Ni Hms 88603
Ni Hms 88603
Ni Hms 88603
Author Manuscript
Schizophr Res. Author manuscript; available in PMC 2011 October 7.
Published in final edited form as:
NIH-PA Author Manuscript
Abstract
BackgroundOne of the cardinal features of schizotypal personality disorder (SPD) is language
abnormalities. The focus of this study was to determine whether or not there are also processing
abnormalities of pure tones differing in pitch and duration in SPD.
MethodsThirteen neuroleptic-nave male subjects met full criteria for SPD and were group-
matched on age and parental socioeconomic status to 13 comparison subjects. Verbal learning was
NIH-PA Author Manuscript
measured with the California Verbal Learning Test. Heschls gyrus volumes were measured using
structural MRI. Whole-brain fMRI activation patterns in an auditory task of listening to tones
including pitch and duration deviants were compared between SPD and control subjects. In a
*
Corresponding author. Serious Mental Illness Program, VA Boston Healthcare System, Psychiatry 116A-RC, 940 Belmont St.,
Brockton, MA 02401, United States. Psychiatry Neuroimaging Laboratory, 1249 Boylston St, Boston, MA 02115, United States. Tel.:
+1 857 225 0093 (Cell); fax: +1 617 525 6150.
Contributors
Dickey: designed the study, collected the fMRI data, analyzed the clinical and fMRI data, and wrote the manuscript.
Morocz: analyzed the fMRI data.
Niznikiewicz: collected and analyzed the ERP data, and contributed to the manuscript.
Voglmaier: collected and analyzed CVLT data, and performed SCID interview.
Dreusicke: analyzed the fMRI data. Toner: aided in fMRI data collection.
Yoo: aided in initial fMRI data collection.
Khan: aided in final fMRI data presentation.
Shenton: reviewed the manuscript.
McCarley: reviewed the manuscript.
Conflict of interest
All authors declare that they have no conflict of interest.
Dickey et al. Page 2
second and separate ROI analysis we found that peak activation in superior temporal gyrus (STG),
Brodmann Areas 41 and 42, was correlated with verbal learning and clinical measures derived
from the SCID-II interview.
NIH-PA Author Manuscript
ResultsIn the region of the STG, SPD subjects demonstrated more activation to pitch deviants
bilaterally (p<0.001); and to duration deviants in the left hemisphere (p=0.005) (two-sample t).
SPD subjects also showed more bilateral parietal cortex activation to duration deviants. In no
region did comparison subjects activate more than SPD subjects in either experiment. Exploratory
correlations for SPD subjects suggest a relationship between peak activation on the right for
deviant tones in the pitch experiment with odd speech and impaired verbal learning. There was no
difference between groups on Heschls gyrus volume.
ConclusionsThese data suggest that SPD subjects have inefficient or hyper-responsive
processing of pure tones both in terms of pitch and duration deviance that is not attributable to
smaller Heschls gyrus volumes. Finally, these auditory processing abnormalities may have
significance for the odd speech heard in some SPD subjects and downstream language and verbal
learning deficits.
Keywords
Schizotypal personality disorder; Auditory; Schizophrenia; fMRI; MRI; Imaging; Tone
processing; Evoked potential; ERP; Mismatch negativity; Pitch; Frequency; Duration; MMN
NIH-PA Author Manuscript
1. Introduction
Auditory sensory processing has been found to be impaired in schizophrenia (Salisbury et
al., 1998; Javitt et al., 2000) and correlate with clinical features, particularly negative
symptoms (Javitt et al., 2000), (Kasai et al., 2002) (Leitman et al., 2005) and cognitive
impairment (Baldeweg et al., 2004). Abnormalities in the superior temporal gyrus (STG)
have been implicated in the processing of pure tones, a fundamental element of complex
sounds and language, using fMRI (Wible et al., 2001) and event-related potentials (Salisbury
et al., 1998). Hallucinations, which may be considered an error in auditory sensory
processing, have been associated with the STG (Dierks et al., 1999) (Cleghorn et al., 1992),
with the STG more activated during hallucinations than actual speech (David et al., 1996).
The STG has also been implicated in verbal learning deficits in schizophrenia using PET
(Ragland et al., 2001). However, research in schizophrenia has been confounded by
potential modulating effects of neuroleptic medications on fMRI signal (Stephan et al.,
2001; Brassen et al., 2003). Neuroleptic-nave subjects are needed to ensure that research
findings are due to underlying neuropathology rather than iatrogenic effects.
NIH-PA Author Manuscript
Subjects with schizotypal personality disorder (SPD) may provide an ideal population of
neuroleptic-nave subjects for fMRI studies. Although SPD shares many features with
schizophrenia, it is not a direct proxy for schizophrenia, as SPD subjects are not psychotic.
Nonetheless, SPD and schizophrenia have traditionally been considered part of the
schizophrenia spectrum disorders based on epidemiological data (Kety et al., 1967) (Kendler
et al., 1993), shared clinical features such as thought disorder (Dickey et al., 1999) and
paranoia (Dickey et al., 2005), similar biological markers (Siever and Davis, 2004),
comparable cognitive deficits in verbal learning (Voglmaier et al., 1997) (Voglmaier et al.,
2000) (Voglmaier et al., 2005), and overlapping morphometric abnormalities (Dickey et al.,
2002a,b).
Indeed, one brain region critical for early sensory auditory processing, Heschls gyrus (Yoo
et al., 2005), has been shown to have reduced volumes in subjects with SPD (Dickey et al.,
2002a,b), similar to what has been shown in schizophrenia (Hirayasu et al., 2000). Heschls
gyrus lies on the plane of the STG, a region found to have small volumes in males with SPD
(Dickey et al., 1999; Downhill et al., 2001), and in females with SPD with a family history
of mental illness (Dickey et al., 2003). Of particular interest to this report, however,
NIH-PA Author Manuscript
Heschls gyrus is noted to have marked inter-subject morphometric volume and shape
variability (Leonard et al., 1998; Knaus et al., 2006), thus complicating the interpretation of
volume data (Knaus et al., 2006).
