Sensory
Gating
Deficits
Lewis
L. Judd,
Byron
and
extend
results
schizophrenic patients. normal subjects (N=20) potentials
to assess
M.D.,
Budnick,
Objective: It has been widely occur in schizophrenic patients to confirm
in Schizophrenia: LouAnn
B.A.,
hypothesized compared ofearlier
and David
L. Braff,
M.D.
that sensory gating failures to normal subjects. The authors studies
that
showed
specific
and sensory overload ofthis study sought
sensory
gating
overall
competence
ofthe
subjects’
central
sensory
deficits
in
(N=20) and event-related
inhibitory
capacity
P50 area responses to two 75-dB (conditioning and averaged over 60 trials, were recorded for each click stimulus induces gating mechanisms that result
in diminished
potentials
(Am
or gated
P50
event-related
The schizophrenic subjects manifested and parietal electrode placement sites,
J
Psychiatry
in the frontal areas of of normal sensory gating
1992;
ince the time of Kraepelin and Bleuler, abnormalities of attention and information processing have been identified among patients with diagnoses in the schizophrenic spectrum of disorders. In recent years, investigators have used information-processing paradigms in studies of both humans and animals to specify the time course of information-processing deficits and the possible underlying neuroanatomical and neurophysiological abnormalities that evoke these deficits (1). Earlier, McGhie and Chapman (2) discussed “disorders of attention and perception in early schizophrenia” and noted that organisms integrate sensory data by utilizing a buffer capacity that allows for perceptual constancy by reduction of the “otherwise chaotic flow of information reaching consciousness.” This is very similar to Venables’ view (3) that in some schizophrenic individuals there is stimulus “flooding” through sensory overload based on a failure of normal filtering and
Aug. 1991.
San
30, 1990; revision received Aug. 26, 1991; accepted From the Department of Psychiatry, University of Diego.
Address
reprint
requests
in response
to the
a significant sensory with a nonsignificant
second
click
gating deficit tendency for
the brain. Conclusions: in schizophrenic patients.
These
data
stimulus. at frontal, the deficit reflect
a
149:488-493)
S
to Dr.
Judd,
Depart-
ment of Psychiatry (0603), UCSD School of Medicine, 9500 Gilman Dr., La Jolla, CA 92093. Supported in part by NIMH grant MH-42228 to Dr. Braff and by a grant from the State of California Department of Mental Hygiene to Dr. Judd. The authors thank Nancy Ostrowski, Ph.D., Mark Geyer, Ph.D., and Joyce Sprock for editorial suggestions and John Bartko, Ph.D., for assistance with statistical analysis and data presentation. Copyright © I 992 American Psychiatric Association.
488
Ph.D.,
Method: Ageand sex-matched schizophrenic patients were tested using electrophysiologically recorded P50
the
to be most prominent regionally diffuse loss
California,
McAdams,
Results
by measuring sensory filtering or gating. test) click stimuli of O.04-msec duration, subject. Normally, the first (conditioning) Results: central,
Received Sept. 27,
New
gating mechanisms. Still, theories of sensory gating abnormalities in schizophrenia are often confusing and difficult to operationalize. This led Neale and Cromwell (4) to call for “refined behavioral techniques utilizing relevant behavioral cues” in gating studies. In more recent years, a number of investigators have used innovative techniques to specify the sensory gating deficits and corresponding sensory overload and inundation that occur in some schizophrenic spectrum patients. Gottschalk et al. (5) used a novel paradigm to illustrate that sensory overload (or LSD treatment) could induce changes in normal subjects’ scores on social alienationpersonal disorganization and cognitive-intellectual impairment scales that are similar to the test results of schizophrenic patients. Specifically, Gottschalk et al. subjected normal control subjects to sensory overload by using an “overload apparatus” that consisted of a geodesic dome into which cacophonous sounds and fragmenting, bizarne, and complex visual images were projected. The resuits were consistent with sensory gating deficit theories, since cognitive fragmentation occurred when individuals experienced sensory overload. Other studies of sensory overload and sensory gating deficits in patients with schizophrenic disorders have been reported in which cortically based event-related potentials were used. Braff et al. (6) used a self-stimulation paradigm in which the subject uses a key release to initiate a stimulus that elicits event-related potentials. The P300 wave component of event-related potentials reflects the amount of novel or unanticipated infonma-
Am
J
Psychiatry
1 49:4,
April
1992
