the BN reference condition at T3 and T4 but not in those of the LE reference condition. P50 peaks were clearly seen in the AEPs of the count task, but those in the simple reaction task were not well defined, especially in the BN reference condition. Fig. 3 shows the peak latencies of average AEPs at C3 for each subject in the simple reaction task with BN reference. The peaks show a widespread distribution.
4. Discussion We measured P50 peak latencies of AEPs elicited during a count task and a simple reaction task using both linked earlobe and balanced non-cephalic electrodes as references. The high positive correlation for the reaction times in the BN reference condition and those in the LE condition (r = 0.851) suggests that subjects performed at the same level in both tasks. Several issues are discussed below.
4.1. Identification of P50 Previous studies on P50 suppression defined peak latency as the most positive voltage between 25 and 75 msec (Judd et al., 1992) and we adopted the same criterion. Although Jerger et al. (1992) described an N100 onset occurring at P50 latency, sometimes causing the entire P50 to be negative, no P50 peak latencies in the present study
A reference electrode site is a crucial issue in measuring evoked potentials. AEPs tend to be influenced by activity recorded at the ear reference sites, because the neural sources of AEPs are close to the ears. As shown in Fig. 2, small negative waves indicated by asterisks were found at around 45 msec at T3 and T4 in the BN reference condition, but no clear negativity was seen in the LE condition. The small negative waves might correspond to the negative waves reported by Wood and Wolpaw (1982), which occurred at 3 2 / 5 7 msec and were dominant in the posterior temporal area. Volume conduction of the negative waves might inteffere with the real potentials of the recording electrodes for the LE reference and thus result in no significant correlation between the reaction times and the P50 peak latencies.
4.3. Was the component correlated with the reaction time really P50? As shown in Figs. 1 and 3, P50 peak latencies showed a relatively wide distribution in the BN reference condition, which might be one of the reasons that the grand average P50 peaks were not well defined, especially in the simple reaction task with the BN reference condition. In the simple reaction task, the simple reaction times and the P50 peak latencies showed significant negative correlation at Fz, Cz, Pz, C3 and C4 when EEGs were recorded with the BN reference. The most common components in the interval (25-75 msec) reported by studies on MLAEPs are Pa at around 30 msec and P50 (Pb) at around 50 msec. Erwin and Buchwald (1986) claimed that the neural source of P50 is an ascending reticular activating system. In this case our results indicate that subjects who responded quickly might activate the reticular system effectively with consistent timing, which would yield a clear P50, whereas those who responded slowly might not and the Pa would be taken as P50. The fact that the highest correlation was found at Fz might support this view. On the other hand, Wood and Wolpaw (1982) suggested that the neural source of a positive wave (FP55) was in the posterior part of the superior temporal plane, including the temporo-parietal junction. Taking the latency and source localization of the positive wave into consideration, the wave might reflect a potential occurring after the primary processing of auditory information. In this case, our results indicate that reaction times get faster if more time is taken processing auditory information in the primary auditory cortex. To put it figuratively, total cooking time gets shorter the longer the time taken for preparation.
H. Ninomiya et al./Electroencephalography and clinical Neurophysiology 100 (1996) 74-77
In our previous study using a G O / N O GO paradigm we found almost the same findings (Ninomiya and Kawasaki, 1991), i.e., reaction times showed a significant negative correlation with P50 peak latencies and a significant positive correlation with N2 peak latencies. In our previous experiment, however, 4 normals served as subjects and CCs were calculated on the basis of repeated measurements. Thus the findings could not be generalized to the general population. The results of the present study suggest the possibility that reaction times are regulated at the very early stage of auditory information processing. Further experiments are needed to confirm our findings and to clarify their nature.
Acknowledgements We express our thanks to Miss N. Kinugawa at the Department of Medical Informatics, Kyushu University, for her help with the statistics.
