Fine Structure of the Auditory M100 in Schizophrenia
and Schizoaffective Disorder
Peter Teale, Martin Reite, Donald C. Rojas, Jeanelle Sheeder, and David Arciniegas
Background:
Several studies have demonstrated
anoma-lous asymmetry of the 100-msec latency auditory-evoked
field (M100) in schizophrenia. Recent evidence suggests
this may be a compound component, however. Our study
examines the localization of two M100 subcomponents in
patients with schizophrenia and schizoaffective disorder.
Methods:
Magnetoencephalographic recordings of
audi-tory-evoked fields were obtained for 14 subjects with
schizophrenia, 12 with schizoaffective disorder, and 23
control subjects. Two M100 subcomponents were
identi-fied and localized in each hemisphere.
Results:
Both patient groups exhibited different
lateral-ization compared with control subjects, with the second
subcomponent tending to be less lateralized.
Conclusions:
The second subcomponent may be the major
contributor to previously reported laterality differences.
Future studies might benefit by separating M100
subcom-ponents so that specific functions could be addressed.
Biol Psychiatry 2000;48:1109 –1112 ©
2000 Society of
Biological Psychiatry
Key Words:
Schizophrenia, schizoaffective disorder,
magnetoencephalography, M100, auditory-evoked fields,
laterality
Introduction
P
sychosis has been associated with anomalous brain
asymmetry (Crow 1990). Several studies have
de-scribed abnormal asymmetry of the 100-msec latency
auditory-evoked field component (M100) in schizophrenia
(less asymmetric or more variable in schizophrenia) as a
manifestation of this anomaly (Hajek et al 1997; Reite et
al 1989, 1997; Tiihonen et al 1998). Some evidence
suggests such anomalies in M100 lateralization may be a
manifestation of a different auditory cortical organization
in schizophrenia (Rojas et al 1997). All previous
magne-toencephalography (MEG) studies in schizophrenia,
how-ever, have considered the M100 a unitary component. The
associated evoked potential, designated N100, is generally
considered to be a superposition of responses produced by
as many as six separate sources (Na¨a¨ta¨nen and Picton
1987) having different spatial and temporal characteristics.
Several of these generators are thought to be radial in
orientation and therefore will not produce measurable
magnetic fields. Recent MEG work suggests the M100 is
a compound component (Lu et al 1992; Sams et al 1993;
Teale et al 1998; Zouridakis et al 1998). In normal
subjects, the M100 component may actually be generated
by two active regions separated in time (by 25 msec) and
in space (later source about 0.1 cm anterior in the L
hemisphere, and 0.8 cm anterior in the right hemisphere;
Teale et al 1998).
Independent localization of these two components has
not been reported in subjects with psychosis. In light of the
laterality differences reported in the literature for the
combined component in subjects with psychosis, we
wished to determine if one of the two putative
subcompo-nents might be selectively contributing to the variance
reported for M100 lateralization.
Methods and Materials
Our subject population consisted of 14 individuals with schizo-phrenia (mean age 36.7 6 6 years), 12 with schizoaffective
disorder (mean age 36.8 6 10 year), and 23 control subjects
(mean age 34610 year) recruited from the Denver metropolitan
area. All patients were in outpatient treatment, and most were medicated. All were men. Diagnosis was determined using the Structured Clinical Interview for DSM-IV and review of medical records. Mean age of onset for the subjects with schizophrenia was 21.62 6 5.35 years and 22.00 6 8.24 years for the
schizoaffective group. All were right handed based on the Annett criteria (Annett 1985). All had normal hearing thresholds at 1 kHz as determined by audiometry. Informed consent was ob-tained from all subjects in accordance with the Colorado Multi-ple Institutional Review Board.
Recordings were obtained inside a magnetically shielded room, with subjects semi-reclining in a nonmagnetic chair, using a BTI Model 607 seven-channel second order biomagnetometer with 18-mm coil diameter and 4-cm baseline. A custom bite plate was fabricated for each subject. Sufficient measurements
cen-From Biomagnetic Imaging Laboratory, Department of Psychiatry, University of Colorado Health Sciences Center, and Denver VA Medical Center, Denver, Colorado.
Address reprint requests to Martin L. Reite, M.D., University of Colorado Health Sciences Center, 4200 East Ninth Avenue, Box C268-68, Denver CO 80262. Received January 14, 2000; revised May 17, 2000; accepted May 23, 2000.
tered over the temporal lobe were made using five instrument positions (53 7 5 35 B-field measurements) on each
hemi-sphere to assure adequate modeling of both generators with a single equivalent current dipole (ECD) at their respective laten-cies. Subjects viewed a silent video throughout the duration of the recording to maintain a reasonably constant attentional state. Stimuli were computer-generated, 30-msec duration (5-msec rise and fall), 85-dB sound pressure level (measured at the ear), 1-kHz tone pips delivered via polyurethane tubing to the ear opposite the hemisphere being recorded. The interstimulus inter-val was fixed at 6 sec.