One question in the literature is how to best measure deficits in early auditory sensory
NIH-PA Author Manuscript
the current report. Structural MRI and neuropsychological testing procedures are also
included. The possible relationship between early auditory processing and downstream
language and other cognitive functions, as well as the highly complex clinical manifestations
of SPD, is also evaluated in an exploratory fashion. These questions are important to ask in
SPD subjects as they are neuroleptic-nave, thus, a complicated confound in similar studies
of auditory function with schizophrenic subjects is removed (Umbricht et al., 1998).
2. Methods
2.1. Subject recruitment
All subjects were male; right-handed; between 18 and 55 years old; neuroleptic-nave; on no
psychotropic medications; had no history of ECT, neurologic disorder, substance abuse in
the last 1 year or substance dependence in the last 5 years; and were recruited from the
community through newspaper and subway advertisements (for recruitment specifics, see
Dickey et al., 2005). Note that data for past psychotropic use of any kind or substance
dependence beyond five years ago was not available. SPD subjects met DSM-IV criteria for
SPD using the SCID and SCID-II interviews and had no personal history of bipolar disorder
NIH-PA Author Manuscript
nor psychosis. Thirteen SPD subjects were group-matched on age, parental socio-economic
status, and estimated IQ to 13 comparison subjects who had additional exclusionary criteria
of no personal history with Axis I or Axis II disorder as determined by SCID, or first-degree
relative history of Axis I disorder.
volumes (Dickey et al., 2002a,b). The anterior boundary was the temporal stem; the
posterior boundary was the complete crux of the fornix; the lateral boundary was determined
by a horizontal line extending laterally from the superior most white matter track of the
STG. There was one major methodological difference in the drawing between the two
reports, however. In this report axial views were used initially to guide the definition of the
extent of Heschls. Axial views were not available previously. With the use of Slicer
software (www.slicer.org) one could now visualize whether there was a single transverse
gyrus, a common medial stem branching to two more laterally in which case both would be
included per Steinmetzs criteria (Steinmetz et al., 1986), or two medial stems joining more
laterally in which case only the more anterior portion would be selected. This ability to
visualize and draw in three dimensions represents a significant technological advancement
compared with previous capabilities and is important given the marked inter-subject and
even inter-hemispheric variability of this gyrus (Knaus et al., 2006). Volumes were
corrected for total intracranial contents using a regression procedure (Dickey et al.,
2002a,b). Inclusion of Heschls volume measurement was important for the interpretation of
the functional data.
Whole-brain images were acquired on a 3.0T GE system in the oblique axial plane parallel
to the superior temporal plane prescribed from the localizer images. Functional image
parameters included whole-brain coverage, 30 slices, 5 mm thick with 1 mm gap, 162
acquisitions the first 6 of which were removed, FOV 24, matrix 6464, TR 2.5, TE 35. The
same prescription (orientation/slice thickness) was applied to obtain a low-resolution SPGR
sequence. These low-resolution SPGR images were reoriented, realigned, and normalized to
the Montreal Neurological Institute (MNI) T1 template with the resulting matrix files
applied to the functional images using SPM2. Functional images were subsequently
smoothed (using 12 FWHM 3-dimensional Gaussian kernel).
2.5. Stimuli
Two experiments were employed to activate the auditory cortex, one using the pitch
deviants and the second using the duration deviants. Tones were transmitted via sound-
insulated and cushioned earphones (Silent Scan, Avotec, Jensen Beach, FL) at 80 db SPL
(Sound Pressure Level). All tones had 10 ms rise and fall times and an interstimulus interval
of 300 ms and were played at 80 db. Experiment 1: Pitch. The standard tone was 500 Hz and
NIH-PA Author Manuscript
the deviant tone was 2000 Hz, all tones 100 ms in duration (Fig. 1a). Experiment 2:
Duration. The standard tone was 50 ms in duration, the deviant tone was 200 ms in duration,
with a frequency of 500 Hz for all tones (Fig. 1b and c). Presentation was block design with
30 s blocks of tones alternating with 30 s rest (Fig. 1d). Each tone block consisted of 100
tone presentations, alternating between blocks of 100% standard tones and mixed blocks of
75% standard tones and 25% deviant tones. In the mixed standard and deviant blocks the
order of standard and deviant tones was randomly determined. Only one deviant (pitch or
duration) was presented in a given run. Run duration was 645. There were two runs for
pitch and two for duration for a total of four runs, order counterbalanced among subjects.
Subjects heard 1050 standard tones and 150 deviant tones for a total of 1200 tones per
experiment (Fig. 1e). For technical reasons, data were acquired on only 11 SPD subjects for
the duration experiment, therefore, the sample size differs in the two fMRI experiments.
Task for both experiments was passive listening with eyes closed. Following each run
subjects were queried as to whether they heard the tones to ensure adequate hearing and
wakefulness. All subjects affirmed that they did indeed hear the tones after each run.
Subjects were not asked nor expected to differentiate between the tones.