References Adler, L.E., Pachtman, E., Franks, R.D., Pecevich, M., Waldo, M.C. and Freedman, R. Neurophysiological evidence for a defect in neuronal mechanisms involved in sensory gating in schizophrenia. Biol. Psychiat., 1982, 17: 639-654. Boutros, N., Zouridakis, G., Rustin, T., Peabody, C. and Warner, D. The P50 component of the auditory evoked potential and subtypes of schizophrenia. Psychiat. Res., 1993, 47: 243-254. Cacace, A.T., Satya-Murti, S. and Wolpaw, J.R. Human middle-latency
77
auditory evoked potentials: vertex and temporal components. Electroenceph, clin. Neurophysiol., 1990, 77: 6-18. Erwin, R.J. and Buchwald, J.S. Midlatency auditory evoked responses: differential recovery cycle characteristics. Electroenceph. clin. Neurophysiol., 1986, 64: 417-423. Jerger, K., Biggins, C. and Fein, G. P50 suppression is not affected by attentional manipulation. Biol. Psychiat., 1992, 31: 365-377. Johnson, J.D. A mechanism to inhibit input activation and its dysfunction in schizophrenia. Br. J. Psychiat., 1985, 146: 429-435. Judd, L.L., McAdams, L., Budnick, B. and Braff, D.L. Sensory gating deficits in schizophrenia: new results. Am. J. Psychiat., 1992, 149: 488-493. Kathmann, N. and Engel, R.R. Sensory gating in normals and schizophrenics: a failure to find strong P50 suppression in normals. Biol. Psychiat., 1990, 21: 1216-1226. Lee, Y.S., Lueders, H., Dinner, D.S., Lesser, R.P., Hahn, J. and Klem, G. Recording of auditory evoked potentials in man using chronic subdural electrodes. Brain, 1984, 107:115-131. Ninomiya, H. and Kawasaki, H. Decomposition of sensory discrimination and motor response processes by analyzing the auditory evoked potentials. Jpn. J. EEG-EMG, 1991, 19: 32-39. (In Japanese.) Pelizzone, M., Hari, R., Mäkelä, J.P., Huttunen, J., Ahlfors, S. and Hämäläinen, M. Cortical origin of middle-latency auditory evoked responses in man. Neurosci. Lett., 1987, 82: 303-307. Picton, T.W., Hillyard, S.A., Krausz, H.I. and Galambos, R. Human auditory evoked potentials. I. Evaluation of components. Electroenceph. clin. Neurophysiol., 1974, 36: 179-190. Schwender, D., Kaiser, A., Klasing, S., Peter, K. and Poppel, E. Midlatency auditory evoked potentials and explicit and implicit memory in patients undergoing cardiac surgery. Anesthesiology, 1994, 80: 493501. Wolpaw, J.R. and Wood, C.C. Scalp distribution of human auditory evoked potentials. I. Evaluation of reference electrode sites. Electroenceph, clin. Neurophysiol., 1982, 54: 15-24. Wood, C.C. and Wolpaw, J.R. Scalp distribution of human auditory evoked potentials. II. Evidence for overlapping sources and involvement of auditory cortex. Electroenceph. clin. Neurophysiol., 1982, 54: 25-38.
