Acta Psychiatr Scand 2016: 134: 16–30 © 2016 John Wiley & Sons A/S.

16 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd
All rights reserved ACTA PSYCHIATRICA SCANDINAVICA
DOI: 10.1111/acps.12564

Review
Neuropsychological and neuroimaging
underpinnings of schizoaffective disorder: a
systematic review
Madre M, Canales-Rodr
ıguez EJ, Ortiz-Gil J, Murru A, Torrent C, M. Madre1, E. J. Canales-
Bramon E, Perez V, Orth M, Brambilla P, Vieta E, Amann BL. Rodr
ıguez1, J. Ortiz-Gil1,2,
Neuropsychological and neuroimaging underpinnings of schizoaffective A. Murru3, C. Torrent3,
disorder: a systematic review. E. Bramon4, V. Perez5,6, M. Orth7,
P. Brambilla8,9, E. Vieta3,
Objectives: The neurobiological basis and nosological status of
schizoaffective disorder remains elusive and controversial. This study
B. L. Amann1
provides a systematic review of neurocognitive and neuroimaging
1
FIDMAG Research Foundation Germanes Hospitalaries,
CIBERSAM, Barcelona, Spain, 2Hospital General de
findings in the disorder.
Granollers, Granollers, Catalonia, Spain, 3Bipolar
Methods: A comprehensive literature search was conducted via Disorders Unit, Institute of Neuroscience, Hospital
PubMed, ScienceDirect, Scopus and Web of Knowledge (from 1949 to Clinic, IDIBAPS, CIBERSAM, University of Barcelona,
31st March 2015) using the keyword ‘schizoaffective disorder’ and any Barcelona, Catalonia, Spain, 4Division of Psychiatry,
of the following terms: ‘neuropsychology’, ‘cognition’, ‘structural University College London, London, UK, 5Institut de
neuroimaging’, ‘functional neuroimaging’, ‘multimodal’, ‘DTI’ and Neuropsiquiatria i Addiccions, Hospital del Mar,
‘VBM’. Only studies that explicitly examined a well defined sample, or Barcelona, Spain, 6CIBERSAM, IMIM (Institut Hospital
subsample, of patients with schizoaffective disorder were included. del Mar d’Investigacions Mediques), Psiquiatria,
Results: Twenty-two of 43 neuropsychological and 19 of 51 Universitat Autonoma de Barcelona, Barcelona, Spain,
7
neuroimaging articles fulfilled inclusion criteria. We found a general Department of Neurology, Ulm University, Ulm,
Germany, 8Department of Neurosciences and Mental
trend towards schizophrenia and schizoaffective disorder being related
Health, Psychiatric Clinic, Fondazione IRCCS Ca’ Granda
to worse cognitive performance than bipolar disorder. Grey matter Ospedale Maggiore Policlinico, University of Milan,
volume loss in schizoaffective disorder is also more comparable to Milan, Italy and 9Department of Psychiatry and
schizophrenia than to bipolar disorder which seems consistent across Behavioural Sciences, University of Texas Health
further neuroimaging techniques. Science Center at Houston, Houston, TX, USA
Conclusions: Neurocognitive and neuroimaging abnormalities in Key words: schizoaffective disorder; schizophrenia;
schizoaffective disorder resemble more schizophrenia than bipolar bipolar disorder; neuroimaging; neurocognition
disorder. This is suggestive for schizoaffective disorder being a subtype Benedikt L. Amann, FIDMAG Research Foundation,
of schizophrenia or being part of the continuum spectrum model of CIBERSAM, Dr. Antoni Pujadas 38, 08830 Sant Boi de
psychosis, with schizoaffective disorder being more skewed towards Llobregat, Barcelona, Spain. E-mail: benedikt.
schizophrenia than bipolar disorder. [email protected]

Accepted for publication February 8, 2016

Summations
• The nosological status of schizoaffective disorder remains still controversial.
• We performed a systematic review of neurocognition and neuroimaging of this disorder showing that
abnormalities in patients with schizoaffective disorder and schizophrenia are similar with bipolar
patients being less affected.
• Schizoaffective disorder is nosologically more skewed towards schizophrenia.

16
 Neurobiology of schizoaffective disorder

Considerations
• Small number of neuroimaging studies were included per modality because of which data could not
be meta-analysed.
• Further comparative neuroimaging trials across mood and psychotic disorders with other neuroimag-
ing techniques are needed.

Introduction disorder. Only two reviews on schizoaffective dis-

Over 120 years ago, Emil Kraepelin differentiated order including also neuroimaging data have been
dementia precox, later to be called schizophrenia, conducted so far (16, 17), but conclusions were not
from manic-depressive insanity, later to be termed firm because of the limited literature at that stage.
bipolar disorder (1). Kraepelin questioned himself Structural imaging data in schizophrenia suggest a
in 1920: ‘It is becoming increasingly obvious that reduction in whole-brain volume of about 2%
we cannot satisfactorily distinguish these two dis- compared with controls, with greater volume loss
eases’ (2). Thirteen years later, Kasanin introduced in regions such as the frontal lobe and the hip-
the term schizoaffective psychosis, when he pocampus (18). In contrast, in bipolar disorder
described nine patients with good premorbid func- only minor whole-brain and regional volume
tioning, who then developed a mixture of psychotic changes have been reported (19, 20). Evidence
and affective symptoms with a full recovery after a from functional imaging studies suggests that a
few months (3). More than eighty years after the distinction between bipolar disorder and
term was introduced, the nosological status of schizophrenia is traceable. Of note, functional net-
schizoaffective disorder still remains elusive and works studies reveal an altered connectivity gener-
controversial. At the centre of the debate is the alized to all brain networks in schizophrenia and
question of the relationship and boundaries restricted to limbic, paralimbic, and interhemi-
between affective and schizophrenic disorders. The spheric networks in bipolar disorder (21). Cogni-
illness is often viewed as a heterogeneous spectrum tive impairment is common to both disorders after
disorder. Some patients seem more prone towards onset but the magnitude of the impairment seems
schizophrenia, and others more towards affective greater in schizophrenia than in bipolar disorder
disorders (4–7). The categorical model proposes, (22). Hence, there is good evidence to indicate that
however, that schizophrenia and affective disorders the neurobiological basis of schizophrenia and
are distinct and mutually exclusive illnesses so that bipolar disorder differs. In schizoaffective disorder,
schizoaffective disorder is viewed as either a form the available neurocognitive, functional and struc-
of schizophrenia (8, 9), a form of affective disorder tural imaging data have increased recently.
(10), or an illness distinct from both schizophrenic
and affective disorders (11). In contrast, a spec- Aims of the study
trum model posits that psychosis severity varies on
a continuous scale with schizophrenia and affective It was the aim of this review to first summarize
disorders at opposite ends, and schizoaffective dis- these data, and, second, to examine if there is evi-
order in-between. Finally, it has been provoca- dence to indicate neurobiological differences
tively questioned if schizoaffective disorder exists between schizoaffective disorder and schizophrenia
at all (12, 13) proposing its elimination from the or bipolar disorder, including first longitudinal
current diagnostic systems (13). This was initially data.
considered for the new DSM-5 (14), with mood
symptoms instead being added as a dimension to
Material and methods
schizophrenia and schizophreniform disorder.
However, the category for schizoaffective disorder A comprehensive literature search of neurocogni-
was ultimately maintained, with a perceived lack tive, functional and structural neuroimaging
of neurobiological validating data being cited as studies in schizoaffective disorder was conducted
the reason (15). using the PubMed, ScienceDirect, Scopus and
Structural and functional neuroimaging tech- Web of Knowledge electronic databases (from
niques can probe the neurobiological underpin- 1949 to 31st March 2015) and the search terms
nings of the clinically defined entities ‘schizoaffective disorder’ and any of the following
schizophrenia, bipolar disorder and schizoaffective terms: ‘neuropsychology’, ‘cognition’, ‘structural