A sonic digitizer was used to establish a head frame coordinate system and coregister the position of instrument. The X axis was determined by the line connecting the preauricular points (posi-tive X is outward through the right ear). The Y axis was the line orthogonal to the midpoint of the X axis (the origin) and contained in the plane established by the X axis and the nasion (positive Y anterior). The Z axis was normal to X and Y at the midpoint and exits the top of the head (positive Z is above the origin).
Magnetoencephalographic recordings were wide band (0.1–1 kHz) and digitized at a rate of 2.275 kHz. Recordings included a 200-msec prestimulus baseline and 250 msec following stimulus delivery. Artifact-free responses to 50 stimuli were averaged and then filtered with a phase invariant digital filter (24 dB/octave slope) with a pass band of 0 to 55 Hz.
Source localization analysis was performed across the entire poststimulus window from 0 to 245 msec using a single moving ECD in a conductive sphere model (Sarvas 1987) and a sliding 10-msec width time window at 1 msec increments for each hemisphere. The mean B-field amplitude for each coil for the 10-msec window (starting 5 msec before and ending 5 msec after, each time point) was computed by averaging the samples (23) in this interval. To identify the two candidate sources, a goodness of fit measure (F) was computed at each time point by dividing the root mean square (RMS) value of the error (i.e., the difference between the measured field amplitudes and the final iteration of the forward solution) by the RMS value of the data. This measure was then evaluated in the window from 60- to 145-msec poststimulus by searching for local minima in 5-msec blocks and then discarding all but the two lowest values (i.e., these are the latencies when the ECD best modeled the data). Mean latencies were 77611 msec poststimulus for the early
source and 100 6 11 msec for the late source. We have
previously reported that the two sources, so determined, were significantly different from one another. Additional details are provided in Teale et al (1998).
Statistical Procedures
All statistical analyses were performed using Statistica 5.5 (StatSoft, Tulsa OK). Analyses were conducted usinga 5.05.
Sums of squares for all analyses of variance designs were type III. We employed a 33232 analysis of variance (ANOVA;
group by hemisphere by source latency), with hemisphere and source latency treated as repeated measures and y location of the sources as the dependent variable, to test differences in source location. To examine our hypothesis concerning schizophrenia
and schizoaffective disorder, two orthogonal interaction contrasts were coded within this ANOVA to test 1) whether the schizoaf-fective and schizophrenia groups, when combined together, differed from the control group and 2) whether the schizophrenia group and schizoaffective group differed from each other.
Results
The first contrast was significant [
F
(1,46)
5
4.03,
p
,
.05], indicating that schizophrenic and schizoaffective
subjects differed from control subjects. The second
con-trast was nonsignificant [
F
(1,46)
5
0.03,
p
.
.05],
indicating that schizoaffective and schizophrenic patients
did not differ significantly from each other in overall
source location. In control subjects, the second source was
more lateralized than the first [
F
(1,46)
5
3.76,
p
,
.06].
The early and late sources were not lateralized in either
patient group [
F
(1,46)
5
0.03,
p
.
.05]. Figure 1
illustrates the mean anteroposterior dipole locations for
each group for the early and late source.
Discussion
Our findings suggest the second subcomponent could be a
major contributing factor to the laterality differences
previously reported for the combined M100 in
schizophre-nia. Thus, experimental demands, or signal processing
differences that inadvertently emphasize one of the two
subcomponents, could influence the laterality differences
reported in the literature. Low-pass filtering, for example,
would likely preferentially emphasize the subcomponent
with the highest amplitude (assuming the subcomponents
are similar in waveform) and so shift source parameter
estimates. A recent report by Rockstroh and colleagues
(Rockstroh et al 1998) demonstrated a failure in the
development of contralateral dominance in patients with
schizophrenia as indexed by M100 ECD strength. Source
estimates, however, did not differentiate between control
subjects and patients with respect to left–right, anterior–
posterior location. In this case, data were low-pass filtered
at 20 Hz. Consideration of a two source model might
produce a larger difference in these parameters.
Anomalous lateralization of the M100 component
pre-viously described in patients with schizophrenia has been
suggested to reflect evidence of a different cortical
orga-nization in schizophrenia (Rojas et al 1997). The present
findings do not directly address the issue of cortical
organization but suggest that such work might profitably
include independent consideration of the several
subcom-ponents contributing to the M100. Schizophrenic subjects
did not differ from subjects with schizoaffective disorder
on this metric, implicating a commonality of the temporal
lobe pathology in the schizophreniform psychoses.
Sub-1110 BIOL PSYCHIATRY P. Teale et al
jects with affective psychoses have yet to be studied in this
regard.
The M100 complex has previously been associated with
auditory sensory memory or echoic memory (Imada et al
1993; Lu et al 1992; Sams et al 1993). Lu et al (1992)
suggested that two components of the M100 complex,
what they termed L100m (an earlier, posterior component)
and N100m (a later, more anterior component) had
differ-ent trace lifetimes and suggested differdiffer-ent functional roles
for each source. In a later paper, Sams et al (1993) also
observed two sources underlying the M100 complex,
which they termed N100m
pand N100m
a(early posterior
and later anterior, respectively). Sams et al (1993) also
observed differing trace decays for the two sources, with
the N100m
pexhibiting shorter time constant of decay than
the N100m
a, in agreement with the findings of Lu et al
(1992). Loveless et al (1996) suggested that these two
sources might reflect the short and long term auditory
stores first proposed by Strous and colleagues (Strous et al
1995) on the basis of behavioral data. Our data indicate
that the differences between the control group and two
psychotic groups are largely confined to the later source
(N100m
a), suggesting an impairment in the longer term
auditory store. By implication, experimental paradigms
with different memory demands may preferentially
acti-vate one subcomponent. If, for example, the amplitude of
one was increased, the apparent localization of the
com-bined source might appear to move forward or backward.