Electroencephalographyand clinical Neurophysiology 100 (1996) 74-77
o
Negative correlation of P50 peak latencies and reaction times in a simple reaction task Hideaki Ninomiya a,* Chung-Ho Chen a, Toshiaki Onitsuka b Atsushi Ichimiya a Department ofNeuropsychiatry, Faculty ofMedicine, Kyushu University, 3-1-1 Maidashi, Higashi-ku, Fukuoka, Japan b lpponmatsu Hospital, 142 Natsuyoshi, Tagawa City, Fukuoka, Japan
Accepted for publication: 12 July 1995
Abstract We investigated the influence of stimulus characteristics and tasks (count and simple reaction tasks) on auditory P50. Ten normal volunteers served as subjects. EEGs and auditory evoked potentials were obtained from Fz, Cz, Pz, C3, C4, T3 and T4 referred to linked earlobes (LE) and balanced non-cephalic (BN) electrodes. In the recordings using a BN reference there were significant negative correlations between the reaction times and P50 peak latencies at Fz, Cz, Pz, C3 and C4. In the LE reference recordings there were negative correlations but these did not reach statistical significance. The results are discussed with reference to the activation of the LE reference and the neural sources of P50. The findings suggest the possibility that reaction times are determined at the very early stage of auditory information processing. Keywords: Auditory information processing; Middle latency auditory evoked potential; P50; Reaction time
I. Introduction A positive wave at around 50 msec (P50) elicited by auditory stimuli has been used as an indicator of sensory gating in the field of psychiatric research (Johnson, 1985). Adler et al. (1982) reported that in normal subjects the amplitude of P50 in the second stimulus of a pair was suppressed when the interval between the two stimuli was regular, whereas in schizophrenics the suppression did not occur. They claimed that the deficit of the suppression in schizophrenics indicated a defect in an unconscious and hardwired sensory gating system, which caused sensory flooding in schizophrenics and resulted in thought disorders. On the other hand, Kathmann and Engel (1990) reported that they did not find strong P50 suppression even in normal subjects. Judd et al. (1992) made an objection to the results of Kathmann and Engel and claimed that the main cause of the differing results was the total power of the stimuli used. Linked earlobe electrodes were used as the reference site in these studies.
* Corresponding author. Tel: 92 641 1151, ext. 2537; Fax: 92 632 3558.
Although Picton et al. (1974) defined middle latency auditory evoked potentials ( M L A E P s ) to occur from PO (12 msec) to Nb (36 msec), some studies include P50 (P1) in the definition ( W o o d and Wolpaw, 1982; Erwin and Buchwald, 1986; Cacace et al., 1990). The neural sources of M L A E P s have been uncertain (Cacace et al., 1990), although adoption of chronic subdural electrodes and topographic analysis (Lee et al., 1984; Cacace et al., 1990), selection of an adequate reference (Wolpaw and Wood, 1982), and the recent development of a dipole trace method in magnetoencephalography (Pelizzone et al., 1987) have all been instrumental in revealing the neural sources of the M L A E P components. In previous studies, stimulus variables and reference electrode sites have been examined, but tasks manipulating psychological state have received little attention. Jerger et al. (1992) investigated the influence of alterations in attention on P50 amplitude in the paired-click c o n d i t i o n i n g / testing paradigm, using linked earlobe electrodes as a reference and reported that the manipulations affected N1 but not P50 amplitude. The reference electrode is an important factor in determining the wave shape of M L A E P s ( W o l p a w and Wood, 1982). Boutros et al. (1993) reported that paranoid and non-
0168-5597/96/$15.00 © 1996 Elsevier Science Ireland Ltd. Ail rights reserved SSDI 0013-4694(95)00160-3
EEP 94205
H. Ninomiya et al. / Electroencephalography and clinical Neurophysiology 1 O0 (1996) 74-77
paranoid schizophrenics showed different patterns of P50 suppression, and Schwender et al. (1994) reported that the latency and amplitude of an MLAEP component (Pa) were related to explicit and implicit memory in patients undergoing cardiac surgery. These reports suggest potential applications of MLAEPs in clinical use. We investigated the influence of stimulus characteristics and tasks (count and simple reaction tasks) on P50 using both linked earlobes and balanced non-cephalic electrodes as references. We found a negative correlation between P50 latency and simple reaction time. In this paper we report the findings briefly and discuss their significance.