JUDD,
tion delivered by a stimulus. Thus, the P300 component is diminished when the subject has elicited the stimulus (in contrast to a randomly generated stimulus), and this diminution is greatest at brief intervals (e.g., 100 msec) between key release and stimulus onset, when the subject is aware that the stimulus is about to “arrive.” Schizophrenic subjects, however, show a paradoxically decreased P300 amplitude in the 250-S00-msec interval, indicating a diminution of the normal inhibitory mechanisms that operate in this time frame. In a related series of studies, Freedman, Adler, Siegel, and associates found that medicated (7) and unmedicated (8) schizophrenic patients, as well as relatives of schizophrenic patients (9), showed deficits in a sensory gating study in which a two-stimulus (conditioning and test) event-related potentials paradigm was used. Both of the stimuli used in this paradigm evoke P50 waves that occur in the epoch about SO msec after stimulus onset. The second stimulus usually evokes a P50 wave that is inhibited or gated by the effect of the first stimulus. Schizophrenic patients exhibit a lack of this normal inhibitory influence and thus show a larger than normal P50 wave elicited by the second stimulus that could reflect a deficit in sensory gating. This deficit is not an artifact of antipsychotic medications, since unmedicated schizophrenic patients and family members of schizophrenic patients show similar deficits. Kathmann and Engel (10) reported that this loss of P50 gating in schizophrenic patients was not reproduced in their study. It is noteworthy, however, that the click stimulus they used in their study was many times more powerful than that used by Freedman and associates because of its longer duration (1.5 msec versus 0.04 msec). In related work with animals, Adler et al. ( 1 1 ) proposed that these deficits are caused by a state of hyperarousal, with hyperexcitable neurons causing a defect in normal cortical and subcortical inhibitory mechanisms. The existence of this viable animal model allows for novel and enlightening studies of the mechanisms underlying gating failures in schizophrenic patients. Since their initial reports (7, 8), there has not been independent confirmation in another laboratory of Freedman and associates’ observations of sensory gating deficits in schizophrenic patients compared with normal or patient control subjects. The only report from another laboratory (10) failed to confirm Freedman and associates’ basic finding. Our study was designed to use the P50 event-related potentials paradigm to replicate and extend Freedman and associates’ findings of sensory gating deficits in schizophrenic patients.
METHOD The subjects were 20 medicated schizophrenic patients recruited from the inpatient psychiatry service of a university medical center. The patients were between 1 8 and 55 years of age and met the DSM-III criteria for schizophrenia. In addition, the patients were administered the Schedule for Affective Disorders and Schizo-
Am
J
Psychiatry
1 49:4,
April
1992
MCADAMS,
BUDNICK,
ET AL.
phnenia-Lifetime Version (SADS-L) (12), and patients meeting the Research Diagnostic Criteria (RDC) (13) for schizophrenia were included in the study. The Brief Psychiatric Rating Scale (14) and the Global Assessment Scale (15) were also administered to all patients. The mean age of the schizophrenic patients ( I S male and five female) was 28.8 years (SD=8.49, range=1853 years). Their average dose of antipsychotic medications was I 996 mg/day of chlorpromazine equivalents. Eighteen of the patients were right-handed and two were left-handed. Potential subjects who had current or recent histories of alcoholism, drug abuse, dementia, or ECT were excluded, as in our previous method (16). Twenty normal subjects (14 male and six female) with no active psychopathology or family history of schizophrenia were recruited from advertisements in local papers and from among the medical center staff. All were administered the SADS-L. Only subjects who were medically healthy and met the RDC for “never mentally ill” were included in the normal group. The mean age for this group was 27.9 years (SD=S.99, range=18-39 years). Seventeen of these subjects were right-handed and three were left-handed. Informed consent was obtained from all subjects. All recordings were performed in the electrophysiology laboratory ofone ofus (L.L.J.) in a shielded chamber with a DEC-MINC computer and Grass amplifiers. Recordings were made from gold cup electrodes (resistances less than S placed at the midline frontal, central, and parietal sites according to the I 0/20 system. Electrodes were referenced to linked ear electrodes. In general, the recording methods closely followed those