2.6.1. Whole-brain analysisThis was the primary statistical analysis for the pitch and
duration fMRI experiments. Whole-brain event-related procedures were employed
statistically in order to single out the effect of all tones (standard+deviant), and in a separate
analysis, the differential effect of hearing only deviant tones. Restated, the hemodynamic
responses were modeled first in order to examine the effect of being in the scanner and
hearing all tones, both standard and deviant, regardless of block type (main effect) (Fig. 1e,
in blue and green). Subsequently, the hemodynamic effect of hearing only deviant tones
(parametric effect) was modeled (Fig. 1e, in green). This isolates the effects of hearing
deviant tones on top of the effect of hearing all tones, that is, standard and deviant (green
only on top of green and blue). Specifically, the main effect of hearing all tones (standard
+ deviant, regardless of block) vs. rest were modeled using one regressor (effect of hearing
deviant tones only) in the General Linear Model for each subject. Second, the differential
effect of hearing the deviant tones was modeled as a parametric regressor (deviant tones as 1
and standard tones with value 0) in order to measure the deviant-related modulation of the
BOLD signal time course for each subject. Thus the effect of hearing all tones (both
standard and deviant combined, the effect of being in a scanner and hearing tones, the main
effect) could be compared as well as the effect of only hearing deviant tones (parametric
NIH-PA Author Manuscript
effect). Subject specific whole-brain contrast images from the two groups were pooled and
were compared using one-sample (all tone and deviant tone conditions for both experiments)
and two-sample t tests (all tone and deviant tone conditions for both experiments) for the
random-effects analysis. Note that these analyses were whole-brain analyses.
regions were selected a priori as they likely are the regions involved in tone processing (Yoo
et al., 2005; Kropotov et al., 1995; Alho, 1995). Note that Brodmann Area 41 derived from
PickAtlas does not directly correspond to the manual Heschls drawing above. They are
NIH-PA Author Manuscript
slightly different measurements but both roughly correspond to presumed areas of pure tone
processing. Threshold was set at 0.05 corrected.
3. Results
NIH-PA Author Manuscript
areas of significantly more activation in both the temporal and parietal lobes in SPD subjects
compared with controls (Fig. 2c and d).
NIH-PA Author Manuscript
3.5. Clinical correlations with fMRI activation patterns in the duration experiment
Of note, correlations were found on the right side between the clinical symptom of odd
thinking/speech in the pitch experiment (Table 2). The more abnormal the activation (t-
score) while hearing deviant pitch tones, the more the impairment due to the clinical
symptom. Verbal learning, a key abnormality in SPD, correlated with all right-sided regions
such that the fewer words learned in trials 15, the more abnormal the activation while
hearing deviant pitch tones. The number of correlations performed was high given the
exploratory nature of the study and results would not have withstood a Bonferroni
correction. However, they are included here to generate future hypotheses regarding the
functional/anatomic relationships in SPD. As stated in the Methods section, included in
Table 2 are only those correlations which were found in at least 2/3 regions/side (Fig 3).
4. Discussion
The main finding of this report is increased activation in the region of the STG bilaterally in
SPD subjects compared with controls while subjects heard deviant tones regardless of
whether the deviance was in the pitch or duration. In the whole-brain analysis, in no region
of the brain did comparison subjects activate more than SPD subjects. These findings
NIH-PA Author Manuscript
suggest that SPD subjects compared with controls had inefficient or hyper-responsive
processing of two of the most basic aspects of auditory sensory stimuli. These data cannot
unequivocally address the question as to whether the SPD subjects have inefficient
processing and, thus, needed to recruit more neurons to process the tones with a resulting
larger hemodynamic response, or, the reverse, that SPD subjects have an exaggerated
response to subtle changes in sensory inputs with a resulting larger hemodynamic response
curve. More basic research is required to differentiate those two possible interpretations.
Nonetheless, these current findings cannot be attributed to small Heschls gyrus volume as
the groups did not differ on that measure.
Deficits in early auditory sensory processing, including N1, MMN, and P300, have been
shown using ERP in schizophrenic subjects (Baldeweg et al., 2004; Kasai et al., 2002;
Salisbury et al., 1998; Javitt et al., 2000; Javitt et al., 1995; Elvevag et al., 2004), in
prodromal subjects (Brockhaus-Dumke et al., 2005), and in SPD subjects (Salisbury et al.,
1996; Niznikiewicz et al., 2000; Liu et al., 2007). Moreover, progressive changes in
schizophrenic subjects can be indexed using a MMN paradigm in the ERP environment
(Salisbury et al., 2007). Indeed, ERP has certain advantages in testing auditory sensory
processing in that tones can be played in sound-proof laboratories with little to no
NIH-PA Author Manuscript
extraneous noise, experimental design allows for a large number of deviants, greater than
160 in most studies (Shelley et al., 1991; Javitt et al., 1995; Umbricht et al., 1998; Michie et
al., 2000; Baldeweg et al., 2004; Brockhaus-Dumke et al., 2005) (exception, Kirino and
Inoue, 1999) and allows for precise time recordings on the order of milliseconds at the
expense of poor spatial resolution.
In contrast, fMRI allows for more accurate anatomic localization at the expense of poor time
resolution and high ambient scanner noise, although some have utilized the scanner noise as
stimuli (Mathiak et al., 2002; Kircher et al., 2004). Applying a prototypic mismatch
paradigm has been challenging in the fMRI environment resulting in few papers on
schizophrenia (Wible et al., 2001; Kircher et al., 2004). Even using scanner noise as the
stimulus causes limitations as variation in pitch cannot be tested (Mathiak et al., 2002;
Kircher et al., 2004).
How one might measure the effect of tone deviants also differs between ERP and fMRI.
ERP MMN studies rely on traditional subtraction of waveforms (i.e.: waveforms from
standard tones subtracted from waveforms from deviant tones) (Javitt et al., 1995; Umbricht
NIH-PA Author Manuscript
et al., 1998). Unfortunately, in fMRI aberrant activation patterns can result from subtraction.
