SCHRES-06321; No of Pages 8

Schizophrenia Research xxx (2015) xxxxxx

Contents lists available at ScienceDirect

Schizophrenia Research
journal homepage: www.elsevier.com/locate/schres

Language-related gamma EEG frontal reduction is associated with

positive symptoms in schizophrenia patients
Chiara Spironelli a,b,, Alessandro Angrilli a,b,c
a
b
c

Department of General Psychology, University of Padova, via Venezia 8, 35131 Padova, Italy
CCN, Center for Cognitive Neuroscience, via Venezia 8, 35131 Padova, Italy
CNR Neuroscience Institute, Padova Section, via G. Colombo 3, 35121 Padova, Italy

a r t i c l e

i n f o

Article history:
Received 30 October 2014
Received in revised form 27 February 2015
Accepted 5 April 2015
Available online xxxx
Keywords:
Gamma EEG rhythm
Psychosis
Delusions
Hallucination
Lateralization

a b s t r a c t
Objectives: Frontal hypoactivation has been consistently found in schizophrenia. We hypothesized that patients'
decit is asymmetrical, i.e., centred over the left frontal locations, associated with loss of language-related
asymmetry, and correlated with positive symptoms.
Method: The amplitude of EEG gamma band (3648 Hz) was measured during the processing of three linguistic
(Phonological vs. Semantic vs. Visuo-perceptual) tasks and used as index of activation/connectivity in 18 schizophrenia patients and 18 healthy participants.
Results: Healthy controls showed higher gamma in frontal sites, revealing a signicantly greater left vs. right
asymmetry in all linguistic tasks, whereas patients exhibited decreased and bilateral gamma amplitude
(i.e., reduced activation/connectivity) in frontal regions. The patients' left hypofrontality during phonological
processing was positively correlated with higher levels of Delusions (P1) and Hallucination (P3) PANSS subscales. A signicantly greater left posterior gamma amplitude was found in patients compared with controls.
Conclusion: Results suggest, in schizophrenia patients, a functional decit of left frontal regions including Broca's
area, a key site playing a fundamental hierarchical role between and within hemispheres which integrates many
basic processes in linguistic and conceptual organization. The signicant correlation between lack of the left anterior asymmetry and increased positive symptoms is in line with Crow's hypothesis postulating the aetiological
role of disrupted linguistic frontal asymmetry on the onset of the key symptoms of schizophrenia.
2015 Elsevier B.V. All rights reserved.

1. Introduction
Abnormal brain organization and functioning in schizophrenia
patients have been consistently investigated, one of the most important and reliable nding being a signicant dysfunction in the
activation of the prefrontal cortex, which is typically associated
with decits in attention, action planning and working memory
(Andreasen et al., 1992; Goldberg and Gold, 1995; Wienberger and
Berman, 1996; Tan et al., 2007 for an extensive review). Schizophrenia patients' hypofrontality has been especially demonstrated with
electrophysiological/neuromagnetic methods analysing slow and
fast EEG/MEG rhythms, and dipole density (Tauscher et al., 1998;
Fehr et al., 2001; Mientus et al., 2002). Most EEG studies carried
out in schizophrenia patients found, compared with controls, both increased levels of delta activity (0.13.9 Hz), and reduced delta coherence
in frontal sites, which was usually inversely correlated with the PANSS
positive syndrome subscale, thus suggesting a direct link between the
severity of positive symptoms and patients' hypofrontality (Tauscher
Corresponding author at: Department of General Psychology, University of Padova, via
Venezia 8, 35131 Padova, Italy. Tel.: +39 049 827 6619; fax: +39 049 827 6600.
E-mail address: chiara.spironelli@unipd.it (C. Spironelli).

et al., 1998). Whereas past literature on slow EEG rhythms provided

converging evidence that increased delta EEG activity is a quantitative
and reliable index of schizophrenia patients' frontal inhibition, most of
past research measuring beta (2035 Hz) and gamma (3650 Hz) oscillations during auditory or visual working memory tasks revealed less
consistent results (Kwon et al., 1999; Basar-Eroglu et al., 2007; Barr
et al., 2010). In particular, increased or reduced gamma activity has
been found depending on modality of stimulus presentation (auditory
vs. visual), levels of task complexity (and the extent of working memory
engagement) as well as prevalent symptom content (positive vs. negative; Herrmann and Demiralp, 2005). Research carried out by measuring
high-frequency neural oscillations in schizophrenia patients and healthy
controls includes various visual perceptual tasks (e.g., Tan et al., 2013),
or working memory paradigms mainly based on the N-back task
(Basar-Eroglu et al., 2007; Barr et al., 2010). In these studies, the main
result was a signicant greater frontal gamma activity in patients
regardless of cognitive load (low vs. high) manipulations (BasarEroglu et al., 2007; Barr et al., 2010), the N-back performance being
negatively correlated with negative symptoms (Barr et al., 2010).
In three previous studies we used a working memory paradigm, validated for language (Spironelli and Angrilli, 2010), in order to assess
Crow's hypothesis (1997, 2000) on schizophrenia disorder. By analysing

http://dx.doi.org/10.1016/j.schres.2015.04.003
0920-9964/ 2015 Elsevier B.V. All rights reserved.