published by Freedman and associates (7-9), so that the results would be comparable to their published data. One difference, however, was that we used 60 pairs of stimuli in these expeniments, whereas Adler et al. (7) used 32 pairs. Electrical activity was amplified 20,000 times by a 1-Me impedance amplifier with band pass filters (-3 dB) at 0 and 1500 Hz. The EEG was monitored on an oscilloscope, and acquisition of data was interrupted when voltage exceeded 100 j.tV or when there was evidence of artificial muscle activity. Data were digitized at a rate of 1000 Hz for on-line averaging, and simultaneous averages were recorded from lateral canthus electro-oculography (EOG) electrodes and monitored in order to exclude EOG contribution to the data. The artifact rejection was automated by computer program and preset to exclude any sweep that was larger than 90% of the A-to-D capacity for amplitude and that exceeded 15% of the entire sweep for duration. With these exclusion parameters it was felt that EOG artifact as well as other artifacts (e.g., swallowing, minor muscular movement) would be fully removed. Grass 5-88 stimulator-generated clicks of 0.04-msec duration were amplified within a band width of 20-10,000 Hz and delivered through earphones with a mean intensity of 75 dB, as measured by a sound meter at the earphone. Stimuli were presented as a series of paired conditioning and testing stimuli. As described by Freedman et al. (8), the interval between the pains of stimuli was kept
489
SENSORY FIGURE
GATING
DEFICITS
1. Representative
Responses
IN SCHIZOPHRENIA
Data for One Normal
to Two Click Stimuli
Averaged First
25
50
75
100
Subject
on P50 Area
Over 60 Trialsa
Click
125
Second
150
.,o
175
200
225
2
175
200
225
250
Click
0 0.
a) >
25
50
75
l0O
125
150
+
10 IJV 50 msec aRobust
sensory
of the P50 wave the firstclick.
gating
is indicated
following
by the large decrement
the second
click relative
to that
in the area following
constant at 10 seconds, and the interval between the conditioning and test pulses was kept constant at 500 msec in order to test the previously identified interval of maximal differences between schizophrenic and normal subjects in this paradigm. Sixty pairs of responses were averaged. The P50 response data were analyzed as follows. Averaged evoked potentials were measured for peak amplitude and area ofthe P50 wave. The P50 peak was defined as the maximal positive activity occurring between 25 and 75 msec after the stimulus and comesponds to peak amplitude data. Both amplitudes and areas were measured by two methods: 1) from prestimulus baseline to peak and 2) from trough to peak by using the segments of potential immediately pmeceding and following the P50 peak to construct the local baseline. This second method was used in order to avoid all possible problems caused by a small P50 peak riding on a large, long-lasting negative potential and those associated with using a short interval between conditioning and test stimuli (i.e., 500 msec). Analyses of peak amplitude and area measurements of absolute
490
and relative change scores yielded similar results. Therefore, only area data are reported. Three specific prospective hypotheses for P50 area responses were formulated. 1 ) A gating effect would be observed across both groups of subjects and all electrode positions and would be evidenced by a decrement in average P50 response to the second click relative to that of the conditioning first click stimulus. 2) If schizophrenic patients lack robust gating, then the overall decrement in the P50 wave from the response to the first click stimulus to the response to the second click stimulus would be larger in the normal group than in the schizophrenic group (the Diagnosis by Gating intenaction). 3) The hypothesized “hypofrontality” in schizophrenic subjects led to the prediction that regional differences between the two groups would be found. Specifically, the magnitude of the differences in P50 responses between normal and schizophrenic subjects would vary among electrode positions, being largest at the frontal electrode position and smallest at the panetal position (the Electrode Position by Diagnosis by Gating interaction). A three-way repeated measures multivaniate analysis of variance (MANOVA) was conducted using BMDP software (17, 18). The MANOVA included a factor for diagnosis (normal or schizophrenic subjects), a repeated measures factor for the stimulus presentations (stimulus 1, the conditioning click; and stimulus 2, the test click), and a factor for the repeated measures at the three electrode positions (frontal, central, and panietal). A square root transformation of P50 area data was used to preserve assumptions of homogeneity of variance and normality.