It might be important to determine the extent to which
attentional mechanisms may be involved in the activation
of the second course. Such mechanisms, thought to be
disturbed in schizophrenia, might preferentially alter
strength of location of this source. Clearly, there are
concerns relating to the variable nomenclature of these
auditory evoked field components, which to date, in the
absence of agreed upon terminology, tend to be laboratory
specific. We hope our discussion will help relate the
somewhat varying nomenclatures.
One caveat in considering these findings is that they are
based on a sequential source model. Improvements in the
accuracy of localization might be obtained with the
appli-cation of a true two-source model in the time window
during which both sources are active. This might involve
a two-pass modeling strategy wherein the first pass would
be as reported here and the second pass would involve a
pair of simultaneously active spatiotemporal ECDs
cover-ing the latency window between the two sources identified
in pass one.
This work was supported by National Institute of Mental Health Grant No. MH47476.
References
Annett M (1985):Left, Right, Hand and Brain: The Right Shift Theory.Hillsdale, NJ: Lawrence Erlbaum.
Crow TJ (1990): Temporal lobe asymmetries as the key to the etiology of schizophrenia.Schizophr Bull16:433– 443. Hajek M, Huonker R, Boehle C, Volz H-P, Nowak H, Sauer H
(1997): Abnormalities of auditory evoked magnetic fields and structural changes in the left hemisphere of male schizophren-ics—a magnetoencephalographic-magnetic resonance imag-ing study.Biol Psychiatry42:609 – 616.
Imada T, Hari R, Loveless N, McEvoy L, Sams M (1993): Determinants of the auditory mismatch response. Electroen-cephalogr Clin Neurophysiol87:144 –153.
Figure 1. Mean anteroposterior M100 generator locations6SEM.
Fine Structure of the Auditory M100 BIOL PSYCHIATRY 1111
Loveless N, Levanen S, Jousmaki V, Sams M, Hari R (1996): Temporal integration in auditory sensory memory: Neuro-magnetic evidence. Electroencephalogr Clin Neurophysiol
100:220 –228.
Lu Z, Williamson SJ, Kaufman L (1992): Behavioral lifetime of human auditory sensory memory predicted by physiological measures.Science258:1668 –1670.
Na¨a¨ta¨nen R, Picton T (1987): The N1 wave of the human electric and magnetic response to sound: A review and an analysis of the component structure. Psychophysiology
24:375– 425.
Reite M, Sheeder J, Teale P, Adams M, Richardson D, Simon J, et al (1997): Magnetic source imaging evidence of sex differences in cerebral lateralization in schizophrenia. Arch Gen Psychiatry54:433– 440.
Reite M, Teale P, Goldstein L, Whalen J, Linnville S (1989): Late auditory magnetic sources may differ in the left hemi-sphere of schizophrenic patients. A preliminary report.Arch Gen Psychiatry46:565–572.
Rockstroh B, Clementz B, Pantev C, Blumenfeld L, Sterr A, Elbert T (1998): Failure of dominant left-hemisphere activa-tion to right-ear stimulaactiva-tion in schizophrenia. Neuroreport
9:3819 –3822.
Rojas D, Teale P, Sheeder J, Simon J, Reite M (1997): Sex-specific expression of Heschl’s gyrus functional and struc-tural abnormalities in paranoid schizophrenia.Am J Psychi-atry154:1655–1662.
Sams M, Hari R, Rif J, Knuutila J (1993): The human auditory sensory memory trace persists about 10 sec: Neuromagnetic evidence.J Cogn Neurosci5:363–370.
Sarvas J (1987): Basic mathematical and electromagnetic con-cepts of the biomagnetic inverse problem. Phys Med Biol
32:11–22.
Strous RD, Cowan N, Ritter W, Javitt DC (1995): Auditory sensory (“echoic”) memory dysfunction in schizophrenia.
Am J Psychiatry152:1517–1519.
Teale P, Sheeder J, Rojas DC, Walker J, Reite M (1998): Sequential source of the M100 exhibits inter-hemispheric asymmetry.Neuroreport9:2647–2652.
Tiihonen J, Katila H, Pekkonen E, Jaaskelainen IP, Huotilainen M, Aronen HJ, et al (1998): Reversal of cerebral asymmetry in schizophrenia measured with magnetoencephalography.
Schizophr Res30:209 –219.
Zouridakis G, Simos PG, Papanicolaou A (1998): Multiple bilateral asymmetric cortical sources account for the auditory N1m component.Brain Topogr10:183–189.
1112 BIOL PSYCHIATRY P. Teale et al