Please cite this article as: Spironelli, C., Angrilli, A., Language-related gamma EEG frontal reduction is associated with positive symptoms in
schizophrenia patients, Schizophr. Res. (2015), http://dx.doi.org/10.1016/j.schres.2015.04.003

C. Spironelli, A. Angrilli / Schizophrenia Research xxx (2015) xxxxxx

automatic recognition potential (Spironelli et al., 2008), Contingent

Negative Variation (Angrilli et al., 2009) and delta hypofrontality
(Spironelli et al., 2011), we found evidence, in schizophrenia patients,
of a signicant lack of left-hemisphere dominance for language as a consequence of their primary decit in functional linguistic integration.
This was in line with prior literature showing lack of left hemisphere
dominance for linguistic stimuli (e.g., Bruder et al., 1999; Rockstroh
et al., 2001; McCarley et al., 2002) Therefore, the inter- and intrahemispheric hierarchical roles of Broca's area (Bookheimer, 2002;
Hagoort, 2005) would explain the lack of linguistic integration and the
metalinguistic disorganization observed in schizophrenia (Crow 1997,
2000; Kuperberg, 2010a,b). In turn, the decit in linguistic dominance,
leads to lack of hemispheric integration, confusion between inner and
external voices (hallucinations), thought disorders and delusions
(Crow 1997, 2000; Kuperberg, 2010a,b). From a more general point of
view, in agreement with Andreasen et al. (1999), schizophrenia patients
are characterized by abnormal integration of brain networks: for this
reason, gamma EEG band should represent the most appropriate marker for studying patients' dysfunctional cortical integration. The present
experiment analysed the amplitude of gamma EEG band as an index
of cognitive functioning and integration mechanisms (Herrmann et al.,
2004): compared with controls, schizophrenia patients were expected
to exhibit lower gamma amplitude over anterior left cortical sites,
i.e., in regions currently considered essential for the organization of
the whole linguistic network. In addition, we expected to nd a correlation between schizophrenia key psychiatric indices (i.e., delusions and
hallucinations measured by the Positive and Negative Syndrome Scale,
PANSS; Kay et al., 1987) and the frontal linguistic left asymmetry, in
line with Crow's etiological hypothesis (1997, 2000) on the main
mechanism postulated at the origin of the schizophrenic disorder, the
disruption of the typical language-related asymmetry measured in
healthy individuals.

2. Materials and methods

2.1. Participants
The psychiatric group consisted of 18 schizophrenia inpatients
(4 women, 14 men; mean age SD: 39.11 11.05 years, range 24
70 years; educational level: 10.11 2.70 years, range 717 years) recruited from the Ospedale Psichiatrico Giudiziario of Castiglione delle
Stiviere (Mantova, Italy) according to the following criteria: all patients
were right-handed (Oldeld, 1971); they had been diagnosed as schizophrenic during the acute phase, on the basis of positive or negative
symptoms exhibited for more than 6 months according to DSM-IV-R
criteria; at the time of the study, all patients were in a chronic state
(average time from onset: 14 8.57 years). The diagnosis, ascertained
by the psychiatrists of the ward at the time of the experiment by administering Structured Clinical Interview for DSM Disorders, classied two
patients as disorganized (ICD-10 F20.1), two with paranoid/residual
(F20.0/F20.5) and fourteen with paranoid schizophrenia (F20.0). In
addition, prior to the experimental session, schizophrenia patients
were screened to assess the severity of symptoms according to the
Italian version of PANSS (Table 1).
Six patients were treated with typical antipsychotic drugs
(i.e., chlorpromazine, clotiapine, clucopenthixol, haloperidol and
methotrimeprazine), six patients with atypical antipsychotic
drugs (i.e., aripiprazole, clozapine, olanzapine, quetiapine, risperidone), and six patients with both typical and atypical antipsychotic
drugs.
The control group consisted of 18 right-handed healthy volunteers
(7 women, 11 men; 2(1) = 1.18, ns), matched for age (mean SD:
42.39 19.80 years, range 2169 years; t(34) = 0.61, ns), educational
level (11.72 3.00 years, range 817 years; t(34) = 1.69, ns) and handedness (98.47 3.55%; t(34) = 1.08, ns) to the patient group.

Table 1
Demographic characteristics of healthy controls and schizophrenia patients, and average
scores obtained from the Italian version of PANSS (Kay et al., 1987), administered to ascertain patients' symptoms' severity.
Healthy controls
Mean
Age (years)
Gender
Education (years)
Handedness
Years from onset

SD

42.39
3.00
7 females 11 males
11.72
3.00
98.47%
3.55%

Schizophrenia
patients
Mean

Negative symptoms
N1 (blunted affect)
N2 (emotional withdrawal)
N3 (poor rapport)
N4 (passive/apathetic social withdrawal)
N5 (difculty in abstract thinking)
N6 (lack of spontaneity and ow of
conversation)
N7 (stereotyped thinking)
Total

t(34) = 0.61
2(1) = 1.18
t(34) = 1.69
t(34) = 1.08

ns
ns
ns
ns

SD

39.11
11.05
4 females 14 males
10.11
2.70
93.05% 10.01%
14.00
8.57

Positive And Negative Syndrome Scale (PANSS)

Positive symptoms
P1 (delusions)
4.56
P2 (conceptual disorganisation)
3.56
P3 (hallucinatory behaviour)
2.94
P4 (excitement)
2.83
P5 (grandiosity)
3.06
P6 (suspiciousness/persecution)
4.11
P7 (hostility)
2.78
Total
3.40

Test

1.58
1.29
1.63
1.38
1.86
1.64
1.40
1.65

4.28
4.44
4.11
4.22
4.28
3.78

0.89
1.04
0.83
1.17
1.27
1.00

4.28
4.20

1.02
1.04

Both healthy adults and patients signed their informed consent to

participate in this study, which was approved by the Ethics Committee
of the Department of General Psychology and the local Ethics Committee of the Ospedale Psichiatrico Giudiziario, and performed in accordance
with the ethical standards laid down in the Declaration of Helsinki.