RESULTS Figures 1 and 2 show averaged P50 area responses to the first and second clicks for one representative subject from the normal group and one from the schizophrenic group, respectively. Primary
Analysis
A significant gating effect was observed in the P50 area response. Across both groups and all three electrode positions, the mean P50 area response to the experimental second click was significantly diminished from that of the conditioning first click (gating effect, F=176.98, df=1, 38, p
Gating
Deficits
Lewis
L. Judd,
Byron
and
extend
results
schizophrenic patients. normal subjects (N=20) potentials
to assess
M.D.,
Budnick,
Objective: It has been widely occur in schizophrenic patients to confirm
in Schizophrenia: LouAnn
B.A.,
hypothesized compared ofearlier
and David
L. Braff,
M.D.
that sensory gating failures to normal subjects. The authors studies
that
showed
specific
and sensory overload ofthis study sought
sensory
gating
overall
competence
ofthe
subjects’
central
sensory
deficits
in
(N=20) and event-related
inhibitory
capacity
P50 area responses to two 75-dB (conditioning and averaged over 60 trials, were recorded for each click stimulus induces gating mechanisms that result
in diminished
potentials
(Am
or gated
P50
event-related
The schizophrenic subjects manifested and parietal electrode placement sites,
J
Psychiatry
in the frontal areas of of normal sensory gating
1992;
ince the time of Kraepelin and Bleuler, abnormalities of attention and information processing have been identified among patients with diagnoses in the schizophrenic spectrum of disorders. In recent years, investigators have used information-processing paradigms in studies of both humans and animals to specify the time course of information-processing deficits and the possible underlying neuroanatomical and neurophysiological abnormalities that evoke these deficits (1). Earlier, McGhie and Chapman (2) discussed “disorders of attention and perception in early schizophrenia” and noted that organisms integrate sensory data by utilizing a buffer capacity that allows for perceptual constancy by reduction of the “otherwise chaotic flow of information reaching consciousness.” This is very similar to Venables’ view (3) that in some schizophrenic individuals there is stimulus “flooding” through sensory overload based on a failure of normal filtering and
Aug. 1991.
San
30, 1990; revision received Aug. 26, 1991; accepted From the Department of Psychiatry, University of Diego.
Address
reprint
requests
in response
to the
a significant sensory with a nonsignificant
second
click
gating deficit tendency for
the brain. Conclusions: in schizophrenic patients.
These
data
stimulus. at frontal, the deficit reflect
a
149:488-493)
S
to Dr.
Judd,
Depart-
ment of Psychiatry (0603), UCSD School of Medicine, 9500 Gilman Dr., La Jolla, CA 92093. Supported in part by NIMH grant MH-42228 to Dr. Braff and by a grant from the State of California Department of Mental Hygiene to Dr. Judd. The authors thank Nancy Ostrowski, Ph.D., Mark Geyer, Ph.D., and Joyce Sprock for editorial suggestions and John Bartko, Ph.D., for assistance with statistical analysis and data presentation. Copyright © I 992 American Psychiatric Association.
488
Ph.D.,
Method: Ageand sex-matched schizophrenic patients were tested using electrophysiologically recorded P50
the
to be most prominent regionally diffuse loss
California,
McAdams,
Results
by measuring sensory filtering or gating. test) click stimuli of O.04-msec duration, subject. Normally, the first (conditioning) Results: central,
Received Sept. 27,
New
gating mechanisms. Still, theories of sensory gating abnormalities in schizophrenia are often confusing and difficult to operationalize. This led Neale and Cromwell (4) to call for “refined behavioral techniques utilizing relevant behavioral cues” in gating studies. In more recent years, a number of investigators have used innovative techniques to specify the sensory gating deficits and corresponding sensory overload and inundation that occur in some schizophrenic spectrum patients. Gottschalk et al. (5) used a novel paradigm to illustrate that sensory overload (or LSD treatment) could induce changes in normal subjects’ scores on social alienationpersonal disorganization and cognitive-intellectual impairment scales that are similar to the test results of schizophrenic patients. Specifically, Gottschalk et al. subjected normal control subjects to sensory overload by using an “overload apparatus” that consisted of a geodesic dome into which cacophonous sounds and fragmenting, bizarne, and complex visual images were projected. The resuits were consistent with sensory gating deficit theories, since cognitive fragmentation occurred when individuals experienced sensory overload. Other studies of sensory overload and sensory gating deficits in patients with schizophrenic disorders have been reported in which cortically based event-related potentials were used. Braff et al. (6) used a self-stimulation paradigm in which the subject uses a key release to initiate a stimulus that elicits event-related potentials. The P300 wave component of event-related potentials reflects the amount of novel or unanticipated infonma-