17
Madre et al.

neuroimaging’, ‘functional neuroimaging’, ‘multi- than those with schizophrenia (31, 32, 41). Find-
modal’, ‘DTI’ and ‘VBM’. In addition, manual ings from the first study (31) suggest that patients
searches were conducted within review papers and with schizoaffective disorder performed signifi-
reference sections of individual papers. All studies cantly better on a motor function and a visuospa-
reporting a well-defined schizoaffective sample or a tial memory task than patients with schizophrenia,
subanalysis of patients with schizoaffective disor- but not on tasks measuring reaction time and exec-
der within another diagnostic group were eligible utive skills. Fiszdon et al. (32) found that patients
for inclusion. Studies with a mixed diagnostic sam- with schizoaffective disorder performed signifi-
ple without sub-analysis or case series were cantly better than patients with schizophrenia on a
excluded. Neuropsychological studies evaluating theory-of-mind task, but both groups did not differ
cognitive rehabilitation or response to treatment on other measures assessed, including tests of exec-
with neuropsychological batteries were excluded. utive function, verbal and non-verbal memory,
Depending on each study, schizoaffective disorder processing speed and social cognition. The third
was diagnosed following DSM-III, DSM-IV, ICD- study found that patients with schizoaffective dis-
10 or Research Diagnostic Criteria. As seen in Fig- order performed better than subjects with
ure S1, the search retrieved a total of 94 potentially schizophrenia in tests of verbal ability, processing
suitable studies. Of those, 42 manuscripts were speed, visual working memory and verbal memory
excluded based on our exclusion criteria. Finally, (41). Notably, irrespective of the diagnosis,
22 of 43 neuropsychological and 19 of 51 neu- patients with the most severe negative symptoms
roimaging articles fulfilled inclusion criteria for had the most severe cognitive impairment, whereas
this review. We also incorporated information mood symptoms were not related to cognitive per-
from two reviews and one meta-analysis in neu- formance. Another study compared cognitive pat-
ropsychological assessments. This review was con- terns of schizoaffective disorder with different
ducted following the PRISMA guidelines (23). subtypes of schizophrenia (30). Patients with
Specifically, we followed the 27 steps indicated in schizoaffective disorder and paranoid schizophre-
the checklist available on the PRISMA statement nia were neuropsychologically indistinguishable
website: http://www.prisma-statement.org/state- and tended to be preserved, but both groups dif-
ment.htm. fered cognitively from patients with undifferenti-
ated and residual schizophrenia subtypes.
Taken together, cognitive impairment of
Neuropsychological evidence in schizoaffective disorder
patients with schizoaffective disorder seems similar
Neuropsychological findings. Twenty-two studies to those with schizophrenia, even though some
were included measuring cognitive performance of measures were in favour of schizoaffective disor-
patients with schizoaffective disorder (n = 802) in der. However, differences were usually minor. This
comparison with schizophrenia (n = 1956), bipolar suggests that, neuropsychologically, schizoaffective
disorder (n = 991) or healthy subjects (n = 733) disorder resembles schizophrenia.
(24–45). Table 1 summarizes the main findings
from these studies. Comparison of schizoaffective disorder with bipolar
disorder. Studies comparing schizoaffective disor-
Comparison of schizoaffective disorder with der with bipolar disorder are fewer in number and
schizophrenia. In some studies, cognitive deficits of have had conflicting findings (33, 38, 39, 44). Fleck
schizoaffective disorder were similar to schizophre- and colleagues (44) compared selective attention in
nia (24, 25, 27, 29, 37). As in schizophrenia, 12 individuals with schizoaffective disorder, 12
schizoaffective disorder has been associated with with psychotic bipolar disorder and 12 healthy
impairment in frontal lobe mediated cognition, controls, and found a greater attentional interfer-
including working memory, alternating attention, ence in subjects with schizoaffective disorder. This
information recall, category generation, abstrac- is in line with a study by Torrent et al. (33) who
tion and motor planning. Heinrichs et al. (35) found that patients with schizoaffective disorder
found that patients with schizophrenia were more showed significantly more impairment than
impaired on all cognitive measures than patients patients with bipolar disorder on tests of short-
with schizoaffective disorder and healthy controls, and long-term verbal memory, and on two out of
but group differences in cognitive performance four executive measures. Another study revealed
were insufficient to separate these syndromes of that patients with schizoaffective disorder did
psychotic illness. In contrast, three studies found worse than patients with bipolar disorder on tests
that patients with schizoaffective disorder per- of attention, psychomotor speed and memory, but
formed better in some neuropsychological tasks there were no significant differences on measures

18
 Neurobiology of schizoaffective disorder

Table 1. Neuropsychological findings in patients with schizoaffective disorder

Main findings: neuropsychological

References Population Measures performance

SA vs. SZ
Miller et al. (24) SZ = 26, SA-M = 9, SA-D = 17 LNNB, BVRT, AVLT, WAIS-R, TMT SA = SZ and SA-M = SA-D
Manschreck et al. (25) SZ = 19, SA = 19, UD = 19, HC = 19 Miller–Selfridge task SA = SZ<UD = HC
Gooding et al. (27) SZ = 34, SA = 23, HC = 30 SWM, WCST SZ = SA<HC
Goldstein et al. (30) SZ-UD = 29, SZ-P = 20, SZ-R = 14, SAD = 20 WAIS-R, HCT, HTPT, TMT-B, WCST SA = SZ-P<SZ-U = SZ-R
Stip et al. (31) SA = 13, SZ = 44 MST, Reaction Time Test, PAL, SOC SA≥SZ
Roofeh et al. (29) SZ = 37, SA = 20, HC = 60 CVLT SA = SZ<HC
Fiszdon et al. (32) SA = 73, SZ = 199 WMS-R, WAIS-III, HVLT-R, WCST, SA = SZ but SA>SZ
Hinting Task, BLERT on the Hinting Task
Heinrichs et al. (35) SZ = 103, SA = 48, HC = 72 WAIS, CVLT SZ<SA<HC (differences insufficient
to separate them)
Hooper et al. (37) SZ = 79, SA = 40 FTT, CPT, HVLT, WRAML, VSWM, SA = SZ, SZ<SA in spelling
WJ-III, COWAT, RFFT, WCST,
Eyes Test, WAIS, WRAT
Torniainen et al. (41) SZ = 218, SA = 62, HC = 123 WAIS-R, TMT-A&B, WMS-R, CVLT SZ<SA<HC
SA vs. BD
Fleck et al. (44) SA = 12, P-BD = 12, HC = 12 SCWT SA<P-BD = HC
Torrent et al. (33) SA = 34, NP-BD = 41, HC = 35 WAIS, WCST, CVLT, SCWT, TMT SA<NP-BD = HC
Studentkowski et al. (38) SA-R = 28, BD-R = 32 MWT, TMT, WAIS, VLMT SA-R<BD-R
Simonsen et al. (39) SA = 27, P-BD = 75, NP-BD = 61 WMS, CVLT, WAIS, D-KEFS SZ = SA = P-BD≤NP-BD
SA vs. SZ vs. BD
Evans et al. (26) SA = 29, SZ = 154, NP-BD = 27 WAIS, ASTVS, TWV, TMT,TPT, DVT, SA = SZ<NP-BD
BCT, WCST, CVLT, FMT, SMT, FTT
Glahn et al. (28) SA = 15, SZ = 5, P-BD = 11, NP-BD = 15 SDRT, WAIS Backward digit span: BD = SZ = SA.
Hx. of psychosis in spatial delayed
response task: SA = SZ = P-BD = NP-BD
Szoke et al. (34) SA = 26, SZ = 48, P-BD = 52, NP-BD = 40 WCST, TMT WCST: SZ<SA<P-BD<NP-BD. TMT: SA = SZ
Reichenberg et al. (36) SZ = 94, SA = 15, BD = 78, UD = 48 WAIS-R, WMS-R, SCWT, TMT-A&B, SZ<SA = BD = UD
FTT, FRT, LFSRT
Amann et al. (40) SZ = 45, SA-M = 26, BD-M = 51, HC = 65 WMS, BADS SZ = SA-M = BD-M<HC
DeRosse et al. (42) SZ( ) = 371, SZ(+) = 224, SA = 129, P-BD = 269 WAIS-R digit Span, CVLT, COWAT, SA<SZ( ) = SZ(+)<P-BD
Animal Naming, TMT-A&B, WRAT
Hill et al. (45) SZ = 293, P-BD = 227, SA = 165, HC = 295 BACS SZ<SA<P-BD
Longitudinal study SA
Madre et al. (43) SA-M = 11, SA-D = 11, SA-R = 22, HC = 22 WMS, BADS SA-R<HC, SA-D = SA-R and SA-M<SA-R

ASTVS, Aphasia Screening Test Verbal Score; AVLT, Rey Auditory Verbal Learning Test; BADS, Behavioural Assessment of the Dysexecutive Syndrome; BCT, Booklet Category
Test; BLERT, Bell Lysaker Emotion Recognition Task; BACS, Brief Assessment of Cognition in Schizophrenia; BVRT, Benton Visual Retention Test; COWAT, Controlled Oral Word
Association Test; CPT, Continuous Performance Test; CVLT, California Verbal Learning Test; D-KEFS, Delis-Kaplan Executive Function Scale; DVT, Digit Vigilance Test; FRT, Facial
Recognition Test; FMT, Figure Memory Test; FTT, Finger Tapping Test; HCT, Halstead Category Test; HTPT, Halstead Tactual Performance Test; HVLT, Hopkins Verbal Learning Test;
LNNB, Luria-Nebraska Neuropsychological Battery; LFSRT, Letter Fluency and Sentence Repetition Test; MST, Motor Screening Task; MWT, Mehrfachwahl-Wortschatz-Intelli-
genztest; PAL, Paired Associates Learning task; RFFT, Ruff Figural Fluency Test; SCWT, Stroop Colour-Word Interference test; SDRT, Spatial Delayed Response Task; SMT, Story
Memory Test; SOC, Stockings of Cambridge items; SWM, Spatial working memory; TMT, Trail Making Test; TPT, Tactual Performance Test; TWV, Thurstone Written Fluency;
VLMT, Verbal Learning Memory Test; VSWM, Visuospatial working memory; WAIS, Wechsler Adult Intelligence Scale; WAIS-R, Wechsler Adult Intelligence Scale Revised;
WCST, Wisconsin Card Sorting Test; WMS, Wechsler Memory Scale; WMS-R, Wechsler Memory Scale Revised; WJ-III, Woodcock-Johnson Test of Cognitive Abilities, third Edi-
tion; WRAML, Wide Range Assessment of Memory and Learning; WRAT, Wide Range Achievement Test.
SA, schizoaffective disorder; SA-M, schizoaffective disorder, manic episode; SA-D, schizoaffective disorder, depressive episode; SA-R schizoaffective disorder in clinical remis-
sion; SZ, schizophrenia; SZ-P, schizophrenia paranoid; SZ-U, schizophrenia undifferentiated; SZ-R, schizophrenia residual; SZ( ), schizophrenia with no history of mood syndrome,
SZ(+), schizophrenia with a superimposed mood syndrome; BD, bipolar disorder; NP-BD, non-psychotic bipolar disorder; P-BD, psychotic bipolar disorder; BD-M, bipolar disorder,
manic episode; BD-R, bipolar disorder in clinical remission; UD, Unipolar depression; HC, Healthy Controls; Hx, History.