2. Materiai and methods Ten right-handed healthy volunteers (age: 26-36 years, mean age: 27.6) served as subjects. Informed consent was obtained from each subject. A count task and a simple reaction task were included. Subjects were instructed to count silently the number of auditory stimuli in the count task, and to push the button in the right hand with their thumb as soon as possible to every auditory stimulus in the simple reaction task. In each task EEGs were recorded referred to linked earlobe (LE) and balanced non-cephalic (BN) electrodes separately. Thus 4 runs in all were carried out for each subject (count/BN, count/LE, simple reaction/BN, simple reaction/ LE), and the order of the runs was counterbalanced. An auditory stimulus was a t o n e burst (duration: 50 msec, fise and fall time: 5 msec, 70 dB SPL), and the stimuli were presented through a plastic tube (5 cm) and ear pieces to the ears with 1-3 sec interstimulus intervals (mean interval: 2 sec). Ag/AgC1 cup electrodes were placed at Fz, Cz, Pz, C3, C4, T3 and T4, according to the international 10-20 system. Electrode resistances were all below 5 k/2. Electrodes were fixed above and below the right eye, and on the lateral canthi to monitor vertical and horizontal eye movements. The bandpass filter was 0.5-500 Hz. For each condition, 100 artefact-free trials were averaged with respect to the onset of the auditory stimulus. The analysis window of 512 msec (1024 sampling points) included 32 msec pre stimulus and 480 msec post stimlus. The latency of the most positive voltage between 25 and 75 msec was defined as P50 peak latency.
3. Results The group mean reaction time was 221 + 34 msec in the LE reference and 215 _ 27 msec in the BN reference conditions. Pearson's correlation coefficient (CC) between reaction times in the LE and BN reference conditions were 0.851 (n = 10, P < 0.01). The number of the pairs is 10 for ail CCs reported in the text. In the BN reference condition there were significant
••
70.
60 . ~ , ~
~lDe
C3/LE
so.l "%~=;~" , 70. °
so
~"--...
•
75
,
C4/LE
•
Q
~ I
=- ,
C3/BN =- .
~
•
,
C4/BN =- .
o C
40
~
30 20 70.
60.l 50.
~ o • r=-0.026 Cz/LE __ ~ , ~
ee• •
40
•
~
"-.~
•
3O 20 100
200
CzJBN r=-0.704
300 4;0100 200 Reaction time(msec)
300
4;0
Fig. 1. Scattergrams of simple reaction times and P50 peak latencies at C3, C4 and Cz. BN signifies balanced non-cephalic electrodes used as reference and LE indicates linked earlobe references. Small r identifies Pearson's correlation coefficient between the reaction times and P50 peak latencies. Solid lines in the scattergram are regression lines.
negative correlations at Fz ( r = - 0 . 7 8 3 , P < 0.01), Cz (r = - 0 . 7 0 4 , P < 0.05), Pz (r = -0.645, P < 0.05), C3 ( r = - 0 . 6 5 1 , P <0.05), C4 ( r = -0.675, P <0.05). In the LE reference condition there were negative correlations but the CCs did not reach statistical significance. Fig. 1 shows scattergrams of P50 peak latencies and reaction times at Cz, C3 and C4. No significant correlation was found between the reaction times and the subjects' ages. Fig. 2 shows grand average auditory evoked potentials (AEPs) in both BN and LE reference conditions at Cz, C3, C4, T3 and T4 in the simple reaction task, and those at Cz
[ Cz
:* T3
c3,,
c~~ \ \
i~~v
100msec
P50
Fig. 2. Grand average auditory evoked potentials (AEPs). Thick line signifies AEPs in the BN reference condition and thin line shows AEPs in the LE condition. The left column and the upper and middle rows of the right column show AEPs in the simple reaction task. The bottom row of the right column shows AEPs in the count task. Dotted lines indicate the onset of the auditory stimulus. Asterisks indicate a negative wave at around 45 msec. Negativity is plotted upward.
H. Ninomiya et al. / Electroencephalography and clinical Neurophysiology 100 (1996) 74-77
76
fell on the border ( 2 5 / 7 5 msec) of the criterion except a few cases in the count task, i.e., positive peaks could be identified within the period (e.g. Fig. 3).
