Secondary Schizophrenia (26 page)

subdivisions
[80, 79, 81],
including the mediodorsal
nucleus
[82, 83].

In some of the studies, functional disconnections
fMRI

have been inferred for the fronto-temporal
[55,
84, 85,

The studies that examined the resting state functional
86],
prefronto-cingulate
[87],
prefronto-parietal
[88],

connectivity in patients with schizophrenia in the
contralateral prefronto-hippocampal
[89],
occipito-entire brain reported widespread functional discontemporal
[77],
as well as thalamocortical (specifically
nectivity (mainly reductions in connectivity), includ-involving the mediodorsal nucleus in the left hemi-ing within the so-called “default mode” network of
sphere,
[90]),
temporo-occipito-cerebellar and fronto-brain activity
[109, 110, 111]
. Other more localized
striatal regions
[77]
. In the more recent of these stud-resting state investigations found reduced prefrontal
ies, multiple foci of aberrant activation have been doc-lobe connectivity with the temporal
[112]
as well as
umented.

parietal lobes, thalamus and striatum
[113].
In this lat-A possible relationship of functional disconnec-

ter study, enhanced fronto-temporal resting connectivity to acute state of the illness has been suggested
tivity in the left hemisphere was also reported
[113].
At
by Erkwoh and colleagues
[91]
who, in a SPECT

least some of the functional disconnectivities may be
study, found that fronto-temporal disconnectivity in
present even before the overt onset of illness in individ-first-episode neuroleptic-na¨ıve patients improved with
uals at high genetic risk
[114, 115, 116]
and in never-symptom remission following antipsychotic treat-

medicated, first-outbreak patients
[99, 102,
113],
par-ment.

ticularly in the prefrontal cortex
[117].

Although the vast majority of the functional MRI
(fMRI) studies report diminished connectivity in
schizophrenia brains either diffusely at rest or as
fMRI

inferred from abnormal regional activations under an
Echoing the earlier metabolic PET studies, fMRI eval-assembly of cognitive tasks, one recent study attempted
uations of task-induced changes in the BOLD signal
to investigate a general pattern of connectivity between
in patients with schizophrenia have over the last two
the core network engaged by a task (lexical decision
decades documented abnormal activation patterns

making and retrieval in this case) and the rest of the
involving the prefrontal cortex
[92, 93, 94, 95],
parietal
brain, viewed as unstructured background noise
[118].

cortex
[96],
temporal cortex
[97, 94, 98],
hippocam-The authors report that although the core topography
pal circuitry
[99, 100],
primary visual areas
[101],

per se did not differ among schizophrenia and healthy
anterior cingulate
[102, 95],
thalamus
[93],
amyg-subjects, connectivity between the core and the rest,
dala
[103, 104],
and cerebellum. Interregional dis-as well as within the rest of the brain, was higher in
connectivities under various tasks have been reported
patients with schizophrenia. Moreover, within the core,
in the prefronto-cingulate and fronto-cerebellar cir-fronto-parietal correlations in schizophrenia patients
cuitries, anterior cingulate and supplementary motor
were significantly lower than in healthy subjects. The
cortices.

authors conclude that there appears to be diminished
Importantly, the number of negative studies pur-

separation of the core task-related functional activ-porting the lack of between-group differences among
ity from the background noise, as well as dimin-

patients with schizophrenia and healthy controls has
ished antero-posterior interlobar integration within
thus far been minimal, although these notably per-the core. This finding may provide one possible expla-tained to the oft-reported anterior cingulate region
nation for the complicated patterns of both increased
[105]
and the fusiform gyrus
[106].
Whereas poor
and decreased functional activations found with a vari-task performance in many of these studies has been
ety of tasks as reviewed in the previous chapters. In
associated with hypoactivations in the relevant cir-a similar vein, Winder and colleagues
[119]
report
cuitries as compared to healthy control subjects, poor
diminished fronto-temporal connectivity in the left
patient performance in some other tasks has actually
hemisphere and strengthened fronto-temporal con-

been associated with hyperactivations, deemed reflec-nectivity in the right hemisphere during a matching
tive of an increased effort on the patients’ part
[107,

task. Gur and colleagues
[120]
likewise found that
64

108].

patients not only underactivated target-specific areas
Chapter 5 – Functional neuroimaging in schizophrenia

to attentional demands, but also displayed an exces-Studies evaluating changes in correlations of
sive employment of networks apparently committed to
regional EEG activity with cognitive (working mem-processing irrelevant stimuli.

ory) tasks have found reduced network recruitments
in patients
[137, 138].
Similarly, normal gamma-band
synchronization preceding a willed motor act, talk-

EEG

ing or thinking (thought to reflect corollary dis-Gamma-band electrical activity (high-frequency oscil-charge, important in theorizations on hallucinatory
lations at
∼
40 Hz) has in recent years been at the
and delusional passivity phenomena, cf.
[139])
has
very center of theoretical interest in the pathophys-recently been reported to be diminished in patients
iology of schizophrenia because synchronization of
with schizophrenia
[140].
These studies generally use
activities in the gamma band is viewed as a likely
techniques akin to evoked response potentials.