of cognitive flexibility and emotional memory (38). disorder differ from those with bipolar disorder.
Interestingly, another study found no differences Subjects with schizoaffective disorder may be cog-
between patients with schizoaffective and bipolar nitively more affected than subjects with bipolar
disorders with a history of psychotic symptoms on disorder, at least in patients with bipolar disorder
a battery of seven tests covering memory, process- without psychotic symptoms. However further
ing speed and executive function; however, there studies are needed to substantiate any conclusions.
were significant differences compared to patients
with non-psychotic forms of bipolar disorder on Comparison of schizoaffective disorder with
four of these tests (39). schizophrenia and bipolar disorder. To date, seven
Taken together, it seems not entirely clear if neu- studies have compared cognitive functioning
ropsychologically patients with schizoaffective across all three disorders (26, 28, 34, 36, 40, 42,

19
Madre et al.

45). Evans et al. (26) found that overall perfor- with schizophrenia (n = 293), psychotic bipolar
mance on a neuropsychological test battery was disorder (n = 227), schizoaffective disorder (manic,
similar in patients with schizophrenia and patients n = 110; depressed, n = 55), their first-degree rela-
with schizoaffective disorder, but both groups were tives (n = 316, n = 259, n = 133, and n = 64,
more impaired than patients with non-psychotic respectively) and healthy controls (n = 295) were
mood disorder on a composite neuropsychological examined via the Brief Assessment of Cognition in
score. More recently, Amann et al. (40) showed a Schizophrenia (BACS) neuropsychological bat-
broadly similar degree of executive and memory tery. They found that cognitive impairment was
deficits in acutely ill patients with schizophrenia, the smallest in patients with bipolar disorder and
schizoaffective disorder, currently in a schizomanic worst in schizophrenia with schizoaffective disor-
phase, and bipolar disorder, currently in a manic der in-between. This finding is compatible with a
episode, supporting that a categorical differentia- continuum model, in which more prominent affec-
tion across different acute states was not possible tive features and less enduring psychosis are associ-
with regard to neuropsychological deficits. Two ated with less cognitive impairment (45).
further studies have separated both psychotic and Taken together, the results of the different stud-
non-psychotic forms of bipolar disorder and com- ies differ, even though they show a general trend
pared them with groups of patients with towards schizophrenia and schizoaffective disorder
schizophrenia and schizoaffective disorder (28, 34). being related to worse cognitive performance than
In the first study, patients with schizoaffective dis- bipolar disorder.
order did not significantly differ from patients with
either form of bipolar disorder, nor from patients Evidence from reviews or meta-analysis. Two
with schizophrenia in measures of verbal working reviews and one meta-analysis have tried to clarify
memory, spatial delayed response task, forward the evidence of neuropsychological studies in
and backward digit span (28). In contrast, in the schizoaffective disorder so far. A review by Bucha-
second study Szoke et al. (34) found evidence sup- nan et al. (46) concluded that schizophrenia and
porting a progressive impairment from bipolar dis- schizoaffective disorder shared a similar pattern of
order without psychotic symptoms through cognitive deficits, which was distinct from that seen
bipolar disorder with psychotic symptoms, to in major depression and bipolar disorder. How-
schizoaffective disorder and finally to schizophre- ever, that review could only be based on only two
nia on the number of perseverative errors in the of the above mentioned studies (24, 26). Similarly,
Wisconsin Card Sorting Test (WCST) but not on in a later review it was suggested that subjects with
the Trail Making Test (TMT). Reichenberg et al. schizoaffective disorder are similar to those with
(36) additionally included a group of patients with schizophrenia on most neuropsychological mea-
unipolar depression and found quite similar neu- sures, especially when information processing was
ropsychological performance profile patterns the focus of the study (16). More recently, a meta-
among the four diagnostic groups: schizophrenia, analysis of 31 studies compared the cognitive per-
schizoaffective disorder, bipolar disorder and formance of patients with schizophrenia and
unipolar depression. All groups demonstrated patients with affective psychosis and schizoaffec-
impairments in memory, executive functions, tive disorder (47). The patients with schizophrenia
attention and processing speed. Nevertheless, were found to perform significantly worse than
patients with schizophrenia were more impaired those with schizoaffective disorder or affective psy-
than the other groups on all cognitive domains. chosis on intelligence, mental speed, verbal work-
A more recent study with a larger sample ing memory, verbal declarative memory and the
(n = 993) compared schizoaffective disorder with WCST (category formation and set shifting), but
bipolar disorder and two groups of patients with not on other measures of executive function such
schizophrenia, with or without a superimposed as the TMT-B, nor on fluency, visual declarative
mood syndrome (42). In contrast to the above memory, digit span or non-verbal working mem-
mentioned studies, they found that cognitive dete- ory. Nonetheless, it is important to point out that
rioration across the course of illness was signifi- the between-group differences were small and
cantly worse in patients with schizoaffective mainly driven by negative symptoms, a younger
disorder than in patients with psychotic bipolar age at onset of the illness and a higher percentage
disorder. Interestingly, in both groups with of males in the group with schizophrenia. When
schizophrenia the speed of decline in cognitive schizophrenia was compared separately with
functioning was between that of patients with schizoaffective disorder and affective psychosis, the
bipolar and schizoaffective disorder. In the other magnitude of the differences between schizophre-
study using a large sample (45), groups of subjects nia and affective psychosis were larger than those

20
 Neurobiology of schizoaffective disorder

between schizophrenia and schizoaffective disorder schizoaffective disorder, and only larger third ven-
in verbal memory and executive functions, despite tricle in affective psychosis. Another study acquired
not being statistically significant. magnetic resonance imaging (MRI) data from 108
subjects with schizophrenia, 20 with schizoaffective
Longitudinal studies. Until now it is not well estab- disorder, 27 with major depression, 20 with bipolar
lished whether cognitive impairment represents a disorder and 150 healthy volunteers, all of which
stable, trait-like feature in schizoaffective disorder were categorized into one of five groups: normal,
comparable to that of schizophrenia (48) or bipo- hyperintensity signals, volume loss, ventricular
lar disorder (49). Several reviews (e.g. (50)) and a anomaly or ‘other’ abnormality (65). They found
recent longitudinal study in schizophrenia (51) did that the prevalence of morphologic abnormalities
not support the notion of a consistent decline for was statistically higher in all disorders than in
most neuropsychological measures. Likewise, the healthy volunteers, but not significant differences
evidence base for changes in cognitive function in were found among disorders, suggestive for a non-
bipolar disorder across relapse and remission is specific role for morphologic anomaly in psychiatric
not well-established, and controversial findings disorders. They also proposed that sex has a sub-
have been reported (52, 53), with very limited evi- stantial impact as only the male patients with
dence of cognitive deterioration over time after the schizophrenia had significantly more morphologic
initial episodes (54, 55). Recently, our group anomalies than healthy male volunteers, especially
assessed whether neuropsychological changes in in the lateral ventricles. Moreover, independent of
patients with schizoaffective disorder persist into diagnosis, women were more likely than men to
remission or not (43). No changes in executive dys- have hyperintensity signals. A further MRI study
function were found comparing acute phase and found that 12 patients with schizoaffective disorder
remission but memory improved in patients with and 12 patients with bipolar disorder showed simi-
schizoaffective disorder, once they were in remis- larly reduced whole-brain volume compared with
sion from a manic episode of their schizoaffective 12 healthy controls (58). Other neuroimaging stud-
disorder. In contrast, in patients with schizoaffec- ies in schizoaffective disorder have focused on speci-
tive disorder, memory and executive dysfunction fic brain structures. Among these, a CT study
in the acute depressive episode were similar to examined thalamic morphology in groups of
when they were in clinical remission. The results of patients with schizophrenia and schizoaffective dis-
this study suggest an underlying trait-like cognitive order (72). They found that both patient groups
impairment in schizoaffective disorder, especially showed similar volume reductions as compared
executive dysfunction that persisted despite symp- with controls, and also similar shape distortions in
tomatic remission. medio-dorsal and ventro-lateral thalamic regions.
Moreover, distinct deformations in medial and lat-
eral thalamic regions were only found in patients
Neuroimaging findings
with schizoaffective disorder. Two other MRI stud-
Nineteen neuroimaging studies fulfilled inclusion ies focused on hippocampal volume revealed that
criteria with a total of 725 patients with schizoaf- this structure was significantly reduced in groups of
fective disorder, 933 with schizophrenia, 484 with patients with schizophrenia and schizoaffective dis-
bipolar disorder, 27 with unipolar depression and order, as compared with patients with bipolar disor-
1336 healthy subjects (43, 56–73). der and healthy subjects (57, 70). In contrast, in a
Table 2 summarizes the main findings from more recent study based on a larger sample (i.e.
structural MRI studies. n = 219 patients with schizophrenia, n = 142
patients with schizoaffective disorder, n = 188
Volumetric studies. A total of eight studies used patients with psychotic bipolar disorder and
volumetric techniques. The earliest study in this cat- n = 337 healthy controls) comparable hippocampal
egory found no differences in lateral ventricular vol- volume reductions were detected in all disorders
ume and an index of cortical atrophy in 28 patients (67). Notably, the amount of volume loss was posi-
with schizophrenia, 19 with bipolar disorder and 15 tively correlated with psychosis severity and cogni-
with schizoaffective disorder (71). In 1994, Jones tive impairment.
et al. published a computed tomography (CT) study
measuring lateral and third ventricles in 121 Voxel-based morphometry of grey matter vol-
patients with schizophrenia, 41 with schizoaffective ume. Brain structural changes in schizoaffective
disorder, 54 with affective psychosis and 67 healthy disorder, and how far they resemble those seen in
controls (62). They detected both larger lateral and schizophrenia and bipolar disorder, have been
third ventricle volumes in schizophrenia and studied to only a limited extent. To date, there