2.2. Stimuli, tasks and procedure

Stimuli consisted of bi- or trisyllabic Italian content words selected
from a frequency dictionary of 5000 written Italian words, and presented in pairs on a 17 computer monitor one at a time with an InterStimulus Interval (ISI) of 2 s: thus, the rst word (W1) remained on
the screen for 1 s and, after an interval of 2 s, the second word (W2, or
target) appeared on the screen, until the subject responded by pressing
a keyboard button, in no case longer than 5 s (11). Word pairs were administered in three separate blocks, which corresponded to three linguistic tasks: according with the Hillyard principle, the same words
were presented as W1, but in different randomized order across tasks.
Upon W2-target presentation, participants had to decide whether
word pairs rhymed (Phonological task), whether target word W2 was
of the same category as W1 (Semantic task), and whether word pairs
were written in the same upper or lower case1 (Visuo-perceptual, control task; Spironelli and Angrilli, 2010). For motor responses, subjects
used their left index or middle nger to press the keyboard buttons
corresponding to matchmismatch conditions. Each task included 80
word-pairs, 50% matches being randomly interspersed with 50%
mismatch trials; task order was randomly varied across participants.
Since this paradigm requires efcient verbal working memory skills,
we administered the forward digit span test to verify that groups had
similar levels of working memory.

All the letters in the same word were upper or lower case.

Please cite this article as: Spironelli, C., Angrilli, A., Language-related gamma EEG frontal reduction is associated with positive symptoms in
schizophrenia patients, Schizophr. Res. (2015), http://dx.doi.org/10.1016/j.schres.2015.04.003

C. Spironelli, A. Angrilli / Schizophrenia Research xxx (2015) xxxxxx

2.3. Data recording and analysis

Behavioural measures included response times and error rates to the
second stimulus, and mean performance was compared between
groups, among tasks. Electrophysiological activity was continuously recorded in DC mode by 38 tin electrodes, 31 placed on an elastic cap according to the International 1020 system (Oostenveld and Praamstra,
2001); the other seven applied below each eye (Io1, Io2), on the two external canthi (F9, F10), nasion (Nz) and mastoids (M1, M2). Amplitude
resolution was 0.1 V; bandwidth ranged from DC to 100 Hz (6 dB/
octave). Sampling rate was set at 500 Hz, and impedance kept below
5 K. All cortical sites were online referred to Cz, and ofine rereferenced to the average reference.
In this study, analysis is focused on induced gamma activity which
was measured by traditional Fast Fourier Transform (FFT). The FFT
was performed after windowing the EEG signal with a cosine-tapered
window, and included three 1024-ms intervals of 512 samples/lines
corresponding to 0.977 Hz resolution. Given the constraint of the Fast
Fourier Transform (FFT) to use 2n samples, the width of each interval
was necessarily forced to 512 samples, corresponding to a 1024-ms interval. Each interval represented a different processing phase required
by tasks: 1024 ms after W1 onset (W1 interval); 10002024 ms after
W1 onset (initial Inter-Stimulus-Interval); and 19763000 ms after
W1 onset (terminal Inter-Stimulus-Interval). As eye movement artifacts
may affect EEG amplitude, particularly over frontal locations, each trialepoch was corrected for blinks and eye movements according to the
methods of Ille et al. (2002), by means of BESA software. Artifact rejection was automatically applied to all epochs, with amplitude (150 V)
and derivative (100 V/ms) thresholds. Remaining epochs were visually
inspected for any residual artifacts, and to ensure that all trials corresponding to wrong responses were discarded, so that a total of 6.11%
and 17.08% of trials was rejected for controls and patients, respectively,
evenly distributed across tasks. For each participant, FFT was performed
on all artifact-free epochs, which were averaged for each interval and
task. The last step consisted of normalizing gamma band amplitude
(nominally 3648 Hz, effective range: 36.14947.879 Hz2) for all sites,
by computing the percentage of gamma amplitude for each electrode
in the 0.977100 Hz spectrum range. After transformation, the KolmogorovSmirnov test was applied to ensure that EEG normalized
gamma activity was normally distributed (d = 0.08576, ns). Therefore,
EEG data underwent analysis of variance (ANOVA) to compare the average gamma amplitude measured by four groups of ve electrodes,
representing different regions of interest: Anterior Left (Io1, F7, FT7,
F3, FC3), Anterior Right (Io2, F8, FT8, F4, FC4), Posterior Left (CP3, P3,
P7, TP7, T7) and Posterior Right (CP4, P4, P8, TP8, T8).
Since less consistent ndings for gamma could at least in part
arise from the greater likelihood of confounding by activity from muscle
artifacts, we carried out an additional analysis of a very high frequency
EEG band, detecting mainly muscular activity (EMG band, nominally
80100 Hz, effective range: 80.11499.654 Hz). Indeed, since EMG provides a wide band spectral response (10250 Hz; Van Boxtel et al.,
1984), when signicant activity and patterns are detected both on
high-end EEG bands and in all lower bands, this may indicate that
some EMG contamination is present in the EEG of interest. In addition,
in order to establish whether the pattern of lateralization/lack of lateralization expected was specic for gamma amplitude rather than a general effect evident throughout the whole EEG activity, we also analysed
two slow frequency rhythms, i.e., theta and alpha amplitudes (nominally 48 and 813 Hz, effective range: 4.8857.816 and 8.79312.724 Hz,
respectively).
As for gamma band, the EMG band, theta and alpha bands have been
normalized, then tested to ensure a normal distribution (Kolmogorov
2
More precisely, gamma amplitude was obtained by the sum of all values within the
gamma spectral band excluding the 50 Hz frequency line to avoid possible noise artifact
due to main power.