These areas of unusual activation may reflect spontaneous neuronal activity, that is, areas
not predicted by task demands, which cannot be experimentally controlled (Binder et al.,
1999). A mismatch task is not cognitively demanding, in fact it is arguably pre-attentive
(Naatanen 1990), possibly allowing for more task-unrelated thoughts (Binder et al., 1999).
These task-unrelated thoughts can result in changes in blood flow detected by BOLD
method (Binder et al., 1999). In fact, the hemodynamic effect of such thoughts using the
subtraction method has been demonstrated in a tone task (Binder et al., 1999). Moreover,
depending on the statistical approach, subtraction or other, results can differ markedly
(Friston et al., 1996). The subtraction method assumes pure insertion, that the cognitive
process is irrespective of the cognitive or physiological context (Friston et al., 1996), an
assumption which may not be valid. Indeed, the two fMRI papers in schizophrenia
compared groups on standard and deviant tones separately, thus avoiding the difficulties of
the subtraction method in an fMRI environment (Wible et al., 2001; Kircher et al., 2004).
In the current fMRI study the effect of deviants was measured using a parametric analysis.
Specifically, the differential processing of deviant tones against the background of hearing
all tones was measured (i.e.: the first parameter is the effect of hearing all tones, both
NIH-PA Author Manuscript
standard and deviant; and the second parameter is the effect of hearing deviant tones only)
(Fig. 1e). This has the benefit of making no assumption of pure insertion (Friston et al.,
1996) and isolates the effect of hearing deviant tones on top of hearing all tones. In
addition, this method may possibly minimize the hemodynamic effect of task-unrelated
thoughts (Binder et al., 1999).
Another aspect of this study which differed from the MMN ERP literature was the use of
disparate frequencies and durations between the standard and deviant tones. Stimuli used
across studies have varied with differences between standard and deviant pitches as small as
24 Hz (10001024 Hz) (Javitt et al., 1995) to as large as 1000 Hz (10002000 Hz) (Kirino
and Inoue, 1999) and differences in duration as small as 25 ms (2550 ms) (Baldeweg et al.,
2004). Nonetheless, the presence of the MMN-like deflection in both conditions suggests
that these experimental parameters engage the early auditory sensory processing stream.
Indeed, stimulus presentation features including probability and degree of separation of the
deviants, and interstimulus interval, are all important variants which may affect the results
(Michie et al., 2000). However, as SPD has demonstrated less severe abnormalities than
schizophrenia on electrophysiological measures (Trestman et al., 1996), more extreme
differences in terms of pitch and duration between the standard and deviant tones were
NIH-PA Author Manuscript
selected to amplify the fMRI signal. Moreover, a larger differentiation between standard and
deviant tones would help to compensate for any potential deficit in tone discrimination
processing described in the schizophrenia literature (Javitt et al., 2000; Leitman et al., 2005).
Although a lower percentage of deviant tones would have similarly increased an
electrophysiological signal (Javitt et al., 2000), in order to achieve enough trials for an
adequate hemodynamic response, a higher percentage of deviant tones was selected.
Although the paradigm used in this study is not standard mismatch, evidence from the MMN
literature may inform the current findings. For example, generators for the MMN are
thought to be from the auditory cortex (Kropotov et al., 1995) with physical separation of
foci for processing of pitch and duration (Molholm et al., 2005), similar to the current
findings. Some workers have also noted activation from frontal generators thought to reflect
a shift of attention toward the deviant, although recent work suggests that the role of the
frontal generator may not depend on attention requirements (Shalgi and Deouell, 2007).
Indeed, a recent review by Naatanen et al. (2007) suggests that frontal activity may be due to
the sum of supratemporal generators (Naatanen et al., 2007). Our report did not find
evidence of fMRI activation in either group in the frontal lobe, possibly due to the use of
NIH-PA Author Manuscript
largely disparate standard and deviant tone features (Shalgi and Deouell, 2007). The
activation of the parietal lobe in the duration experiment is consistent with the hypothesis
that the parietal lobe is important for time perception (Harrington et al., 1998) and change-
detection (Molholm et al., 2005). Indeed, bilateral activation has been demonstrated with
deviants in duration (Molholm et al., 2005). Finally, abnormalities in MMN have been
documented in clinically diagnosed patients with SPD (Liu et al., 2007).
In this report, in contrast to our previous report (Dickey et al., 2002a,b), we did not show
any difference in Heschls gray matter volume. We would argue, however, that, in general,
in the peri-Heschls/STG region, SPD subjects likely have smaller volumes (Dickey et al.,
2002a,b, 2003, 1999; Downhill et al., 2001 {Koo, 2006 #3366}). There are several
differences between our two reports on Heschls gyrus ((Dickey et al., 2002a,b) and current)
and that volumes measured depend on many factors including subject demographics, co-
morbidity, subject N, image viewing tools which can affect landmark detection, and inter-
subject and inter-hemispheric variability of Heschls morphology, particularly in
schizophrenia spectrum disorders. First, in the earlier report there was a significant
difference between IQ and personal socio-economic status between groups suggesting that
cohort may have represented a more impaired group of SPD subjects. Second, and delving
NIH-PA Author Manuscript
deeper into the issue of subject characteristics, the percentage of subjects meeting DSM-IV
criteria for other Axis II disorders differs between samples. In the prior report there were
more borderline personality disorder subjects (31% vs. 15% in current sample) and fewer
who met criteria for paranoid personality disorder (37.5% vs. 46% in current report). Little
is known about brain morphology of the STG region in paranoid and borderline personality
disorders (no papers found per PubMed search performed 4/17/08). Whether the presence of
co-morbid personality disorders has an effect on Heschls measurement cannot be
addressed. Third, in this current sample, the subject N is smaller, leading to possibly less
stable data. Fourth, the current study used Slicer as the image processing software tool, a
more sophisticated tool than previously available (www.slicer.org) as this tool now allows
for simultaneous 3D viewing. As noted under the Methods section above, the delineation of
Heschls previously may have included more peri-Heschls area posteriorly thus leading to
the differences between groups in that report (Dickey et al., 2002a,b). In the current report
the more posterior transverse gyri may not have been included. Fifth, there is marked inter-
subject and even inter-hemispheric variability of the STG, particularly, Heschls gyrus
(Lange et al., 1997; Leonard et al., 1998; Sweet et al., 2005; Knaus et al., 2006). To clarify,
it is quite common for there to be two parallel transverse or Heschls gyri (Knaus et al.,
2006; Sweet et al., 2005). However, these parallel transverse gyri can be completely separate
NIH-PA Author Manuscript
or partially separate. If they are partially separate they can be merged medially, part way
down the gyrus, or laterally. What should one consider as Heschls gyrus is not clear.