21
Madre et al.

Table 2. Structural neuroimaging studies in patients with schizoaffective disorder

Method
Population (Image-modality) Findings

Rieder et al. (71) SA = 15, SZ = 28, BD = 19 Volumetry (CT) Increased lateral ventricle volumes and cortical atrophy:
SA = SZ = BD
Jones et al. (62) SA = 41, SZ = 121, HC = 67 Volumetry (CT) Increased lateral and third ventricle volumes: SZ = SA>HC
Lewine et al. (65) SZ = 108, SA = 20, BD = 20, UD = 27,HC = 150 Volumetry (MRI) Volume loss and ventricular anomaly: SZ = SA>BD = UD
Getz et al. (58) SA = 12, BD = 12, HC = 12 Volumetry (MR) Decrease in whole-brain volume: SA = BD> HC. Globus pallidal
difference was the largest contribution to this difference
Smith et al. (72) SA = 15, SZ = 47, HC = 47 Volumetry (MRI) Volume reduction and thalamus shape deformations:
SA = SZ> HC
Radonic et al. (70) SA = 15, SZ = 15, BD = 15, HC = 15 Volumetry (MRI) Hippocampal volume reduction: SZ = SA>BD = HC
Arnold et al. (57) SA = 70, SZ = 71, P-BD = 86, Volumetry (MRI) Hippocampal volume reduction: SZ = SA>P-BD = HC =
SZ-Rel = 74, SA-Rel = 62, BD-Rel = 88, HC = 145 SZ-Rel = SA-Rel = BD-Rel
Mathew et al. (67) SA = 142, SZ = 219, P-BD = 188, HC = 337 Volumetry (MRI) Hippocampal volume reduction: SA = SZ = P-BD. Positive
correlation with psychosis severity and cognitive impairment
Ivleva et al. (61) SZ = 19, SA = 16, BD = 17, HC = 10 VBM (MRI) Widespread gray matter volume reductions: SZ = SA>BD>HC
Ivleva et al. (60) SA = 90, SZ = 146, P-BD = 15, VBM (MRI) Widespread gray matter volume reductions: SZ = SA> P-BD.
SZ-Rel = 134, SA-Rel = 106, BD-Rel = 129, HC = 200 Gray matter reductions: relatives with psychosis spectrum
disorders (n = 34)>relatives without psychosis spectrum
(n = 332) = HC
Amann et al. (56) SA = 45, SZ = 45, BD = 45, HC = 45 VBM (MRI) Widespread gray matter volume reductions: SZ = SA> BD>HC
Padmanabhan et al. (69) SA = 117, SZ = 181, P-BD = 157, HC = 352 CTh (MRI) CTh: SA = SZ = P-BD. CTh strongly associated with PANSS
positive subscale
Landin-Romero et al. (64) SA = 45, HC = 45 SBM and DTI (MRI Widespread CTh reduction and white matter disruption: SA>HC
and diffusion MRI)

SA, schizoaffective disorder; SA-M, schizoaffective disorder, manic episode; SA-D, schizoaffective disorder, depressive episode; SA-R schizoaffective disorder in clinical remis-
sion; SZ, schizophrenia; BD, bipolar disorder; P-BD, psychotic bipolar disorder; SZ-Rel, first-degree schizophrenia relatives; SA-Rel, first-degree schizoaffective disorder relatives;
BP-Rel, first-degree bipolar relatives; UD, Unipolar depression; HC, healthy controls.
CT, computed tomography, MRI, magnetic resonance imaging, DTI, diffusion tensor imaging; VBM, voxel-based morphometry; CTh, cortical thickness; SBM, surface-based mor-
phometry (including cortical thickness, surface area and cortical volume).

have been only three voxel-based morphometry disorder and healthy controls respectively (56). In
(VBM) studies of schizoaffective disorder, two of line with results from the Ivleva studies, this study
them from the same group. In the first of these, Ivl- also detected widespread reductions in GM volume
eva et al. (61) examined 19 patients with in schizoaffective disorder that more closely resem-
schizophrenia, 16 with schizoaffective disorder and ble those seen in schizophrenia.
17 with psychotic bipolar disorder, as well as 10
healthy subjects. They found decreased grey matter Diffusion tensor imaging of white and grey matter tis-
(GM) volume in the subjects with schizophrenia, sues. In the only available diffusion tensor imag-
and a similar, albeit less extensive, volume reduc- ing (DTI) study for schizoaffective disorder,
tion in the patients with schizoaffective disorder; differences in fractional anisotropy (FA) and mean
however, the patients with bipolar disorder did not diffusivity (MD) were investigated in 45 patients
differ from the controls. In a subsequent larger compared with 45 matched healthy controls (64).
VBM study which combined MRI data from four The tract-based spatial statistics analysis in white
different centres, the same research team found matter (WM) reported significant FA reduction in
that both groups of 146 patients with schizophre- patients relative to controls. The observed wide-
nia and 90 patients with schizoaffective disorder spread abnormalities were mainly located in the
showed GM volume reduction in numerous and corpus callosum, extending to the anterior limb of
overlapping areas as compared with 200 healthy the internal capsule bilaterally, the left superior
controls (60). In contrast, the GM volume reduc- longitudinal fasciculus, as well as tracts adjacent to
tion in the group of 115 patients with psychotic GM tissues in temporal structures, the right thala-
bipolar disorder was limited to the frontotemporal mus and the left lingual. On the other hand, the
cortex. It should be noted that in both studies no whole-brain voxel-based comparison of MD
statistical differences were detected in pairwise showed significant increases in patients in both
comparisons between the three patient groups. The WM and GM tissues.
last VBM study examined a group of 45 patients
meeting DSM-IV and RDC criteria for schizoaf- Surface-based morphometry of the cortical man-
fective disorder, and three groups of 45 well- tle. Two studies have used surface-based mor-
matched subjects with schizophrenia, bipolar phometry techniques. In the first study,

22
 Neurobiology of schizoaffective disorder

measurements of CV, CT and cortical SA were patients with schizoaffective disorder as a distinct
compared in 45 patients with schizoaffective disor- clinical group. One study compared the fMRI
der and 45 matched healthy controls (64). Briefly, activity of the neural networks involved in both
a widespread bilateral pattern of CT reduction was verbal and visuospatial working memory process-
observed in patients that resembles results from ing in terms of behavioural performance in 14
the above mentioned VBM studies. In addition, patients with schizophrenia, 14 with schizoaffec-
decreased cortical SA was observed in the left fusi- tive disorder and 14 healthy controls (59). They
form region. Finally, there was a widespread pat- found that patients with schizophrenia had pro-
tern of reduced CV, similar (although more nounced impairments in both neural networks,
restricted) to that observed in the CT analysis. As whereas verbal working memory performance
this study also conducted a DTI analysis (see the was preserved in patients with schizoaffective dis-
previous subsection), overlapping regions of multi- order. A proton magnetic resonance spectroscopy
modal abnormalities were detected by the different study compared 15 patients with schizoaffective
contrasts. Taken together, the reduced GM and disorder, 15 with schizophrenia and 15 with bipo-
WM volume in patients and the surface-based lar disorder, as well as 15 healthy subjects (63).
morphometry results indicate that GM abnormal- They found reduced levels of N-acetyl aspartate
ity is mainly driven by cortical thinning. Interest- in all patients groups in the dorsolateral pre-
ingly, the multimodal abnormalities were mainly frontal cortex (DLPFC), but this anomaly was
detected in areas that have been consistently not a distinctive feature among these three ill-
reported to be altered in schizophrenia (74), and to nesses. The levels of compounds containing cho-
some extent in bipolar disorder (75), which may line and creatine were similar in patients with
explain part of the common symptomatology schizoaffective and bipolar disorders. However,
related to these disorders. In the second available the levels of these compounds were not related to
study, Padmanabhan and colleagues (69) analysed executive functions and attention performance, as
the correlations between surface-based morphome- determined from the ‘Wisconsin Card Sorting’
try measurements and symptom dimensions of and ‘Stroop’ Tests. Another study by Ongur et al.
psychosis in 455 individuals with schizophrenia, (68) investigated resting state functional connec-
schizoaffective, or bipolar I disorders. They found tivity using independent component analysis and
that the positive symptom subscale correlated compared the wider default mode network
inversely with grey matter volume and CT in fron- (DMN) at rest between schizophrenia, bipolar
tal and temporal regions. Moreover, CT demon- disorder, and healthy subjects. In a subanalysis,
strated more associations with psychopathology they found reduced connectivity in the medial
than cortical SA. frontal cortex in seven patients with schizoaffec-
Figure 1 shows an anatomical summary of the tive disorder, which was similar in degree to that
main structural findings in schizoaffective disorder. seen in seven patients with schizophrenia. How-
ever, this region was less disconnected in 14
Functional neuroimaging. Brain activity can be patients with bipolar disorder. Recently, our
assessed with different neuroimaging techniques group reported on task fMRI using a working
such as magnetic resonance spectroscopy, PET or memory task in 32 acutely ill patients with
functional MRI. Functional MRI assesses neural schizoaffective disorder. Patients showed less neu-
activity while performing a task (task based ral activity in prefrontal, parietal and temporal
fMRI) or while participants are at rest (resting regions than healthy controls (66). In addition,
state fMRI). Few MRI studies so far considered patients did not de-activate the medial frontal

Fig. 1. Principal regions of altered gray matter morphology in the schizoaffective brain. V, volume; CT, cortical thickness; SA, sur-
face area.