Smirnov d = 0.10202, d = 0.08392 and d = 0.08985, all ns, respectively), and clustered in the same four regions of interest.
With regard to behavioural measures, forward digit spans from controls and patients were analysed using t test for independent groups.
Considering participants' performance on experimental tasks, mean
response times and error rates have been analysed with separate
ANOVAs, which included the between-subjects factor Group (two
levels: Controls vs. Patients) and within-subjects factor Task (three
levels: Visuo-perceptual vs. Phonological vs. Semantic task).
Since preliminary analyses carried out on gamma amplitude revealed no effects related to the time course (i.e., neither the main effect
nor interactions involving the temporal interval factor were signicant),
in agreement with Spironelli et al. (2013), we decided to collapse the
three time intervals following the rst word onset by computing their
mean amplitude. This solution allows increasing statistical power and
improving data presentation, making result interpretation easier.
ANOVA was therefore carried out on EEG data (i.e., gamma, EMG
band, theta and alpha amplitude) including Group (two levels: Controls
vs. Patients), Task (three levels: Visuo-perceptual vs. Phonological vs.
Semantic), Region (two levels: Anterior vs. Posterior) and Laterality factors (two levels: Left vs. Right hemisphere). The GreenhouseGeisser
correction was applied whenever sphericity assumption was violated;
in this case, uncorrected degrees of freedom, epsilon values and
corrected probability levels are listed. Post-hoc comparisons were
computed with NewmanKeuls tests at p b 0.05.
In addition, for the patient group only, Pearson's correlation analyses
were carried out between a-priori selected PANSS scores and gamma
amplitude achieved from the left anterior and posterior sites during
linguistic processing (i.e., Phonological and Semantic tasks), to test
whether specic positive symptoms Delusions (P1) and Hallucinatory
behaviour (P3) represent a behavioural correlate signicantly linked
with the gamma left hypofrontality. Since we had no a-priori hypotheses
on negative symptoms and gamma amplitude achieved from the left anterior and posterior sites during linguistic processing (i.e., Phonological
and Semantic tasks), we carried out an explorative correlation analyses
on all negative symptom subscales Blunted affect (N1), Emotional
withdrawal (N2), Poor rapport (N3), Passive/apathetic social withdrawal
(N4), Difculty in abstract thinking (N5), Lack of spontaneity and ow of
conversation (N6) and Stereotyped thinking (N7). Therefore, negative
correlations indicate that those patients with more severe symptoms
(higher scores on P1 and P3 constructs, or one or more negative
subscales) had also lower gamma amplitude in the left frontal cortical
regions (that is, specic frontal hypofunction), left posterior sites (that
is, specic temporo-parietal hypofunction) or both (that is, generalised
hypofunction of the whole left hemisphere).
3. Results
3.1. Behavioural data
Healthy adults and schizophrenia patients had similar scores to the
forward digit span test (6.44 1.25 vs. 5.67 1.28, respectively;
t(34) = 1.84, ns).
RTs showed the signicant main effect of Group (F(1,34) = 22.29,
p b 0.001), patients being slower than controls (1511.59 551.16 ms
vs. 910.81 250.12 ms, respectively), probably due to patients' pharmacological treatment. This pattern was evident also in the signicant
two-way Group by Task interaction (F(2,68) = 3.59, p b 0.05, GG =
0.99): patients had slower RTs than controls, regardless of task (all
p b 0.001; Fig. 1a). In addition, schizophrenia patients revealed longer
RTs for the Semantic than the Phonological task, which in turn induced
longer RTs than the Orthographic task (all p b 0.001), whereas controls
showed longer RTs for the Semantic than both Phonological and Visuoperceptual tasks (p b 0.05 and p b 0.001, respectively).
ANOVA carried out on ERs revealed signicant main effects of Group
(F(1,34) = 21.44, p b 0.001), with patients producing more errors than

Please cite this article as: Spironelli, C., Angrilli, A., Language-related gamma EEG frontal reduction is associated with positive symptoms in
schizophrenia patients, Schizophr. Res. (2015), http://dx.doi.org/10.1016/j.schres.2015.04.003

C. Spironelli, A. Angrilli / Schizophrenia Research xxx (2015) xxxxxx

Fig. 1. Scattergrams of the individual data. Behavioural data analyses revealed a signicant two-way Group by Task interaction in (a) Response Times (RTs) and (b) Error Rate (ER). Visuoperc. = Visuo-perceptual task; Phon. = Phonological task.

controls (7.97% 5.93% vs. 3.08% 2.26%, respectively). However, the

two-way Group by Task interaction (F(2,68) = 3.94, p b 0.05, GG =
0.69) revealed the same percentage of errors (b8%) in controls' tasks
(Fig. 1b), whereas schizophrenia patients showed signicantly higher
error rates in both Visuo-perceptual and Semantic than in the Phonological task (all p b 0.001). Considering between-group differences,
patients made more errors than controls in Visuo-perceptual and
Semantic tasks (all p b 0.001), but not in the Phonological one.

3.2. Electrophysiological data

Fig. 2 shows normalised gamma amplitude in healthy controls (left
panel) and schizophrenia patients (right panel) for each task: higher
levels of gamma are shown in red, lower in blue.
The visual inspection of spline maps suggested that controls had a
left-lateralized distribution of gamma amplitude, as shown by higher
levels in the anterior left vs. right sites, in particular during phonological
and visuo-perceptual processing. Instead, schizophrenia patients exhibited gamma amplitude bilaterally distributed in fronto-polar electrodes,
but higher levels in the left rather than right temporal sites, especially
during phonological and visuo-perceptual processing.
The ANOVA on gamma amplitude revealed the signicant main effect of Region (F(1,34) = 39.63, p b 0.001): regardless of group, higher
gamma amplitude was found in anterior (12.29% 1.48%) compared
with posterior sites (11.38% 1.56%). However, the two-way Group
by Region interaction (F(1,34) = 26.08, p b 0.001) showed that this pattern marked gamma amplitude distribution in controls (p b 0.001),
whereas patients had similar gamma amplitude distribution in anterior
and posterior sites (Fig. 3).
With respect to group differences, healthy controls had signicant
greater gamma amplitude than schizophrenia patients in anterior
(p b 0.001), not in posterior sites.
In addition, the four-way Group by Task by Region by Laterality interaction (F(2,68) = 4.37, p b 0.01, GG = 0.99) revealed specic
laterality patterns between groups. In anterior regions, controls showed
a signicantly higher left vs. right gamma amplitude in all tasks (all
p b 0.001; Fig. 4, top row). In addition, gamma amplitude measured
on the left sites during visuo-perceptual and phonological processing
was signicantly greater with respect to that obtained during Semantic
task (all p b 0.001). Instead, schizophrenia patients exhibited bilaterally
distributed gamma amplitude throughout tasks.
Furthermore, compared with healthy controls, patients showed
signicant lower gamma amplitude in frontal sites (all p b 0.001).