4.2. Influence of the reference electrodes on P50 latency
e:
50msec
Fig. 3. Average AEPs of each subject at C3 in the simple reaction task with BN reference. Arrows show the onset of the auditory stimuli and filled circles signify the P50 peak. Negativity is plotted upward. If the wave ( * ) is taken as the P50 peak, the correlation coefficient (CC) between the reaction times and the P50 peak latencies at C3 is - 0 . 6 4 8 . The CC is still statistically significant.
(bottom of right colunm) in the count task. Smali negative waves at around 45 msec (indicated by asterisks) were seen in AEPs of the BN reference condition at T3 and T4 but not in those of the LE reference condition. P50 peaks were clearly seen in the AEPs of the count task, but those in the simple reaction task were not well defined, especially in the BN reference condition. Fig. 3 shows the peak latencies of average AEPs at C3 for each subject in the simple reaction task with BN reference. The peaks show a widespread distribution.
4. Discussion We measured P50 peak latencies of AEPs elicited during a count task and a simple reaction task using both linked earlobe and balanced non-cephalic electrodes as references. The high positive correlation for the reaction times in the BN reference condition and those in the LE condition (r = 0.851) suggests that subjects performed at the same level in both tasks. Several issues are discussed below.
4.1. Identification of P50 Previous studies on P50 suppression defined peak latency as the most positive voltage between 25 and 75 msec (Judd et al., 1992) and we adopted the same criterion. Although Jerger et al. (1992) described an N100 onset occurring at P50 latency, sometimes causing the entire P50 to be negative, no P50 peak latencies in the present study
A reference electrode site is a crucial issue in measuring evoked potentials. AEPs tend to be influenced by activity recorded at the ear reference sites, because the neural sources of AEPs are close to the ears. As shown in Fig. 2, small negative waves indicated by asterisks were found at around 45 msec at T3 and T4 in the BN reference condition, but no clear negativity was seen in the LE condition. The small negative waves might correspond to the negative waves reported by Wood and Wolpaw (1982), which occurred at 3 2 / 5 7 msec and were dominant in the posterior temporal area. Volume conduction of the negative waves might inteffere with the real potentials of the recording electrodes for the LE reference and thus result in no significant correlation between the reaction times and the P50 peak latencies.
4.3. Was the component correlated with the reaction time really P50? As shown in Figs. 1 and 3, P50 peak latencies showed a relatively wide distribution in the BN reference condition, which might be one of the reasons that the grand average P50 peaks were not well defined, especially in the simple reaction task with the BN reference condition. In the simple reaction task, the simple reaction times and the P50 peak latencies showed significant negative correlation at Fz, Cz, Pz, C3 and C4 when EEGs were recorded with the BN reference. The most common components in the interval (25-75 msec) reported by studies on MLAEPs are Pa at around 30 msec and P50 (Pb) at around 50 msec. Erwin and Buchwald (1986) claimed that the neural source of P50 is an ascending reticular activating system. In this case our results indicate that subjects who responded quickly might activate the reticular system effectively with consistent timing, which would yield a clear P50, whereas those who responded slowly might not and the Pa would be taken as P50. The fact that the highest correlation was found at Fz might support this view. On the other hand, Wood and Wolpaw (1982) suggested that the neural source of a positive wave (FP55) was in the posterior part of the superior temporal plane, including the temporo-parietal junction. Taking the latency and source localization of the positive wave into consideration, the wave might reflect a potential occurring after the primary processing of auditory information. In this case, our results indicate that reaction times get faster if more time is taken processing auditory information in the primary auditory cortex. To put it figuratively, total cooking time gets shorter the longer the time taken for preparation.