23
Madre et al.

cortex, suggestive of DMN dysfunction. Later, 22 to the healthy controls. Table 3 summarizes main
of these patients were followed-up and rescanned findings from functional MRI studies.
after 2 months of clinical remission (43). Those in Figure 2 depicts an anatomical overview of the
remission from a manic episode also reverted to main functional findings in schizoaffective disor-
normal in their frontal lobe task related neural der.
activation patterns while patients in remission Interestingly, the affected brain regions reported
from a depressive episode did not. Interestingly, in the aforementioned functional studies were very
the whole group of patients with schizoaffective similar to those revealing structural abnormalities
disorder in clinical remission showed a failure of (see Figs. 1 and 2). These included the superior
de-activation in the medial frontal gyrus com- prefrontal cortex, the anterior and posterior cingu-
pared with the healthy controls, which reflected lated cortex, the orbitrofrontal and dorsolateral
again a DMN dysfunction, as a trait-like feature prefrontal areas, the parietal, temporal and occipi-
of schizoaffective disorder. Similar results were tal cortex, the insula, the cerebellum, the hip-
observed in bipolar disorder showing that manic, pocampus and thalamic nuclei.
depressed and euthymic patients had a failure of
de-activation in the medial frontal cortex, sup-
Discussion
porting also the concept of ‘trait’ DMN dysfunc-
tion (76). Several studies have also reported The present article provides a systematic review of
DMN dysfunction, in terms of less deactivation neuropsychological and neuroimaging data in
of the medial frontal gyrus, in schizophrenia and schizoaffective disorder, a neglected and controver-
bipolar disorder (77–81). A recent study used a sial condition (82). The reliability (83) and long-
group information guided ICA method, designed term stability (84) of schizoaffective disorder have
to capture individual functional networks and to proven to be high, and therefore a thorough review
simultaneously preserve correspondence of net- of its nosologic validity from the neuropsychologi-
works across subjects (73). Resting-state func- cal and neuroimaging standpoint is fully justified.
tional MRI data of 20 patients with The available data do not yet suffice to define
schizophrenia, 20 with bipolar disorder, 20 with a schizoaffective disorder as a neurobiological entity.
manic and 13 with a depressive episode of their What is evident, though, is that these patients in
schizoaffective disorder were compared with 20 general suffer from neurocognitive and brain struc-
healthy subjects. They found that, at a subject tural and functional deterioration. In the follow-
level, subjects from the same group had in gen- ing, we will synthesize current knowledge on
eral similar network patterns. At a group level, cognitive function, brain structure and brain func-
subjects with a manic and a depressive episode of tion in schizoaffective disorder.
their schizoaffective disorder were the most simi-
lar groups to each other. These two groups also Cognition. As highlighted, cognitive impairment is
shared high similarity with the group of patients a well-established finding in schizophrenia (e.g 48)
with schizophrenia. Moreover, the group of sub- and bipolar disorder (e.g 85, 86). Even though data
jects with bipolar disorder was the most similar are less robust in schizoaffective disorder, the

Table 3. Functional neuroimaging studies in schizoaffective disorder

Population Method Findings

Gruber et al. (59) SA = 14, SZ = 14, HC = 14 fMRI (verbal and Impaired Verbal WM: SZ = SA>HC
visuospatial WM task) Impaired Visuospatial WM: SZ>SA = HC. Activation in
the left premotor–parietal cortex: SA = HC>SZ
Kalayci et al. (63) SA = 5, SZ = 15, BD = 15, HC = 15 H-MRS DLPFC NAA levels reduction: SZ = SA = BD>HC. Levels of
Cho and Cr compounds: SA = BD
Ongur et al. (68) SA = 7 + SZ = 7, BD = 14, HC = 15 fMRI resting (ICA: mPFC) Resting state connectivity reduction in the mPFC: SA = SZ>BD
Madre et al. (66) SA-M = 16, SA-D = 16, HC = 32 fMRI (N-back task) Frontal hypo-activations and failure to deactivate the MFC:
SA-M = SA-D>HC
Madre et al. (43) SA-M = 11, SA-D = 11, SA-R = 22, HC = 22 fMRI (N-back task) Frontal hypo-activations: SA-M>SA-R = HC; SA-D = SA-R = HC.
Failure to deactivate the mPFC: SA-M = SA-D = SA-R>HC
Du et al. (73) BD = 20, SZ = 20, SA-M = 20, SA-D = 13, HC = 20 fMRI (GIG-ICA) 12 regions distinguished SZ, SA and BD. SA-M = SA-D (but SA-
D also similar SZ). BD similar to HC

SA, schizoaffective disorder; SA-M, schizoaffective disorder, manic episode; SA-D, schizoaffective disorder, depressive episode; SZ, schizophrenia; BD, bipolar disorder; HC,
healthy controls.
MRI, magnetic resonance imaging, H-MRS, proton magnetic resonance spectroscopy, DLPFC, Dorsolateral Prefrontal Cortex; NAA, N-acetyl aspartate; Cho, Choline; Cr, Creatine;
mPFC, medial prefrontal cortex; ICA, Independent component analysis, GIG-ICA, Group information guided Independent component analysis; WM, Working memory.

24
 Neurobiology of schizoaffective disorder

Fig. 2. Principal regions of changes in activations in the schizoaffective brain. FDA, failure of deactivations; HP, hypoactivation;
NA, not available.

pattern of impairment appears to be broadly simi- with some, albeit partial improvement when in
lar in all disorders: poor performance is seen in remission (43).
most or all domains of cognition, but deficits are Taken together, the available data suggest that
disproportionately marked in executive function, cognitive impairment in schizophrenia and
long-term memory and sustained attention. Cogni- schizoaffective disorder may be greater than in
tive impairment is often considered to be greater in bipolar disorder. Nevertheless, the patterns of cog-
schizophrenia than in bipolar disorder, but a meta- nitive impairment appear similar in all three disor-
analysis (47) found only partial support for this ders without major improvement in clinical
view. The differences were generally small and not remission.
significant in around half the domains of cognition
examined. When schizophrenia was compared sep- Structural neuroimaging. The available data from
arately with schizoaffective disorder and affective structural MRI suggest that patients with schizoaf-
disorders, the magnitude of the differences between fective disorder show similar grey matter volume
schizophrenia and affective disorders were similar reduction as patients with schizophrenia, and more
or larger than differences between schizophrenia than patients with bipolar disorder. This has been
and schizoaffective disorder in some domains of suggested by three VBM imaging studies (56, 60,
cognitive function, but in others schizoaffective 61). In the study of Ivleva et al. (60), the number
disorder showed non-significantly smaller effect of participants was much larger than in the study
sizes than those for affective disorders. by Amann et al. (56). However, since MRI data
With respect to the state vs. trait characteristics were collected at four different centres this may
of cognitive impairment in schizoaffective disorder, have introduced noise into the analysis. Notable,
results were mixed in a first preliminary trial (43): Ellison-Wright and Bullmore (87) found also evi-
memory but not executive function improved sig- dence for more extensive grey matter volume
nificantly in clinical remission in patients with reductions in schizophrenia compared with bipolar
schizoaffective disorder, currently in a manic epi- disorder in their meta-analysis of VBM studies.
sode, but neither domain of function changed Nevertheless, a later meta-analysis by Bora et al.
between relapse and clinical remission in patients (88) reported that the differences became consider-
with schizoaffective disorder, currently in a ably smaller when correcting for the predominance
depressed episode. In schizophrenia, the difference of male patients in the schizophrenia studies. De
in cognition between relapse and remission has Peri et al. (89) also found overall largely similar
received little attention (48, 50) with evidence to whole-brain volume differences to controls in first-
suggest cognition remains stable. The traditional episode patients with schizophrenia and bipolar
view of cognitive impairment in bipolar disorder is disorder in a meta-analysis of conventional volu-
that it is state-related and reverts to normal in metric MRI studies. Grey matter reductions were
remission (e.g (49)). However, the recognition of more pronounced in schizophrenia and white mat-
euthymic cognitive impairment over the last 15– ter reductions in bipolar disorder.
20 years now makes this assumption questionable.
Indeed, a large study (52) found few significant dif- Functional neuroimaging. Task based functional
ferences between depressed, manic and euthymic MRI studies have not resulted in a clear distinction
patients with bipolar disorder and no clear pattern between bipolar disorder and schizophrenia. How-
could be identified to distinguish the three clinical ever, cohorts were small (59, 63, 68). A study using
states. The neuropsychological findings in a larger sample provided evidence that schizoaffec-
schizoaffective disorder reveals a similar pattern tive disorder is characterized by a general DMN