Should one include both, part of both, or only the more anterior one? Both the prior report
and this paper used Steinmetzs criteria (Steinmetz et al., 1986) as the posterior boundary.
This arbitrary criteria state that if the gyri are merged medially with a common stem, then
one can include both. If they are merged part way down the gyrus or laterally, then they
should be divided into two separate gyri and only include the more anterior gyri as Heschls.
Note that this criteria is helpful, but arbitrary. Subtle differences in visualization abilities or
in morphometry can result in significant volume differences which may not be meaningful
functionally. Indeed, the functional anatomy of AI (primary auditory cortex) may not be
restricted to medial Heschls gyrus, regardless of its definition or criteria (Sweet et al.,
2005). Indeed, recent work suggests that there may be two primary auditory receptive fields
in the human auditory system, one more medial, one more lateral (Engelien et al., 2002).
Close inspection of the receptive field maps (Engelien et al., 2002) suggests that the more
lateral field may be partially located on what may be considered by Steinmetzs criteria to be
the second transverse gyrus, an area more likely captured in our first report but less likely
captured in this current report. That functionally important area (Sweet et al., 2005) may
NIH-PA Author Manuscript
represent the difference in these two reports. Indeed, other workers suggest that volume
asymmetries in the region of the primary auditory cortex do not correspond to functional
activation while hearing tones and word pairs (Yoo et al., 2005), possibly due to the poor
matching of gyral sulcal patterns and cytoarchitechtonically defined acoustic regions
(Morosan et al., 2001), and marked variation in the morphology of Heschls gyrus (Leonard
et al., 1998) (Sweet et al., 2005). Furthermore, regional differences in STG volumes are
more marked in schizophrenia than in controls (Park et al., 2004). In sum, the lack of
replication of our previous report is likely due to a combination of factors: differences in
subject demographics, presence of co-morbid disorders, subject N, difference in
measurement tools, and inter-subject morphometric variability of Heschls gyrus particularly
in the schizophrenia spectrum.
Despite these limitations to direct comparison between the two samples, the lack of
difference between groups in Heschls volume in this study is informative: one cannot
attribute differences in activation in the current report to subtle volume differences of
Heschls gyrus per se. There may be subtle differences in volumes in the adjacent cortical
regions, perhaps in areas 41 and 42 or the larger STG not measured by the current report.
However, even if there were smaller volumes in the adjacent cortex in SPD subjects, one
NIH-PA Author Manuscript
would not have predicted that smaller volumes lead to larger areas of activation, more likely,
smaller volumes would predict smaller areas of activation. Therefore, we do not believe that
possible volume differences in peri-Heschls regions can account for the between group
differences in activation patterns.
The possible relationship between abnormal auditory sensory processing and clinical
measures was also investigated. Exploratory correlations between ROI and clinical/cognitive
symptoms suggest that these early auditory sensory processing problems may have
implications for downstream language functioning. Specifically, in SPD subjects, there was
a correlation between abnormal activation on the right during the pitch experiment with odd
speech. Odd speech and formal thought disorder likely represent a symptom continuum. Our
laboratory previously demonstrated thought disorder correlating with left STG volumes in
female SPD subjects (Dickey et al., 2003). Similarly, in schizophrenic subjects, correlations
have been shown between thought disorder and hallucinations with bilateral measures of
volume and function of temporal lobe regions (Shenton et al., 1992; Dierks et al., 1999;
Kircher et al., 2001; Barta et al., 1990; Woodruff et al., 1997). In this report, however, the
correlation with odd speech occurred with right-sided activation. Odd speech in SPD may be
characterized by impoverished prosody (Dickey et al., unpublished data). Consensus
NIH-PA Author Manuscript
suggests a right hemisphere advantage for emotional (non-semantic) aspects of language and
prosody (Mitchell and Crow, 2005), particularly in the right STS and MTG, whereas more
semantic aspects of language recruit more left-sided regions (Mitchell et al., 2003). This
hemispheric pattern has been seen using fMRI (Mitchell et al., 2003) (Wildgruber et al.,
2005), transcranial Doppler ultrasonography (Vingerhoets et al., 2003), repetitive
transcranial magnetic stimulation (van Rijin et al., 2005), event-related potential (Eckstein
and Friederici, 2005), and in lesion studies (Pell, 2006). Therefore, it is possible, that the
tone processing impairment on the right in the SPD subjects is related to non-semantic
aspects of their odd speech.
Other aspects of language, specifically verbal learning, also correlated with inappropriate
activation of right-sided ROI for the current pitch experiment. Again, this is similar to
findings in schizophrenia. Using PET, Ragland et al. demonstrated impaired verbal learning
correlating with STG activation abnormalities (Ragland et al., 2001).