candidate to account for the mechanism of integraThe fact that some authors found increased syn-

tive binding of disparate cerebral activities into uni-chronizations (i.e. hyperconnectivity), even in the
fied networks
[121].
This unifying mechanism has
lower range of the beta rhythm (<30 Hz), in patients
been considered necessary for the conscious sense of
with schizophrenia
[127, 128,
141]
or in their sub-agency dating back to the Kantian theory of appercep-groups led Gruzelier
[142]
to propose that the directive self-awareness. Disruption in the mechanism of
tion of change may in fact be syndrome dependent,
binding would therefore be resonant with the clinical
namely, diminished with the predominance of the neg-descriptions of schizophrenia in terms of loose asso-ative symptoms and increased with the predominance
ciations
[122]
or cognitive dysmetria
[36],
the patho-of the positive symptoms. Lee and Cooeagyues [143]

physiological concept of disconnection, and indeed
observed that although synchronization in gamma-

the etymology of the nosological term itself (schizo and
band activity was decreased in the left frontal lobe
phrenia).

in schizophrenia patients as a group, dividing the
In spite of this theoretical interest in the gamma
patients into three subgroups by the predominant
band, reports of disturbances in the somewhat lower-clinical syndrome revealed decreased synchrony with
frequency beta band in the recent literature on

psychomotor poverty, right hemispheric increase in
schizophrenia have been no less frequent, which may
synchrony with reality distortion, and widespread
in part be explained by terminological confusion.

increase, involving more posterior regions, with disor-Thus, a study by Strelets
et al.
[123],
which reported
ganization. In a case report, intense tactile hallucina-disruption of interhemispheric “high-frequency beta-tions in a nonschizophrenic patient also were specifi-rhythm” integration in the frontal lobes in patients
cally related to increased synchronicity in the gamma
with schizophrenia, in fact focused on the approxi-band
[144].
Yet, in a study by Strelets and colleagues
mately 40 Hz oscillations that by many authors would
[123]
, both negative and positive syndrome groups of
be considered to lie in the gamma frequency band.

schizophrenics showed decreased synchronicity in the
Taken together, diminished synchronization of local
high-frequency beta/gamma rhythm, so that the issue
oscillations in the high-frequency beta and gamma
remains unsettled at this time.

bands has been reported by several groups
[124].

Some of the previous reports apparently evalu-

Other authors found reduced
[125]
or increased
ated first-episode schizophrenia patients
[136, 133,

[126, 127, 128]
coherence in the theta and delta
125, 126, 127,
145,
124, 128,
146]
indicating that the
frequency bands, generally assessed under resting
abnormalities may already be present close to the
condition. Widespread, particularly fronto-temporal,
clinical onset of the illness
[147]
or even represent
disconnectivity was documented using a related tech-a genetic diathesis for psychosis
[148].
Strelets and
nique of low-resolution electromagnetic tomography
colleagues
[149]
found some lateralized differences
(LORETA) as well
[129,
130, 131, 132].
In addition,
between the first-episode and chronic schizophrenia
microstates of synchronous EEG activity at rest have
patients, although both groups displayed disordered
been noted to be shorter
[133, 134],
sequentially dif-interhemispheric connectivity.

ferent
[135]
and topographically more dispersed
[136]

A recent study, utilizing innovative statistical
in patients with schizophrenia, thus indicating both
methodology, described differential patterns of frac-reduced maintenance of the synchronizations and
tal, higher-order temporal periodicity in the alpha
65

impairment in their sequential integration.

and theta rhythm fluctuations in the frontal lobes
The Neurology of Schizophrenia – Section 2

among patients with schizophrenia and healthy com-in some reports, to the right temporal lobe regions
parison subjects, suggesting increased randomness
as well
[42,
156, 157, 158, 159, 160, 161].
Reduced
(diminished regularity) in the patients in the alpha
fronto-temporal connectivity was also associated with
rhythm dynamics and, in contrast, diminished com-auditory hallucinations
[162]
and there was an obser-plexity (heightened regularity) in the theta rhythm
vation of an accentuated multiregional response to
dynamics
[150].
Another innovative study
[151],

emotion-laden words in hallucinating patients
[163].

which analyzed multivariate S-estimator maps of

Similar to the described activations in auditory corti-synchronized electrical activity across a wide range of
cal areas with auditory hallucinations, a case report of
bandwidths, found a bilateral pattern of increased syn-fMRI in a patient with visual hallucinations revealed
chronization in the temporal regions and decreased
activation in the visual areas and the hippocampus
synchronization in the upper parietal regions,

[164].

both changes more pronounced with more severe

Formal thought disorder, specifically disturbance
psychopathology.

in the train of thought, has been related to underactivation of the posterior (Wernicke’s) part of the supe-

Structural (volumetric)

rior temporal gyrus
[165, 166, 167]
and Broca’s area
in the frontal lobe
[168].
Receptive prosodic distur-

intercorrelations

bances
[169]
and misattribution of speech
[170, 171]

Two reports of detailed analyses of regional cortical
in patients were also associated with primary auditory
intercorrelations of grey matter volumes by individual
cortex and its connectivity with the anterior cingulate.