25
Madre et al.

dysfunction in relation to controls (66), with first nia with regard to neurocognitive and neu-
evidence that this might be a trait characteristic of roimaging data, some studies have found further
the disease (43). This has also been found in similarities between schizophrenia and schizoaf-
schizophrenia (77–81) bipolar disorder (76), as well fective disorder. For instance, the meta-analysis
as in a range of psychiatric and neuropsychiatric by Pagel (104) found a stronger resemblance of
disorders (90). The DMN is active at rest and schizoaffective disorder to schizophrenia than to
reduces its activity when the individual engages in bipolar disorder in most of the demographic and
a cognitive task (91, 92). The medial prefrontal clinical variables. These results support the
cortex and the posterior cingulate cortex form the hypothesis that schizoaffective disorder is not
two midline nodes of the DMN (93). Clues to the primarily an affective disorder; however, the
function of the DMN come from a small number authors stated that it remains unclear whether
of studies using tasks which, rather than producing the findings reflect the nature of the disorder or
de-activation, increase activity in parts of it. Such whether they are simply an artifact of current
tasks often share a component of introspective or diagnostic criteria.
self-related thought: recall of personal experiences, Regardless of its controversial status, schizoaf-
making social and emotional judgements, envision- fective disorder is clinically a frequent diagnosis
ing the future and performing theory of mind tasks (105). Until further studies will be carried out to
(94). So far, only one resting-state fMRI study clarify its nosology, schizoaffective disorder is a
using a group information guided ICA method necessary and useful diagnosis for clinicians. The
compared patients with schizophrenia, bipolar dis- clinical concept of a subgroup of psychotic patients
order, schizoaffective disorder (currently in a with prominent affective symptoms remains impor-
manic or depressed episode) and healthy subjects tant (106), even though the available evidence does
(73). Results from that network-based functional not indicate whether its neurobiological underpin-
study suggest that schizoaffective disorder is an nings are unique. This concept may still be relevant
independent entity but with a high similarity to for clinical practice including therapeutic implica-
schizophrenia. tions (107).
In conclusion, neurocognitive data suggest that
The neurobiology of schizoaffective disorder. One of cognition in schizoaffective disorder is similarly
our aims was to help to understand how affected as in schizophrenia, and more than in
schizoaffective disorder fits within the schizophre- bipolar disorder, especially in absence of psychotic
nia – bipolar disorder nosological framework. symptoms. Functional and structural imaging
Evidence so far is clearly incompatible with the abnormalities point towards more impairment in
argument that schizoaffective disorder is a vari- schizoaffective disorder and schizophrenia than in
ant of bipolar disorder (10, 95, 96). It is also dif- bipolar disorder. This suggests that schizoaffective
ficult to reconcile the available data with the disorder is a subtype of schizophrenia or part of
view that schizoaffective disorder represents the the continuous spectrum model of psychosis, with
midpoint on a continuous spectrum of psychosis, schizoaffective disorder skewed more towards
as conceptualized either genetically (6, 97) clini- schizophrenia than bipolar disorder.
cally (98–101) or both genetically and clinically
(102). This would predict a degree of volume
Limitations
and functional changes in schizoaffective disorder
between schizophrenia and bipolar disorder, Some limitations of this review should be taken
something which cannot be drawn as conclusion into account when results are translated into clini-
of studies included in this review. Findings so far cal praxis. Small number of neuroimaging studies
appear most compatible with the view that were included for each modality because of which
schizoaffective disorder is a variant of data could not be meta-analysed. Therefore, con-
schizophrenia. This is a proposal that is currently clusions in general have to be considered as prelim-
unfashionable, to the extent that it is difficult to inary. Some of the neuroimaging techniques have
find publications that have advocated this as well known limitations. For instance, in structural
Lehman et al. (103) did so 40 years ago. Another imaging, VBM is susceptible to artefacts such as
less controversial possibility might be that poor segmentation and poor registration. It does
schizoaffective disorder is an intermediate form not allow determining whether the identified differ-
of illness, but one that is, for unknown reasons, ences in volume are because of changes in cortical
skewed towards schizophrenia. However, it has thickness or cortical surface area. Likewise, stan-
also to be mentioned, that besides the overlap dard volumetric techniques are also prone to
between schizoaffective disorder with schizophre- motion artefacts, partial volume contamination

26
 Neurobiology of schizoaffective disorder

and manual or automatic segmentation errors.

References
Surface-based morphometry techniques also have
their own limitations, including artefacts because 1. Kraepelin EE. Psychiatrie. 6th ed. Leipzig: Barth; 1899.
of insufficient voxel resolution for optimal calcula- 2. Kraepelin E. Die Erscheinungsformen des Irreseins. Zeits-
chrift f€ur die gesamte Neurologie und Psychiatrie
tion in highly convoluted areas. Thus, without 1920;62:1–29.
standardized quality control protocols, it is diffi- 3. Kasanin K. The acute schizoaffective psychoses. Am J
cult to compare results obtained with different Psychiatry 1933;90:97–126.
scanner types, sequence parameters and field 4. Peralta V, Cuesta MJ. Exploring the borders of the
strengths. White matter macro- and micro-struc- schizoaffective spectrum: a categorical and dimensional
approach. J Affect Disord 2008;108:71–86.
ture have not been investigated in detail. There are 5. Kendler KS, Neale MC, Walsh D. Evaluating the spec-
no reports comparing schizoaffective disorder, trum concept of schizophrenia in the Roscommon Family
schizophrenia and bipolar disorder by means of Study. Am J Psychiatry 1995;152:749–754.
purely discriminative methods like machine learn- 6. Crow TJ. The continuum of psychosis and its implication
ing. Furthermore, more comparative neuroimag- for the structure of the gene. Br J Psychiatry
1986;149:419–429.
ing studies across mood and psychotic disorders
7. Beck AT. Depression: Causes and Treatment. Philadel-
are urgently needed, such as functional MRI stud- phia: University of Pennsylvania Press; 1967.
ies across all three diseases. Last but not least, 8. Welner A, Croughan J, Fishman R, Robins E. The group of
some studies included subjects with unipolar and schizoaffective and related psychoses: a follow-up study.
bipolar schizoaffective disorder; a separate analysis Compr Psychiat 1977;18:413–422.
of both subtypes might change some results. 9. Lehman H. Schizophrenia: clinical features. In: Freedman
A, Kaplan H, Saddock J, editors. Comprehensive Text-
book of Psychiatry. Baltimore: Williams & Wilkins; 1975:
Acknowledgements pp 457–486.
10. Pope HG jr., Lipinski JF, Cohen BM, Axelrod DT.
We acknowledge the generous support by the Centro de Inves- “Schizoaffective disorder”: an invalid diagnosis? A com-
tigaci

on Biom
edica en Red de Salud Mental (CIBERSAM). parison of schizoaffective disorder, schizophrenia, and
This work was also supported by the Catalonian Government affective disorder. Am J Psychiatry 1980;137:921–927.
(2014-SGR-1573 to the Research Unit of FIDMAG) and sev- 11. Procci WR. Schizo-affective psychosis: fact or fiction? A
eral grants from the Plan Nacional de I+D+i and co-funded by survey of the literature. Arch Gen Psychiatry
the Instituto de Salud Carlos III-Subdirecci
on General de 1976;33:1167–1178.
Evaluaci
on y Fomento de la Investigaci
on, Plan Nacional 12. Maier W. Do schizoaffective disorders exist at all? Acta
2008–2011 and 2013–2016, and the European Regional Devel- Psychiat Scand 2006;113:369–371.
opment Fund (FEDER): Stabilization Contract grant (CES12/ 13. Lake CR, Hurwitz N. Schizoaffective disorders are psy-
024) and Research Projects PI07/1278, PI10/02622 and PI/15/ chotic mood disorders; there are no schizoaffective disor-
02242 to Dr. BL. Amann and to Dr. E. Vieta (PI 12/00910 and ders. Psychiatry Res 2006;143:255–287.
2014 SGR 398). Finally, we thank PJ McKenna for his revision 14. APA. Diagnostic and statistical manual of mental disor-
of the PhD thesis of Merce Madre, parts of which were inte- ders, 5th edn (DSM-V). Washington DC: American Psy-
grated in this manuscript. chiatric Association, 2013.
15. Cosgrove VE, Suppes T. Informing DSM-5: biological
boundaries between bipolar I disorder, schizoaffective
Declaration of interests disorder, and schizophrenia. BMC Med 2013;11:127.
This work was supported by the Centro de Investigaci
on 16. Abrams DJ, Rojas DC, Arciniegas DB. Is schizoaffective
Biom
edica en Red de Salud Mental (CIBERSAM), the Catalo- disorder a distinct categorical diagnosis? A critical review
nian Government (2009SGR211 to the Research Unit of Ben- of the literature. Neuropsychiatr Dis Treat 2008;4:1089–
ito Menni) and the Instituto de Salud Carlos III. The funding 1109.
organizations played no role in the study design, data collec- 17. Malhi GS, Green M, Fagiolini A, Peselow ED, Kumari V.
tion and analysis, or manuscript approval. Dr. BL Amann has Schizoaffective disorder: diagnostic issues and future rec-
served as speaker for Janssen, Lundbeck and Otsuka. Dr. A. ommendations. Bipolar Disord 2008;10:215–230.
Murru has received support from Bristol-Myers-Squibb and 18. Haijma SV, Van HAREN N, Cahn W, Koolschijn PC, Hul-
has been speaker for Astra Zeneca. Dr. V. Perez has been a shoff Pol HE, Kahn RS. Brain volumes in schizophrenia:
consultant to or has received honoraria or grants from Astra- a meta-analysis in over 18 000 subjects. Schizophrenia
Zeneca, Bristol-Myers-Squibb, Janssen Cilag, Lundbeck, Bull 2013;39:1129–1138.
Otsuka, Servier and Medtronic. Dr. Vieta has received grants 19. Arnone D, Cavanagh J, Gerber D, Lawrie SM, Ebmeier
and served as consultant, advisor or CME speaker for the fol- KP, McIntosh AM. Magnetic resonance imaging studies
lowing entities: AstraZeneca, Bristol-Myers Squibb, Elan, Eli in bipolar disorder and schizophrenia: meta-analysis. Br
Lilly, Ferrer, Forest Research Institute, Gedeon Richter, J Psychiatry 2009;195:194–201.
Glaxo-Smith-Kline, Janssen, Lundbeck, Otsuka, Pfizer, 20. Kempton MJ, Geddes JR, Ettinger U, Williams SC, Grasby
Roche, Sanofi-Aventis, Servier, Sunovion, Takeda, Teva, the PM. Meta-analysis, database, and meta-regression of 98
Spanish Ministry of Science and Innovation (CIBERSAM), structural imaging studies in bipolar disorder. Arch Gen
the Seventh European Framework Programme (ENBREC), Psychiatry 2008;65:1017–1032.
and the Stanley Medical Research Institute. The rest of the 21. D’Albis MA, Houenou J. The kraepelinian dichotomy
authors have declared that there are no conflicts of interest in viewed by neuroimaging. Schizophrenia Bull
the last 2 years. 2015;41:330–335.