Concerning posterior regions, during phonological processing, controls had signicant higher gamma amplitude in the left vs. right sites
(p b 0.05; Fig. 4b, bottom row). No lateralization was found during
Visuo-perceptual or Semantic task. Instead, patients' gamma amplitude
was greater on the left vs. right sites (all p b 0.01) in all tasks. Compared
with healthy controls, schizophrenia patients exhibited signicant
greater gamma amplitude on the left posterior sites in all tasks (all
p b 0.001), and in the right locations only during phonological processing (p b 0.001).
The ANOVA on the EMG band detecting muscular activity revealed
no signicant main effects or interactions involving Group, Task and
Laterality factors.
The ANOVA on theta amplitude revealed a signicant Group by Region interaction (F(1,34) = 6.38, p b 0.05): higher theta amplitude
was found in controls' posterior (7.20% 2.57%) compared with anterior sites (5.89% 1.88%; p b 0.001), whereas no effect of region appeared
in schizophrenia patients, in which the levels of theta amplitude were
equally distributed on anterior and posterior sites (6.18% 1.77% and
6.57% 2.05%, respectively). Concerning group differences, healthy
controls had signicant greater theta amplitude than schizophrenia patients in posterior (p b 0.05), not in anterior sites. A similar (but not
identical) pattern was found for alpha amplitude: the signicant
Group by Region interaction (F(1,34) = 13.81, p b 0.001) revealed
higher alpha amplitude in posterior compared with anterior sites of
both controls (10.83% 3.99% vs. 7.01% 2.22%, respectively;
p b 0.001) and patients (8.70% 2.07% vs. 6.86% 1.59%, respectively;
p b 0.001). Similar to Theta band analysis, healthy controls had signicant greater alpha amplitude than schizophrenia patients in posterior
(p b 0.001), not in anterior sites. Thus, no other signicant main effects
or interactions involving Group, Task and Laterality factors were found
in theta and alpha EEG bands.
3.3. Pearson's correlations
This analysis provided important information for functional interpretation of schizophrenics' gamma distribution within the linguistic
domain. Only gamma amplitude measured from the anterior left cluster
during the phonological processing was negatively correlated with patients' scores to P1 and P3 symptom scales (r(16) = 0.50, p = 0.03
and r(16) = 0.52, p = 0.02, respectively). The lower the gamma
amplitude in the left anterior cluster, the higher the score achieved in
both Delusion and Hallucinatory behaviour domains (Fig. 5).
No signicant correlations were found with gamma amplitude
measured on the posterior left sites during linguistic tasks.

Please cite this article as: Spironelli, C., Angrilli, A., Language-related gamma EEG frontal reduction is associated with positive symptoms in
schizophrenia patients, Schizophr. Res. (2015), http://dx.doi.org/10.1016/j.schres.2015.04.003

C. Spironelli, A. Angrilli / Schizophrenia Research xxx (2015) xxxxxx

Fig. 2. Spline maps of normalized gamma amplitude for healthy controls (left side) and schizophrenia patients (right side). Note that the ranges of gamma percentages have been scaled to
groups' gamma distribution.

Concerning negative symptoms, no hypothesis was advanced in this

investigation, but reviewers asked this information for completeness.
Gamma band amplitude measured from the anterior left cluster during
the phonological processing was negatively correlated with N6 symptom scale (r(16) = 0.51, p = 0.03), whereas amplitude measured
from the same cluster during both the phonological and the semantic
tasks was negatively correlated with patients' scores in the N7 symptom
scale (r(16) = 0.60, p = 0.009 and r(16) = 0.47, p = 0.04, respectively). The lower the gamma amplitude in the left anterior cluster, the
higher Lack of spontaneity and ow of conversation (N6) and the
Stereotyped thinking (N7). No other signicant correlations were
found for gamma band on the left sites.

4. Discussion
The present study examined the functional alterations in language
lateralization in schizophrenia patients by analysing gamma amplitude an index of cognitive functioning and integration mechanisms
(Herrmann and Demiralp, 2005), being generated by those neurons engaged in high level cognitive processing, which re synchronously at
very high frequency and with low amplitudes (Herrmann et al.,
2004) in a validated linguistic paradigm (Spironelli and Angrilli,
2010).
To this aim, it is important to consider a critical precondition. To
be sure that the gamma activity analysed was not contaminated by

Please cite this article as: Spironelli, C., Angrilli, A., Language-related gamma EEG frontal reduction is associated with positive symptoms in
schizophrenia patients, Schizophr. Res. (2015), http://dx.doi.org/10.1016/j.schres.2015.04.003

C. Spironelli, A. Angrilli / Schizophrenia Research xxx (2015) xxxxxx

Fig. 3. Gamma band analysis: signicant two-way Group by Region interaction. Controls'
and patient's gamma amplitudes (black and dotted lines, respectively) are shown for
the anterior and posterior brain regions. Asterisks: signicant post-hoc comparisons.

confounding muscular activity, we carried out a supplemental analysis on the 80100 Hz EEG band, which may best indicate muscular
artifacts. No between-group differences and laterality effects were
found. At the same time, the analyses carried out on two slow
frequency rhythms, i.e., theta and alpha amplitudes, allowed us to
verify that the lack of lateralization was specic for gamma band
rather than due to a general effect present throughout the whole
EEG activity.
Considering gamma amplitude, as a rst important consideration,
the main effect of Group factor was not signicant: patients had no

Fig. 5. Pearson's correlations between Hallucinatory behaviour construct (P3) of Positive

Scale of the PANSS and the gamma anterior left amplitude recorded during Phonological
task.

overall gamma amplitude difference with respect to healthy controls.