There are several limitations to the current study. One limitation is the large number of
correlations performed as discussed in the Methods section. However, including these
exploratory findings in this paper may serve to generate future hypotheses about the
NIH-PA Author Manuscript
scanner would not scan during stimuli presentation, would more fully remove the confound
of interference from scanner noise. Unfortunately, such a design would also significantly
increase the scanning time (a mean of 14 s between stimuli as proposed by Amaro et al.,
2002). In addition, one cannot modulate scanner frequency or pitch, thus limiting the
exploration of differential pitch discrimination in SPD subjects. Sixth, there is no behavioral
output measure during the experiments. This limits the ability to determine what the subject
is doing while hearing the tones and, therefore, limits the ability to draw firm interpretations
of the results. That is, one cannot definitively conclude that the SPD subjects exhibit
inefficient processing vs. exaggerated responses to deviant tones. However, one can state
that it appears that SPD subjects recruit more neurons while hearing simple tones than
comparison subjects. Having a behavioral response, however, would have added a layer of
complexity by introducing other potential confounds such as SPD subjects ability to make
decisions, processing speed, and motor speed. Finally, one limitation common to the fMRI
literature is that data were collected for each subject on only one time point. Recent work
has shown marked intersession variability for the extent of activation emanating from
listening to tones or word pairs (Yoo et al., 2005).
Nonetheless, these data suggest that neuroleptic-nave SPD subjects, compared with
NIH-PA Author Manuscript
Acknowledgments
Role of funding source
This study was funded by awards to CCD (VA Advanced Career Development Award and Brigham & Womens
Hospital Translational Neuroscience Award), and to RWM (NIMH RO1 MH52807). The VA, BWH, and NIMH
had no involvement in the study design, data collection or interpretation, nor in the writing or submitting of the
report.
We thank Dr. Dean Salisbury for his careful review of this manuscript. We thank Marie Fairbanks, Nancy Maxwell,
and Lori Benjamin for their administrative support.
References
Alho K. Cerebral generators of mismatch negativity (MMN) and its magnetic counterpart (MMNm)
NIH-PA Author Manuscript
Impaired mismatch negativity generation in prodromal subjects and patients with schizophrenia.
Schizophr Res. 2005; 73(23):297310. [PubMed: 15653275]
Brooker BH, Cyr JJ. Tables for clinicians to use to convert WAIS-R short forms. J Clin Psychol. 1986;
42:983986.
Cadenhead K, Light G, Geyer M, Braff D. Sensory gating deficits assessed by the P50 event-related
potential in subjects with schizotypal personality disorder. Am J Psychiatry. 2000; 157 (1):5559.
[PubMed: 10618013]
Cleghorn JM, Franco S, Szechtman B, Kaplan RD, Szechtman H, Brown GM, Nahmias C, Garnett ES.
Toward a brain map of auditory hallucinations [see comments]. Am J Psychiatry. 1992; 149 (8):
10621069. [PubMed: 1353314]
David AS, Woodruff PW, Howard R, Mellers JD, Brammer M, Bullmore E, Wright I, Andrew C,
Williams SC. Auditory hallucinations inhibit exogenous activation of auditory association cortex.
NeuroReport. 1996; 7 (4):932936. [PubMed: 8724677]
Delis, D.; Kramer, J.; Kaplan, E.; Ober, B. California Verbal Learning Test ManualResearch
Edition. The Psychological Corporation; San Diego: 1987.
Dickey CC, McCarley RW, Voglmaier MM, Niznikiewicz MA, Seidman LJ, Hirayasu Y, Fischer I,
Teh EK, Van Rhoads R, Jakab M, Kikinis R, Jolesz FA, Shenton ME. Schizotypal personality
disorder and MRI abnormalities of temporal lobe gray matter. Biol Psychiatry. 1999; 45 (11):
NIH-PA Author Manuscript
Javitt D, Shelley A-M, Ritter W. Associated deficits in mismatch negativity generation and tone
matching in schizophrenia. Clin Neurophysiol. 2000; 111:17331737. [PubMed: 11018486]
Javitt DC, Doneshka P, Grochowski S, Ritter W. Impaired mismatch negativity generation reflects
widespread dysfunction of working memory in schizophrenia. Arch Gen Psychiatry. 1995; 52 (7):
550558. [PubMed: 7598631]
Kasaki K, Nakagome K, Itoh K, Koshida I, Hata A, Iwanami A, Fukuda M, Kato N. Impaired cortical
network for preattentive detection of change in speech sounds in schizophrenia: a high-resolution
event-related study. Am J Psychiatry. 2002; 159 (4):546553. [PubMed: 11925291]
Kendler KS, McGuire M, Gruenberg AM, OHare A, Spellman M, Walsh D. The Roscommon family
study I. Methods, diagnosis of probands, and risk of schizophrenia in relatives. Arch Gen
Psychiatry. 1993; 50:527540. [PubMed: 8317947]
Kety, SS.; Rosenthal, D.; Wender, PH.; Schulsinger, F. The types and prevalence of mental illness in
the biological and adoptive families of adopted schizophrenics. Second Research Conference of
the Foundations Fund for Research in Psychiatry; Dorado, Puerto Rico: Pergamon Press; 1967.
Kimble M, Lyons M, ODonnell B, Nestor P, Niznikiewicz M, Toomey R. The effect of family status
and schizotypy on electrophysiologic measures of attention and semantic processing. Biol
Psychiatry. 2000; 47 (5):402412. [PubMed: 10704952]
Kircher TT, Liddle PF, Brammer MJ, Williams SC, Murray RM, McGuire PK. Neural correlates of
NIH-PA Author Manuscript
Lange N, Giedd JN, Castellanos FX, Vaituzis AC, Rapoport JL. Variability of human brain structure
size: ages 420 years. Psych Res. 1997; 74(1):112.