27
Madre et al.

22. Trotta A, Murray RM, Maccabe JH. Do premorbid and ders depends on history of psychosis rather than diagnos-
post-onset cognitive functioning differ between tic group. Schizophr Bull 2011;37:73–83.
schizophrenia and bipolar disorder? A systematic review 40. Amann B, Gomar JJ, Ortiz-Gil J et al. Executive dysfunc-
and meta-analysis. Psychol Med 2015;45:381–394. tion and memory impairment in schizoaffective disorder:
23. Moher D, Liberati A, Tetzlaff J, Altman DG. Preferred a comparison with bipolar disorder, schizophrenia and
reporting items for systematic reviews and meta-analyses: healthy controls. Psychol Med 2012;42:2127–2135.
the PRISMA statement. BMJ 2009;339:b2535. 41. Torniainen M, Suvisaari J, Partonen T et al. Cognitive
24. Miller LS, Swanson-Green T, Moses JA Jr., Faustman impairments in schizophrenia and schizoaffective disor-
WO. Comparison of cognitive performance in RDC- der: relationship with clinical characteristics. J Nerv
diagnosed schizoaffective and schizophrenic patients with Ment Dis 2012;200:316–322.
the Luria-Nebraska Neuropsychological Battery. J Psy- 42. Derosse P, Burdick KE, Lencz T, Siris SG, Malhotra AK.
chiatr Res 1996;30:277–282. Empirical support for DSM-IV schizoaffective disorder:
25. Manschreck TC, Maher BA, Beaudette SM, Redmond DA. clinical and cognitive validators from a large patient sam-
Context memory in schizoaffective and schizophrenic dis- ple. PLoS ONE 2013;8:e63734.
orders. Schizophr Res 1997;26:153–161. 43. Madre M, Radua J, Landin-Romero R et al. Trait or state?
26. Evans JD, Heaton RK, Paulsen JS, McAdams LA, Heaton A longitudinal neuropsychological evaluation and fMRI
SC, Jeste DV. Schizoaffective disorder: a form of study in schizoaffective disorder. Schizophr Res
schizophrenia or affective disorder? J Clin Psychiatry 2014;159:458–464.
1999;60:874–882. 44. Fleck DE, Delbello MP, Zimmerman ME, Getz GE, Shear
27. Gooding DC, Tallent KA. Spatial working memory per- PK, Strakowski SM. Selective attention in schizoaffective
formance in patients with schizoaffective psychosis versus disorder: pilot data. Schizophr Res 2004;68:109–110.
schizophrenia: a tale of two disorders? Schizophr Res 45. Hill SK, Reilly JL, Keefe RS et al. Neuropsychological
2002;53:209–218. impairments in schizophrenia and psychotic bipolar dis-
28. Glahn DC, Bearden CE, Cakir S et al. Differential work- order: findings from the Bipolar-Schizophrenia Network
ing memory impairment in bipolar disorder and on Intermediate Phenotypes (B-SNIP) study. Am J Psy-
schizophrenia: effects of lifetime history of psychosis. chiatry 2013;170:1275–1284.
Bipolar Disord 2006;8:117–123. 46. Buchanan RW, Davis M, Goff D et al. A summary of the
29. Roofeh D, Cottone J, Burdick KE et al. Deficits in mem- FDA-NIMH-MATRICS workshop on clinical trial
ory strategy use are related to verbal memory impair- design for neurocognitive drugs for schizophrenia. Schi-
ments in adolescents with schizophrenia-spectrum zophr Bull 2005;31:5–19.
disorders. Schizophr Res 2006;85:201–212. 47. Bora E, Yucel M, Pantelis C. Cognitive functioning in
30. Goldstein G, Shemansky WJ, Allen DN. Cognitive func- schizophrenia, schizoaffective disorder and affective psy-
tion in schizoaffective disorder and clinical subtypes of choses: meta-analytic study. Br J Psychiatry
schizophrenia. Arch Clin Neuropsychol 2005;20:153–159. 2009;195:475–482.
31. Stip E, Sepehry AA, Prouteau A et al. Cognitive dis- 48. Palmer BW, Dawes SE, Heaton RK. What do we know
cernible factors between schizophrenia and schizoaffec- about neuropsychological aspects of schizophrenia? Neu-
tive disorder. Brain Cogn 2005;59:292–295. ropsychol Rev 2009;19:365–384.
32. Fiszdon JM, Richardson R, Greig T, Bell MD. A compar- 49. Murray RM, Sham P, van Os J, Zanelli J, Cannon M,
ison of basic and social cognition between schizophrenia McDonald C. A developmental model for similarities and
and schizoaffective disorder. Schizophr Res 2007;91:117– dissimilarities between schizophrenia and bipolar disor-
121. der. Schizophr Res 2004;71:405–416.
33. Torrent C, Martinez-Aran A, Amann B et al. Cognitive 50. Rund BR. Is schizophrenia a neurodegenerative disorder?
impairment in schizoaffective disorder: a comparison Nord J Psychiatry 2009;63:196–201.
with non-psychotic bipolar and healthy subjects. Acta 51. Dickerson F, Schroeder J, Stallings C et al. A longitudi-
Psychiatr Scand 2007;116:453–460. nal study of cognitive functioning in schizophrenia: clini-
34. Szoke A, Meary A, Trandafir A et al. Executive deficits in cal and biological predictors. Schizophr Res
psychotic and bipolar disorders - implications for our 2014;156:248–253.
understanding of schizoaffective disorder. Eur Psychiat 52. Martinez-Aran A, Vieta E, Reinares M et al. Cognitive
2008;23:20–25. function across manic or hypomanic, depressed, and
35. Heinrichs RW, Ammari N, McDermid VAZS, Miles AA. euthymic states in bipolar disorder. Am J Psychiatry
Are schizophrenia and schizoaffective disorder neuropsy- 2004;161:262–270.
chologically distinguishable? Schizophr Res 2008;99:149– 53. Xu G, Lin K, Rao D et al. Neuropsychological perfor-
154. mance in bipolar I, bipolar II and unipolar depression
36. Reichenberg A, Harvey PD, Bowie CR et al. Neuropsy- patients: a longitudinal, naturalistic study. J Affect Dis-
chological function and dysfunction in schizophrenia and ord 2012;136:328–339.
psychotic affective disorders. Schizophrenia Bull 54. Torrent C, Martinez-Aran A, Del Mar Bonnin C et al.
2009;35:1022–1029. Long-term outcome of cognitive impairment in bipolar
37. Hooper SR, Giuliano AJ, Youngstrom EA et al. Neurocog- disorder. J Clin Psychiatry 2012;73:e899–e905.
nition in early-onset schizophrenia and schizoaffective 55. Mora E, Portella MJ, Forcada I, Vieta E, Mur M. Persis-
disorders. J Am Acad Child Adolesc Psychiatry tence of cognitive impairment and its negative impact on
2010;49:52–60. psychosocial functioning in lithium-treated, euthymic
38. Studentkowski G, Scheele D, Calabrese P et al. Cognitive bipolar patients: a 6-year follow-up study. Psychol Med
impairment in patients with a schizoaffective disorder: a 2013;43:1187–1196.
comparison with bipolar patients in euthymia. Eur J Med 56. Amann BL, Canales-Rodr
ıguez E, Madre M et al. Brain
Res 2010;15:70–78. structural changes in schizoaffective disorder compared
39. Simonsen C, Sundet K, Vaskinn A et al. Neurocognitive to schizophrenia and bipolar disorder. Acta Psychiat
dysfunction in bipolar and schizophrenia spectrum disor- Scand 2016;133:23–33.