Am
J
Psychiatry
1 49:4,
April
1992
JUDD,
tion delivered by a stimulus. Thus, the P300 component is diminished when the subject has elicited the stimulus (in contrast to a randomly generated stimulus), and this diminution is greatest at brief intervals (e.g., 100 msec) between key release and stimulus onset, when the subject is aware that the stimulus is about to “arrive.” Schizophrenic subjects, however, show a paradoxically decreased P300 amplitude in the 250-S00-msec interval, indicating a diminution of the normal inhibitory mechanisms that operate in this time frame. In a related series of studies, Freedman, Adler, Siegel, and associates found that medicated (7) and unmedicated (8) schizophrenic patients, as well as relatives of schizophrenic patients (9), showed deficits in a sensory gating study in which a two-stimulus (conditioning and test) event-related potentials paradigm was used. Both of the stimuli used in this paradigm evoke P50 waves that occur in the epoch about SO msec after stimulus onset. The second stimulus usually evokes a P50 wave that is inhibited or gated by the effect of the first stimulus. Schizophrenic patients exhibit a lack of this normal inhibitory influence and thus show a larger than normal P50 wave elicited by the second stimulus that could reflect a deficit in sensory gating. This deficit is not an artifact of antipsychotic medications, since unmedicated schizophrenic patients and family members of schizophrenic patients show similar deficits. Kathmann and Engel (10) reported that this loss of P50 gating in schizophrenic patients was not reproduced in their study. It is noteworthy, however, that the click stimulus they used in their study was many times more powerful than that used by Freedman and associates because of its longer duration (1.5 msec versus 0.04 msec). In related work with animals, Adler et al. ( 1 1 ) proposed that these deficits are caused by a state of hyperarousal, with hyperexcitable neurons causing a defect in normal cortical and subcortical inhibitory mechanisms. The existence of this viable animal model allows for novel and enlightening studies of the mechanisms underlying gating failures in schizophrenic patients. Since their initial reports (7, 8), there has not been independent confirmation in another laboratory of Freedman and associates’ observations of sensory gating deficits in schizophrenic patients compared with normal or patient control subjects. The only report from another laboratory (10) failed to confirm Freedman and associates’ basic finding. Our study was designed to use the P50 event-related potentials paradigm to replicate and extend Freedman and associates’ findings of sensory gating deficits in schizophrenic patients.
METHOD The subjects were 20 medicated schizophrenic patients recruited from the inpatient psychiatry service of a university medical center. The patients were between 1 8 and 55 years of age and met the DSM-III criteria for schizophrenia. In addition, the patients were administered the Schedule for Affective Disorders and Schizo-
Am
J
Psychiatry
1 49:4,
April
1992
MCADAMS,
BUDNICK,
ET AL.
phnenia-Lifetime Version (SADS-L) (12), and patients meeting the Research Diagnostic Criteria (RDC) (13) for schizophrenia were included in the study. The Brief Psychiatric Rating Scale (14) and the Global Assessment Scale (15) were also administered to all patients. The mean age of the schizophrenic patients ( I S male and five female) was 28.8 years (SD=8.49, range=1853 years). Their average dose of antipsychotic medications was I 996 mg/day of chlorpromazine equivalents. Eighteen of the patients were right-handed and two were left-handed. Potential subjects who had current or recent histories of alcoholism, drug abuse, dementia, or ECT were excluded, as in our previous method (16). Twenty normal subjects (14 male and six female) with no active psychopathology or family history of schizophrenia were recruited from advertisements in local papers and from among the medical center staff. All were administered the SADS-L. Only subjects who were medically healthy and met the RDC for “never mentally ill” were included in the normal group. The mean age for this group was 27.9 years (SD=S.99, range=18-39 years). Seventeen of these subjects were right-handed and three were left-handed. Informed consent was obtained from all subjects. All recordings were performed in the electrophysiology laboratory ofone ofus (L.L.J.) in a shielded chamber with a DEC-MINC computer and Grass amplifiers. Recordings were made from gold cup electrodes (resistances less than S placed at the midline frontal, central, and parietal sites according to the I 0/20 system. Electrodes were referenced to linked ear electrodes. In general, the recording methods closely followed those published by Freedman and associates (7-9), so that the results would be comparable to their published data. One difference, however, was that we used 60 pairs of stimuli