H. Ninomiya et al./Electroencephalography and clinical Neurophysiology 100 (1996) 74-77
In our previous study using a G O / N O GO paradigm we found almost the same findings (Ninomiya and Kawasaki, 1991), i.e., reaction times showed a significant negative correlation with P50 peak latencies and a significant positive correlation with N2 peak latencies. In our previous experiment, however, 4 normals served as subjects and CCs were calculated on the basis of repeated measurements. Thus the findings could not be generalized to the general population. The results of the present study suggest the possibility that reaction times are regulated at the very early stage of auditory information processing. Further experiments are needed to confirm our findings and to clarify their nature.
Acknowledgements We express our thanks to Miss N. Kinugawa at the Department of Medical Informatics, Kyushu University, for her help with the statistics.
References Adler, L.E., Pachtman, E., Franks, R.D., Pecevich, M., Waldo, M.C. and Freedman, R. Neurophysiological evidence for a defect in neuronal mechanisms involved in sensory gating in schizophrenia. Biol. Psychiat., 1982, 17: 639-654. Boutros, N., Zouridakis, G., Rustin, T., Peabody, C. and Warner, D. The P50 component of the auditory evoked potential and subtypes of schizophrenia. Psychiat. Res., 1993, 47: 243-254. Cacace, A.T., Satya-Murti, S. and Wolpaw, J.R. Human middle-latency
77
auditory evoked potentials: vertex and temporal components. Electroenceph, clin. Neurophysiol., 1990, 77: 6-18. Erwin, R.J. and Buchwald, J.S. Midlatency auditory evoked responses: differential recovery cycle characteristics. Electroenceph. clin. Neurophysiol., 1986, 64: 417-423. Jerger, K., Biggins, C. and Fein, G. P50 suppression is not affected by attentional manipulation. Biol. Psychiat., 1992, 31: 365-377. Johnson, J.D. A mechanism to inhibit input activation and its dysfunction in schizophrenia. Br. J. Psychiat., 1985, 146: 429-435. Judd, L.L., McAdams, L., Budnick, B. and Braff, D.L. Sensory gating deficits in schizophrenia: new results. Am. J. Psychiat., 1992, 149: 488-493. Kathmann, N. and Engel, R.R. Sensory gating in normals and schizophrenics: a failure to find strong P50 suppression in normals. Biol. Psychiat., 1990, 21: 1216-1226. Lee, Y.S., Lueders, H., Dinner, D.S., Lesser, R.P., Hahn, J. and Klem, G. Recording of auditory evoked potentials in man using chronic subdural electrodes. Brain, 1984, 107:115-131. Ninomiya, H. and Kawasaki, H. Decomposition of sensory discrimination and motor response processes by analyzing the auditory evoked potentials. Jpn. J. EEG-EMG, 1991, 19: 32-39. (In Japanese.) Pelizzone, M., Hari, R., Mäkelä, J.P., Huttunen, J., Ahlfors, S. and Hämäläinen, M. Cortical origin of middle-latency auditory evoked responses in man. Neurosci. Lett., 1987, 82: 303-307. Picton, T.W., Hillyard, S.A., Krausz, H.I. and Galambos, R. Human auditory evoked potentials. I. Evaluation of components. Electroenceph. clin. Neurophysiol., 1974, 36: 179-190. Schwender, D., Kaiser, A., Klasing, S., Peter, K. and Poppel, E. Midlatency auditory evoked potentials and explicit and implicit memory in patients undergoing cardiac surgery. Anesthesiology, 1994, 80: 493501. Wolpaw, J.R. and Wood, C.C. Scalp distribution of human auditory evoked potentials. I. Evaluation of reference electrode sites. Electroenceph, clin. Neurophysiol., 1982, 54: 15-24. Wood, C.C. and Wolpaw, J.R. Scalp distribution of human auditory evoked potentials. II. Evidence for overlapping sources and involvement of auditory cortex. Electroenceph. clin. Neurophysiol., 1982, 54: 25-38.