This result suggests that groups had no overall differences in their cortical activation, notwithstanding patients were treated with different
types of antipsychotic drugs when they participated in the experiment.
However, when considering language-related activity, in patients' frontal regions gamma amplitude was substantially reduced and bilaterally
distributed, whereas in controls it showed a coherent pattern of the left
anterior lateralization across all tasks, in agreement with the rationale of
our paradigm which stresses on verbal working memory, and with past
fMRI studies that showed a signicant, language-related loss of the
frontal left lateralization not only in schizophrenia-spectrum patients,

Fig. 4. Gamma band analysis: signicant four-way Group by Task by Region by Laterality interaction. (a) Visuo-perceptual, (b) Phonological and (c) Semantic tasks in controls (black line)
and schizophrenia patients (dotted black line) in anterior (top panel) and posterior sites (bottom panel).

Please cite this article as: Spironelli, C., Angrilli, A., Language-related gamma EEG frontal reduction is associated with positive symptoms in
schizophrenia patients, Schizophr. Res. (2015), http://dx.doi.org/10.1016/j.schres.2015.04.003

C. Spironelli, A. Angrilli / Schizophrenia Research xxx (2015) xxxxxx

but also in individuals with high genetic risk for schizophrenia

(e.g., Kuperberg et al., 2007; Li et al., 2007, Li, Alapati et al., 2012;
Thermenos et al., 2013). In addition, we investigated the link between
behaviour and brain activity to discover the possible etiological aspects
of schizophrenia. This allowed us to disentangle among different interpretations, and to make stronger assumptions on Crow's specic theory
(1997, 2000): patients with lower gamma amplitude in the left frontal,
but not in posterior sites during the phonological processing showed a
greater extent of Delusions and Hallucinatory behaviour. Thus, the loss
of frontal asymmetry might originate from three interacting mechanisms: (a) the overall frontal hypofunctioning, (b) the lack of the left
hemisphere specialisation for language and (c) the lack of inhibition of
homologous (right-hemisphere) regions. On the one hand, such a result
is in line with prior research carried out with a different method (Slow
Evoked Potentials; 15) on another sample of patients. On the other
hand, this nding highlights an important functional link between the
left frontal reduced integration mechanisms and positive key symptoms
of schizophrenia. These ndings are apparently in contrast with past literature on working memory studies carried out using different version
of N-back tasks, in which greater levels of gamma oscillations were
found in schizophrenia patients than controls (Basar-Eroglu et al., 2007;
Barr et al., 2010). In fact, N-back task mainly activates working memory,
i.e., a relatively limited process able to hold and manipulate on-going
information (Baddeley and Hitch, 1974), and gamma increase was correlated with negative symptoms. In our study, a relatively distributed,
language-related network was necessarily activated, and gamma
amplitude was correlated with greater extent of positive symptoms.
Furthermore, gamma amplitude recorded from posterior regions
was comparable (or even higher in the left hemisphere) to those
measured from healthy controls, suggesting that patients' hypofrontality
was a consequence of their primary decit in functional linguistic integration, hierarchically organized by Broca's area (Koechlin and Jubault,
2006). According to Ferrarelli et al. (2008), which found a decreased
gamma-evoked response in schizophrenia patients after TMS stimulation on the frontal (but not motor) cortex, patients' impaired frontal
thalamo-cortical circuits should lead to a reduced or absent ability to
generate efcient and synchronized gamma oscillations.
Within posterior sites, controls' signicant greater left than right activation during the phonological processing suggests that this task
needs the activation of an anteriorposterior extended network within
the left hemisphere by specically involving, with respect to the other
two tasks, the verbal working memory. However, in posterior locations,
patients had signicantly greater gamma amplitude than controls on
the left hemisphere, but their asymmetrical (left N right) pattern of activation was not modulated by tasks. This result is in line with current
hypotheses postulating a lack of connectivity and monitoring of dominant frontal regions over posterior ones, that become relatively more
activated, or better, disinhibited (Ford et al., 2002; Ford and Mathalon,
2004), and in general to a dysfunction of the language-related left
temporal regions (Bruder et al., 1999; Li, Alapati et al., 2012). Alternatively, this pattern should reect a compensatory mechanism aimed at
carrying out the linguistic tasks administered in the present experiment.
In this case, however, the mechanism should be considered functional
only to task execution, with no generalized advantages for positive or
negative symptoms in everyday life. Although the main sources of auditory hallucinations are in the temporal posterior regions, the extent of
this disinhibited disconnected hallucinatory activity is correlated with
the weakening of dominant control from the left frontal sites: indeed,
we found a correlation between the reduced left frontal gamma asymmetry and the extent of hallucinatory behaviour P3 scale. In addition,
the correlation with two peculiar negative symptoms the Lack of
spontaneity and ow of conversation (N6) and the Stereotyped thinking (N7) together with the other positive construct Delusions
(P1) suggested a clear link between the key role of anterior (rather
than posterior) regions of the left hemisphere for a coherent thinking
and speech skills. The latter represent more specic associations

between our linguistic tasks (and the phonological processing in particular) and the altered (although not so evident) linguistic functioning in
everyday activities of schizophrenia patients, thus conrming the ecological impact of our paradigm for the study of this important disease.
All correlations were relatively high, notwithstanding that the antipsychotic medication tends to reduce positive symptoms, suggesting a
stable network-connectivity pattern probably related to the chronic
evolution of patients' psychotic disorder.
The innovative aspect of the present research lies in the ability of the
EEG gamma amplitude to evidence the specic role of the phonological
process to highlight the left frontal linguistic control-dominance over
both contralateral right homologous locations, and the ipsilateral left
posterior regions. This important outcome allowed us to conrm Crow's
hypothesis on language lateralization as a key probe able to assess, in
the schizophrenic brain, a connectivity, language-associated, core decit
correlated with the main symptoms of this elusive psychiatric disorder.
Role of funding source
The funder had no role in the study design, in the collection, analysis and interpretation of data, in the writing of the manuscript, or in the decision to submit the article for
publication.
Contributors
CS was the Principal Investigator, collected and analysed all data, made statistics, and
wrote the rst and following drafts of the manuscript. AA designed the experiment, and
contributed to the interpretation of results, several drafts, and critical discussion of drafts
of the paper. All authors contributed to and have approved the nal manuscript.
Conict of interest
No author reports any biomedical nancial interest or potential conict of interest.
Acknowledgement
This study was supported by a grant PRAT 2012 from the University of Padova, project
no. CPDA125502 to A.A.