Leitman DI, Foxe JJ, Butler PD, Saperstein A, Revheim N, Javitt D. Sensory contributions to impaired
NIH-PA Author Manuscript
Mitchell RL, Crow TJ. Right hemisphere language functions and schizophrenia: the forgotten
hemisphere? Brain. 2005; 128 (Pt 5):963978. [PubMed: 15743870]
Mitchell RLC, Elliott R, Barry M, Cruttenden A, Woodruff PWR. The neural response to emotional
prosody, as revealed by functional magnetic resonance imaging. Neuropsychologia. 2003;
41:14101421. [PubMed: 12757912]
Molholm S, Martinez A, Ritter W, Javitt DC, Foxe JJ. The neural circuitry of pre-attentive auditory
change-detection: an fMRI study of pitch and duration mismatch negativity generators. Cereb
Cortex. 2005; 15 (5):545551. [PubMed: 15342438]
Morosan P, Rademacher J, Schleicher A, Amunts K, Schormann T, Zilles K. Human primary auditory
cortex: cytoarchi-tectonic subdivisions and mapping into a spatial reference system. NeuroImage.
2001; 13:684701. [PubMed: 11305897]
Naatanen R. The role of attention in auditory information processing as revealed by event-related
potentials and other measures of cognitive function. Behav Brain Sci. 1990; 13:201288.
Naatanen R, Paavilainen P, Rinne T, Alho K. The mismatch negativity (MMN) in basic research of
central auditory processing: a review. Clin Neurophysiol. 2007; 118(12):25442590. [PubMed:
17931964]
Niznikiewicz M, Shenton M, Voglmaier M, Seidman L, Dickey C, Rhoads R, Teh E, McCarley R.
Electrophysiological correlates of language abnormality in schizotypal personality disorder. Am J
Psychiatry. 1999; 156:10521058. [PubMed: 10401451]
NIH-PA Author Manuscript
Niznikiewicz MA, Voglmaier MM, Shenton ME, Dickey CC, Seidman LJ, Teh E, Van Rhoads R,
McCarley RW. Lateralized P3 deficit in schizotypal personality disorder. Biol Psychiatry. 2000;
48 (7):702705. [PubMed: 11032982]
Park HJ, Levitt JJ, Shenton ME, Salisbury DF, Kubiki M, Kikinis R, Jolesz FA, McCarley R. An MRI
study of spatial probability brain map differences between first episode schizophrenia and normal
controls. NeuroImage. 2004; 22 (3):12311246. [PubMed: 15219595]
Pell MD. Cerebral mechanisms for understanding emotional prosody in speech. Brain Lang. 2006;
96(2):221234. [PubMed: 15913754]
Ragland JD, Gur RC, Raz J, Schroeder L, Kohler CG, Smith RJ, Alavi A, Gur RE. Effect of
schizophrenia on fronto-temporal activity during word encoding and recognition: a PET cerebral
blood flow study. Am J Psychiatry. 2001; 158 (7):11141125. [PubMed: 11431234]
Salisbury D, Voglmaier M, Seidman L, McCarley R. Topographic abnormalities of P3 in schizotypal
personality disorder. Biol Psychiatry. 1996; 40:165172. [PubMed: 8830949]
and structural evidence of post-onset brain reduction in schizophrenia. Arch Gen Psychiatry. 2007;
64(5):521529. [PubMed: 17485604]
Shalgi S, Deouell LY. Direct evidence for differential roles of temporal and frontal components of
auditory change detection. Neuropsychologia. 2007; 45 (8):18781888. [PubMed: 17239410]
Shelley A, Ward P, Catts S, Michie P, Andrews S, McConaghy N. Mismatch negativity: an index of a
preattentive processing deficit in schizophrenia. Biol Psychiatry. 1991; 30:10591062. [PubMed:
1756198]
Shenton ME, Kikinis R, Jolesz FA, Pollak SD, LeMay M, Wible CG, Hokama H, Martin J, Metcalf D,
Coleman M, et al. Abnormalities of the left temporal lobe and thought disorder in schizophrenia. A
quantitative magnetic resonance imaging study. N Engl J Med. 1992; 327(9):604612. [PubMed:
1640954]
Siever LJ, Davis KL. The pathophysiology of schizophrenia disorders: perspectives from the spectrum.
Am J Psychiatry. 2004; 161 (3):398413. [PubMed: 14992962]
Steinmetz H, Rademacher JHY. Cerebral asymmetry: MR planimetry of the human planum temporale.
J Comput Assist Tomogr. 1986; 13(6):9961005. [PubMed: 2584512]
Stephan KE, Magnotta VA, White T, Arndt S, Flaum M, OLeary DS, Andreasen NC. Effects of
olanzapine on cerebellar functional connectivity in schizophrenia measured by fMRI during a
simple motor task. Psychol Med. 2001; 31(6):10651078. [PubMed: 11513374]
NIH-PA Author Manuscript
Sweet RA, Dorph-Petersen KA, Lewis DA. Mapping auditory core, lateral belt, and parabelt cortices
in the human superior temporal gyrus. J Comp Neurol. 2005; 491(3):270289. [PubMed:
16134138]
Trestman RL, Horvath T, Kalus O, Peterson AE, Coccaro E, Mitropoulou V, Apter S, Davidson M,
Siever LJ. Event-related potentials in schizotypal personality disorder. J Neuropsychiatry Clin
Neurosci. 1996; 8(1):3340. [PubMed: 8845699]
Umbricht D, Javitt D, Novak G, Bates J, Pollack S, Lieberman J, Kane J. Effects of clozapine on
auditory event-related potentials in schizophrenia. Biol Psychiatry. 1998; 44:716725. [PubMed:
9798075]
van Rijin S, Aleman A, van Diessen E, Berckmoes C, Vingerhoets G, Kahn RS. What is said or how it
is said makes a difference: role of the right fronto-parietal operculum in emotional prosody as
revealed by repetitive TMS. Eur J Neurosci. 2005; 21:31953200. [PubMed: 15978028]
Vingerhoets G, Berckmoes C, Stroobant N. Cerebral hemodynamics during discrimination of prosodic
and semantic emotion in speech studied by transcranial Doppler ultrasonography.