in these expeniments, whereas Adler et al. (7) used 32 pairs. Electrical activity was amplified 20,000 times by a 1-Me impedance amplifier with band pass filters (-3 dB) at 0 and 1500 Hz. The EEG was monitored on an oscilloscope, and acquisition of data was interrupted when voltage exceeded 100 j.tV or when there was evidence of artificial muscle activity. Data were digitized at a rate of 1000 Hz for on-line averaging, and simultaneous averages were recorded from lateral canthus electro-oculography (EOG) electrodes and monitored in order to exclude EOG contribution to the data. The artifact rejection was automated by computer program and preset to exclude any sweep that was larger than 90% of the A-to-D capacity for amplitude and that exceeded 15% of the entire sweep for duration. With these exclusion parameters it was felt that EOG artifact as well as other artifacts (e.g., swallowing, minor muscular movement) would be fully removed. Grass 5-88 stimulator-generated clicks of 0.04-msec duration were amplified within a band width of 20-10,000 Hz and delivered through earphones with a mean intensity of 75 dB, as measured by a sound meter at the earphone. Stimuli were presented as a series of paired conditioning and testing stimuli. As described by Freedman et al. (8), the interval between the pains of stimuli was kept
489
SENSORY FIGURE
GATING
DEFICITS
1. Representative
Responses
IN SCHIZOPHRENIA
Data for One Normal
to Two Click Stimuli
Averaged First
25
50
75
100
Subject
on P50 Area
Over 60 Trialsa
Click
125
Second
150
.,o
175
200
225
2
175
200
225
250
Click
0 0.
a) >
25
50
75
l0O
125
150
+
10 IJV 50 msec aRobust
sensory
of the P50 wave the firstclick.
gating
is indicated
following
by the large decrement
the second
click relative
to that
in the area following
constant at 10 seconds, and the interval between the conditioning and test pulses was kept constant at 500 msec in order to test the previously identified interval of maximal differences between schizophrenic and normal subjects in this paradigm. Sixty pairs of responses were averaged. The P50 response data were analyzed as follows. Averaged evoked potentials were measured for peak amplitude and area ofthe P50 wave. The P50 peak was defined as the maximal positive activity occurring between 25 and 75 msec after the stimulus and comesponds to peak amplitude data. Both amplitudes and areas were measured by two methods: 1) from prestimulus baseline to peak and 2) from trough to peak by using the segments of potential immediately pmeceding and following the P50 peak to construct the local baseline. This second method was used in order to avoid all possible problems caused by a small P50 peak riding on a large, long-lasting negative potential and those associated with using a short interval between conditioning and test stimuli (i.e., 500 msec). Analyses of peak amplitude and area measurements of absolute
490
and relative change scores yielded similar results. Therefore, only area data are reported. Three specific prospective hypotheses for P50 area responses were formulated. 1 ) A gating effect would be observed across both groups of subjects and all electrode positions and would be evidenced by a decrement in average P50 response to the second click relative to that of the conditioning first click stimulus. 2) If schizophrenic patients lack robust gating, then the overall decrement in the P50 wave from the response to the first click stimulus to the response to the second click stimulus would be larger in the normal group than in the schizophrenic group (the Diagnosis by Gating intenaction). 3) The hypothesized “hypofrontality” in schizophrenic subjects led to the prediction that regional differences between the two groups would be found. Specifically, the magnitude of the differences in P50 responses between normal and schizophrenic subjects would vary among electrode positions, being largest at the frontal electrode position and smallest at the panetal position (the Electrode Position by Diagnosis by Gating interaction). A three-way repeated measures multivaniate analysis of variance (MANOVA) was conducted using BMDP software (17, 18). The MANOVA included a factor for diagnosis (normal or schizophrenic subjects), a repeated measures factor for the stimulus presentations (stimulus 1, the conditioning click; and stimulus 2, the test click), and a factor for the repeated measures at the three electrode positions (frontal, central, and panietal). A square root transformation of P50 area data was used to preserve assumptions of homogeneity of variance and normality.
RESULTS Figures 1 and 2 show averaged P50 area responses to the first and second clicks for one representative subject from the normal group and one from the schizophrenic group, respectively. Primary
Analysis
A significant gating effect was observed in the P50 area response. Across both groups and all three electrode positions, the mean P50 area response to the experimental second click was significantly diminished from that of the conditioning first click (gating effect, F=176.98, df=1, 38, p