References
Andreasen, N.C., Rezai, K., Allinger, R., Swayze, V.W., Flaum, M., Kirchner, P., Cohen, G.,
O'Learly, D.S., 1992. Hypofrontality in neuroleptic-nave patients and in patients
with chronic schizophrenia: assessment with Xenon 133 single-photon emission
computed tomography and the Tower of London. Arch. Gen. Psychiatry 49, 943958.
Andreasen, N.C., Nopoulos, P., O'Leary, D.S., Miller, D.D., Wassink, T., Flaum, M., 1999.
Dening the phenotype of schizophrenia: cognitive dysmetria and its neural mechanisms. Biol. Psychiatry 46, 908920.
Angrilli, A., Spironelli, C., Elbert, T., Marano, G., Stegagno, L., 2009. Schizophrenia as failure
of left hemispheric dominance for the phonological component of language. PLoS
ONE 4 (2), e4507. http://dx.doi.org/10.1371/journal.pone.0004507.
Baddeley, A.D., Hitch, G.J., 1974. Working Memory. The Psychology of Learning and
Motivation. Academic Press, New York.
Barr, M.S., Farzan, F., Tran, L.C., Chen, R., Fitzgerald, P.B., Daskalakis, Z.J., 2010. Evidence for
excessive frontal evoked gamma oscillatory activity in schizophrenia during working
memory. Schizophr. Res. 121, 146152.
Basar-Eroglu, C., Brand, A., Hildebrandt, H., Kedzior, K.K., Mathes, B., Schmiedt, C., 2007.
Working memory related gamma oscillations in schizophrenia patients. Int.
J. Psychophysiol. 64, 3945.
Bookheimer, S., 2002. Functional MRI of language: new approaches to understanding the
cortical organization of semantic processing. Annu. Rev. Neurosci. 25, 151188.
Bruder, G., Kayser, J., Tenke, C., Amador, X., Friedman, M., Sharif, Z., Gorman, J., 1999. Left
temporal lobe dysfunction in schizophrenia. Event-related potential and behavioral
evidence from phonetic and tonal dichotic listening tasks. Arch. Gen. Psychiatry 56,
267276.
Crow, T.J., 1997. Schizophrenia as failure of hemispheric dominance for language. Trends
Neurosci. 20, 339343.
Crow, T.J., 2000. Schizophrenia as the price that Homo sapiens pays for language: a
resolution of the central paradox in the origin of the species. Brain Res. Rev. 31,
118129.
Fehr, T., Kissler, J., Moratti, S., Wienbruch, C., Rockstroh, B., Elbert, T., 2001. Source
distribution of neuromagnetic slow waves and MEG-delta activity in schizophrenia
patients. Biol. Psychiatry 50, 108116.
Ferrarelli, F., Massimini, M., Peterson, M.J., Riedner, B.A., Lazar, M., Murphy, M.J., Huber, R.,
Rosanova, M., Alexander, A.L., Kalin, N., Tononi, G., 2008. Reduced evoked gamma oscillations in the frontal cortex in schizophrenia patients: a TMS/EEG study. Am.
J. Psychiatry 165, 9961005.
Ford, J.M., Mathalon, D.H., 2004. Electrophysiological evidence of corollary discharge
dysfunction in schizophrenia during talking and thinking. J. Psychiatr. Res. 38, 3746.
Ford, J.M., Mathalon, D.H., Whiteld, S., Faustman, W.O., Roth, W.T., 2002. Reduced
communication between frontal and temporal lobes during talking in schizophrenia.
Biol. Psychiatry 51, 485492.

Please cite this article as: Spironelli, C., Angrilli, A., Language-related gamma EEG frontal reduction is associated with positive symptoms in
schizophrenia patients, Schizophr. Res. (2015), http://dx.doi.org/10.1016/j.schres.2015.04.003