Neuropsychology. 2003; 17 (1):9399. [PubMed: 12597077]
Voglmaier M, Seidman L, Salisbury D, McCarley R. Neuropsychological dysfunction in schizotypal
personality disorder: a profile analysis. Biol Psychiatry. 1997; 41:530540. [PubMed: 9046985]
Voglmaier M, Seidman L, Niznikiewicz M, Dickey C, Shenton M, McCarley R. Verbal and nonverbal
neuropsychological test performance in subjects with schizotypal personality disorder. Am J
NIH-PA Author Manuscript
Yoo SS, OLeary HM, Dickey CC, Wei XC, Guttmann CR, Park HW, Panych LP. Functional
asymmetry in human primary auditory cortex: identified from longitudinal fMRI study. Neurosci
Lett. 2005; 383(12):16. [PubMed: 15936503]
NIH-PA Author Manuscript
NIH-PA Author Manuscript
NIH-PA Author Manuscript
Fig. 1.
Diagrams of stimuli presentation and processing. a. For the pitch experiment all tones are
100 ms in duration with 200 ms of silence before the next tone. b. For the duration
experiment, the standard tone is 50 ms in duration with 250 ms of silence before the next
tone. c. For the duration experiment the deviant tone is 200 ms in duration with 100 ms of
silence before the next tone. This variation decreases the expectancy factor. d. Tones were
presented in block design. e. Hemodynamic response curves were generated for all tones
together (both standard and deviant, all tone condition) and for only the deviant tones
(deviant condition). ISI = interstimulus interval; ms = milliseconds; s = standard; s/d =
standard and deviant tones intermixed; rest = silence.
NIH-PA Author Manuscript
NIH-PA Author Manuscript
Fig. 2.
Whole-brain statistical parametric maps with consistent p scale shown. Note that regardless
of experiment (pitch or duration) and regardless of tone set heard (all tones or deviants
only), there is no area of the brain in which control subjects activated more than SPD
subjects. a. Pitch experiment, all tones (standard+deviant) heard analyzed together. b. Pitch
experiment deviant condition. Note significantly more activation in SPD subjects compared
NIH-PA Author Manuscript
with control subjects in the region of the STG bilaterally. Thresholded at the p<0.001 level
with an extent of activation cluster size limit set at 240 voxels. MNI coordinates and cluster
significance by two-tailed t tests were: (52, 22, 2, T=5.3, p<0.001), (54, 46, 2, T=5.3,
p<0.001), (54, 46, 18, T=4.85, p<0.001). c. Duration experiment, all tones (standard
+deviant) condition. Thresholded at the p<0.01 level with an extent of activation cluster size
limit set at 0 voxels to ensure all visible super-threshold activation. MNI coordinates and
cluster significance by two-tailed t tests were: (2, 22, 10, T=5.67, p<0.0005) and (14, 4, 24,
T=5.3, p<0.0005). d. Duration experiment, deviant condition. SPD subjects activated more
to deviant tones than comparison subjects in left temporal and bilateral parietal regions.
Thresholded at the p<0.01 level with an extent of activation cluster size limit set at 240
voxels for temporal lobe regions, 0 voxels for parietal regions in order not to lose relevant
findings in that area. MNI coordinates and cluster significance by two-tailed t tests were: for
temporal lobe regions, (48, 28, 4, T=2.83, p=0.005 and 22, 88, 2, T=3.1, p=0.003);
for parietal lobe regions, (26, 42, 38, T=4.0, p<0.001), (24, 20, 42, T=3.2, p=0.002),
(28, 40, 38, T=3.36, p=0.001). Note standardized color bars across images reflect relative T
scores. Neurological convention (left is left).
NIH-PA Author Manuscript
Fig. 3.
Scatterplots of peak activation and clinical and cognitive measures. a. SPD subjects with
greater impairment due to odd thinking or speech (higher score) demonstrated more
activation while passively hearing tones in BA 41 and 42 on the right. b. SPD subjects who
learned fewer words on the CVLT had greater peak activation in the right BA 41, 42, and
STG. BA = Brodmann Area.
NIH-PA Author Manuscript
NIH-PA Author Manuscript
Table 1
Subject demographics and absolute Heschls gyrus volumes in milliliters
NIH-PA Author Manuscript
SPD NC p
N 13 13
Age 36.8 (10.8) 30.4 (11.4) <0.2
PSES 4.0 (1.2) 3.8 (1.4) 0.08
SES 3.0 (1.4) 3.2 (1.4) 0.7
IQ 116 (9.7) 120.4 (12.4) <0.4
Education 15.2 (1.9) 18.0 (3.8) 0.02
SSPT score 55.2 56.2 (N=8) <0.5
Note that statistics were performed on regressed not absolute volumes in order to account for confounding effect of head size. PSES = parental
socio-economic status, SES = socio-economic status, SSPT = Speech Sound Perception Test.
a
F(1,24)=0.109.
b
NIH-PA Author Manuscript
F(1,24)=1.128.
NIH-PA Author Manuscript
Table 2
Clinical/cognitive/functional correlations
NIH-PA Author Manuscript
Correlations between peak t activation values for the ROI, namely, STG, Brodmann Areas 41 and 42, with clinical criteria, and verbal learning
(CVLT) are given.
NIH-PA Author Manuscript
NIH-PA Author Manuscript