C. Spironelli, A. Angrilli / Schizophrenia Research xxx (2015) xxxxxx

Goldberg, T.E., Gold, J.M., 1995. Neurocognitive decits in schizophrenia. In: Hirsch, S.R.,
Weinberger, D.R. (Eds.), Schizophrenia. Blackwell Science, Oxford, pp. 146162.
Hagoort, P., 2005. On Broca, brain, and binding: a new framework. Trends Cogn. Sci. 9,
416423.
Herrmann, C.S., Demiralp, T., 2005. Human EEG gamma oscillations in neuropsychiatric
disorders. Clin. Neurophysiol. 116, 27192733.
Herrmann, C.S., Munk, M.H.J., Engel, A.K., 2004. Cognitive functions of gamma activity:
memory match and utilization. Trends Cogn. Sci. 8, 347355.
Ille, N., Berg, P., Scherg, M., 2002. Artifact correction of the ongoing EEG using spatial
lters based on artifact and brain signal topographies. J. Clin. Neurophysiol. 19,
113124.
Kay, S.R., Fiszbein, A., Opler, L.A., 1987. The positive and negative syndrome scale
(PANSS): development and standardization. Schizophr. Bull. 13, 261276.
Koechlin, E., Jubault, T., 2006. Broca's area and the hierarchical organization of human behavior. Neuron 50, 963974.
Kuperberg, G.R., 2010a. Language in schizophrenia part 1: an introduction. Lang. Linguist.
Compass 4/8, 576589. http://dx.doi.org/10.1111/j.1749-818x.2010.00216.x.
Kuperberg, G.R., 2010b. Language in schizophrenia part 2: what can psycholinguistics
bring to the study of schizophrenand vice versa? Lang. Linguist. Compass 4/8,
590604. http://dx.doi.org/10.1111/j.1749-818x.2010.00217.x.
Kuperberg, G.R., Deckersbach, T., Holt, D., Goff, D., West, W.C., 2007. Increased temporal
and prefrontal activity to semantic associations in schizophrenia. Arch. Gen.
Psychiatry 64, 138151.
Kwon, J.S., O'Donnel, B.F., Wallenstein, G.V., Greene, R.W., Hirayasu, Y., Nestor, P.G.,
Hasselmo, M.E., Potts, G.F., Shenton, M.E., McCarley, R.W., 1999. Gamma frequencyrange abnormalities to auditory stimulation in schizophrenia. Arch. Gen. Psychiatry
56, 10011005.
Li, X., Branch, C.A., Ardekani, B.A., Bertisch, H., Hicks, C., DeLisi, L.E., 2007. fMRI study of
language activation in schizophrenia, schizoaffective disorder and in individuals
genetically at high risk. Schizophr. Res. 96, 1424.
Li, X., Alapati, V., Jackson, C., Xia, S., Bertisch, H.C., Branch, C.A., DeLisi, L.E., 2012. Structural abnormalities in language circuits in genetic high-risk subjects and schizophrenia patients.
Psychiatry Res. Neuroimaging 201, 182189.
McCarley, R.W., Salisbury, D.F., Hirayasu, Y., Yurgelun-Todd, D.A., Tohen, M., Zarate, C.,
Kikinis, R., Jolesz, F.A., Shenton, M.E., 2002. Association between smaller left posterior
superior temporal gyrus volume on magnetic resonance imaging and smaller left
temporal P300 amplitude in rst-episode schizophrenia. Arch. Gen. Psychiatry 59,
321331.

Mientus, S., Gallinat, J., Wuebben, Y., Pascual-Marqui, R.D., Mulert, C., Frick, K., Dorn, H.,
Herrmann, W.M., Winterer, G., 2002. Cortical hypoactivation during resting EEG in
schizophrenics but not in depressives and schizotypal subjects as revealed by low resolution electromagnetic tomography (LORETA). Psychiatry Res. Neuroimaging 116, 95111.
Oldeld, R.C., 1971. The assessment and analysis of handedness: the Edinburgh Inventory.
Neuropsychologia 9, 97113.
Oostenveld, R., Praamstra, P., 2001. The ve percent electrode system for high-resolution
EEG and ERP measurements. Clin. Neurophysiol. 112, 713719.
Rockstroh, B., Kissler, J., Mohr, B., Eulitz, C., Lommen, U., Wienbruch, C., Cohen, R., Elbert,
T., 2001. Altered hemispheric asymmetry of auditory magnetic elds to tones and
syllables in schizophrenia. Biol. Psychiatry 49, 694703.
Spironelli, C., Angrilli, A., 2010. Developmental aspects of language lateralization in delta,
theta, alpha and beta EEG bands. Biol. Psychol. 85, 258267.
Spironelli, C., Angrilli, A., Stegagno, L., 2008. Failure of language lateralization in
schizophrenia patients: an ERP study on early linguistic components. J. Psychiatry
Neurosci. 33, 235243.
Spironelli, C., Angrilli, A., Calogero, A., Stegagno, L., 2011. Delta EEG band as a marker of
left hypofrontality in schizophrenia patients. Schizophr. Bull. 37, 757767.
Spironelli, C., Manfredi, M., Angrilli, A., 2013. Beta EEG band: a measure of functional brain
damage and language reorganization in aphasic patients after recovery. Cortex 49,
26502660.
Tan, H.-Y., Callicott, J.H., Wienberger, D.R., 2007. Dysfunctional and compensatory
prefrontal cortical systems, genes and the pathogenesis of schizophrenia. Cereb. Cortex 17, i171i181.
Tan, H.-R.M., Lana, L., Uhlhaas, P.J., 2013. High-frequency neural oscillations and visual
processing decits in schizophrenia. Front. Psychol. 4, 621. http://dx.doi.org/10.
3389/fpsyg.2013.00621.
Tauscher, J., Fischer, P., Neumeister, A., Rappelsberger, P., Kasper, S., 1998. Low frontal
electroencephalographic coherence in neuroleptic-free schizophrenic patients. Biol.
Psychiatry 44, 438447.
Thermenos, H.W., Whiteld-Gabrieli, S., Seidman, L.J., Kuperberg, G., Juelich, R.J., Divatia,
S., Riley, C., Jabbar, G.A., Shenton, M.E., Kubicki, M., Manschreck, T., Keshavan, M.S.,
DeLisi, L.E., 2013. Altered language network activity in young people at familial
high-risk for schizophrenia. Schizophr. Res. 151, 229237.
van Boxtel, A., Goudswaard, P., Janssen, K., 1984. Amplitude and bandwidth of the frontalis
surface EMG: effects of electrode parameters. Psychophysiology 21, 699707.
Wienberger, D.R., Berman, K.F., 1996. Prefrontal function in schizophrenia: confounds and
controversias. Philos. Trans. R. Soc. Lond. B Biol. Sci. 351, 14951503.

Please cite this article as: Spironelli, C., Angrilli, A., Language-related gamma EEG frontal reduction is associated with positive symptoms in
schizophrenia patients, Schizophr. Res. (2015), http://dx.doi.org/10.1016/j.schres.2015.04.003