28
 Neurobiology of schizoaffective disorder

57. Arnold SJ, Ivleva EI, Gopal TA et al. Hippocampal vol- 73. Du Y, Pearlson GD, Liu J et al. A group ICA based
ume is reduced in schizophrenia and schizoaffective disor- framework for evaluating resting fMRI markers when
der but not in psychotic bipolar I disorder demonstrated disease categories are unclear: Application to schizophre-
by both manual tracing and automated parcellation nia, bipolar, and schizoaffective disorders. NeuroImage
(FreeSurfer). Schizophr Bull 2015;41:233–249. 2015;122:272–280.
58. Getz GE, Delbello MP, Fleck DE, Zimmerman ME, Sch- 74. Pomarol-Clotet E, Canales-Rodr
ıguez EJ, Salvador R
wiers ML, Strakowski SM. Neuroanatomic characteriza- et al. Medial prefrontal cortex pathology in schizophre-
tion of schizoaffective disorder using MRI: a pilot study. nia as revealed by convergent findings from multimodal
Schizophr Res 2002;55:55–59. imaging. Mol Psychiatr 2010;15:823–830.
59. Gruber O, Gruber E, Falkai P. Articulatory rehearsal in 75. Canales-Rodr
ıguez EJ, Pomarol-Clotet E, Radua J et al.
verbal working memory: a possible neurocognitive Structural abnormalities in bipolar euthymia: a multicon-
endophenotype that differentiates between schizophrenia trast molecular diffusion imaging study. Biol Psychiatry
and schizoaffective disorder. Neurosci Lett 2006;405:24– 2014;76:239–248.
28. 76. Pomarol-Clotet E, Alonso-Lana S, Moro N et al. Brain
60. Ivleva EI, Bidesi AS, Keshavan MS et al. Gray matter vol- functional changes across the different phases of bipolar
ume as an intermediate phenotype for psychosis: Bipolar- disorder. Br J Psychiatry 2014;206:136–144.
Schizophrenia Network on Intermediate Phenotypes (B- 77. Pomarol-Clotet E, Salvador R, Sarro S et al. Failure to
SNIP). Am J Psychiatry 2013;170:1285–1296. deactivate in the prefrontal cortex in schizophrenia: dys-
61. Ivleva EI, Bidesi AS, Thomas BP et al. Brain gray matter function of the default mode network? Psychol Med
phenotypes across the psychosis dimension. Psychiatry 2008;38:1185–1193.
Res 2012;204:13–24. 78. Whitfield-Gabrieli S, Thermenos HW, Milanovic S et al.
62. Jones PB, Harvey I, Lewis SW et al. Cerebral ventricle Hyperactivity and hyperconnectivity of the default net-
dimensions as risk factors for schizophrenia and affective work in schizophrenia and in first-degree relatives of per-
psychosis: an epidemiological approach to analysis. Psy- sons with schizophrenia. Proc Natl Acad Sci U S A
chol Med 1994;24:995–1011. 2009;106:1279–1284.
63. Kalayci D, Ozdel O, Sozeri-Varma G, Kiroglu Y, Tum- 79. Milanovic SM, Thermenos HW, Goldstein JM et al. Med-
kaya S. A proton magnetic resonance spectroscopy study ial prefrontal cortical activation during working memory
in schizoaffective disorder: comparison of bipolar disor- differentiates schizophrenia and bipolar psychotic
der and schizophrenia. Prog Neuropsychopharmacol Biol patients: A pilot FMRI study. Schizophr Res
Psychiatry 2012;37:176–181. 2011;129:208–210.
64. Landin-Romero R, Canales-Rodr
ıguez EJ, Kumfor F et al. 80. Salgado-Pineda P, Fakra E, Delaveau P, McKenna PJ,
White matter and cortical thickness abnormalities in Pomarol-Clotet E, Blin O. Correlated structural and
schizoaffective disorder: A multimodal MRI study. Aust functional brain abnormalities in the default mode net-
NZ J Psychiat 2016; pii: 0004867416631827. [Epub ahead work in schizophrenia patients. Schizophr Res
of print]. 2011;125:101–109.
65. Lewine RR, Hudgins P, Brown F, Caudle J, Risch SC. Dif- 81. Schneider FC, Royer A, Grosselin A et al. Modulation of
ferences in qualitative brain morphology findings in the default mode network is task-dependant in chronic
schizophrenia, major depression, bipolar disorder, and schizophrenia patients. Schizophr Res 2011;125:110–117.
normal volunteers. Schizophr Res 1995;15:253–259. 82. Murru A, Pacchiarotti I, Nivoli AM, Colom F, Vieta E. Is
66. Madre M, Pomarol-Clotet E, McKenna P et al. Brain schizoaffective disorder still a neglected condition in the
functional abnormality in schizo-affective disorder: an scientific literature? Psychother Psychosom 2012;81:389–
fMRI study. Psychol Med 2013;43:143–153. 390.
67. Mathew I, Gardin TM, Tandon N et al. Medial Temporal 83. Freedman R, Lewis DA, Michels R et al. The initial field
Lobe Structures and Hippocampal Subfields in Psychotic trials of DSM-5: new blooms and old thorns. Am J Psy-
Disorders: Findings From the Bipolar-Schizophrenia chiatry 2013;170:1–5.
Network on Intermediate Phenotypes (B-SNIP) Study. 84. Salvatore P, Baldessarini RJ, Tohen M et al. McLean-
JAMA Psychiatry 2014;71():769–777. Harvard International First-Episode Project: two-year
68. Ongur D, Lundy M, Greenhouse I et al. Default mode net- stability of ICD-10 diagnoses in 500 first-episode psy-
work abnormalities in bipolar disorder and schizophre- chotic disorder patients. J Clin Psychiatry 2011;72:183–
nia. Psychiatry Res 2010;183:59–68. 193.
69. Padmanabhan JL, Tandon N, Haller CS et al. Correlations 85. Robinson LJ, Thompson JM, Gallagher P et al. A meta-
between brain structure and symptom dimensions of psy- analysis of cognitive deficits in euthymic patients with
chosis in schizophrenia, schizoaffective, and psychotic bipolar disorder. J Affec Disorders 2006;93:105–115.
bipolar I disorders. Schizophr Bull 2015;41:154–162. 86. Mann-Wrobel MC, Carreno JT, Dickinson D. Meta-analy-
70. Radonic E, Rados M, Kalember P, Bajs-Janovic M, Folne- sis of neuropsychological functioning in euthymic bipolar
govic-Smalc V, Henigsberg N. Comparison of hippocam- disorder: an update and investigation of moderator vari-
pal volumes in schizophrenia, schizoaffective and bipolar ables. Bipolar Disord 2011;13:334–342.
disorder. Coll Antropol 2011;35(Suppl 1):249–252. 87. Ellison-Wright I, Bullmore E. Anatomy of bipolar disor-
71. Rieder RO, Mann LS, Weinberger DR, van Kammen DP, der and schizophrenia: a meta-analysis. Schizophr Res
Post RM. Computed tomographic scans in patients with 2010;117:1–12.
schizophrenia, schizoaffective, and bipolar affective disor- 88. Bora E, Fornito A, Yucel M, Pantelis C. The effects of
der. Arch Gen Psychiat 1983;40:735–739. gender on grey matter abnormalities in major psychoses:
72. Smith MJ, Wang L, Cronenwett W, Mamah D, Barch DM, a comparative voxelwise meta-analysis of schizophrenia
Csernansky JG. Thalamic morphology in schizophrenia and bipolar disorder. Psychol Med 2012;42:295–307.
and schizoaffective disorder. J Psychiatr Res 2011;45: 89. de Peri L, Crescini A, Deste G, Fusar-Poli P, Sacchetti E,
378–385. Vita A. Brain structural abnormalities at the onset of

29
Madre et al.

schizophrenia and bipolar disorder: a meta-analysis of 99. Marneros A, Deister A, Rohde A. Psychopathological and
controlled magnetic resonance imaging studies. Curr social status of patients with affective, schizophrenic and
Pharm Des 2012;18:486–494. schizoaffective disorders after long-term course. Acta
90. Broyd SJ, Demanuele C, Debener S, Helps SK, James CJ, Psychiatr Scand 1990;82:352–358.
Sonuga-Barke EJ. Default-mode brain dysfunction in 100. Kendler KS, McGuire M, Gruenberg AM, Walsh D.
mental disorders: a systematic review. Neurosci Biobehav Examining the validity of DSM-III-R schizoaffective dis-
Rev 2009;33:279–296. order and its putative subtypes in the Roscommon Fam-
91. Gusnard DA, Raichle ME. Searching for a baseline: func- ily Study. Am J Psychiatry 1995;152:755–764.
tional imaging and the resting human brain. Nat Rev 101. Benabarre A, Vieta E, Colom F, Martinez-Aran A, Rein-
Neurosci 2001;2:685–694. ares M, Gasto C. Bipolar disorder, schizoaffective disor-
92. Raichle ME, Macleod AM, Snyder AZ, Powers WJ, Gus- der and schizophrenia: epidemiologic, clinical and
nard DA, Shulman GL. A default mode of brain function. prognostic differences. Eur Psychiat 2001;16:167–172.
Proc Natl Acad Sci U S A 2001;98:676–682. 102. van Os J, Kapur S. Schizophrenia. Lancet 2009;374:635–
93. Buckner RL, Andrews-Hanna JR, Schacter DL. The 645.
brain’s default network: anatomy, function, and rele- 103. LEHMAN H, editor. Schizophrenia: clinical features. Bal-
vance to disease. Ann N Y Acad Sci 2008;1124:1–38. timore: Williams & Wilkins; 1975.
94. Gusnard DA. Being a self: considerations from functional 104. Pagel T, Baldessarini RJ, Franklin J, Baethge C. Charac-
imaging. Conscious Cogn 2005;14:679–697. teristics of patients diagnosed with schizoaffective disor-
95. Clayton PJ, Rodin L, Winokur G. Family history studies. der compared with schizophrenia and bipolar disorder.
3. Schizoaffective disorder, clinical and genetic factors Bipolar Disord 2013;15:229–239.
including a one to two year follow-up. Compr Psychiatry 105. Olfson M, Marcus SC, Wan GJ. Treatment patterns for
1968;9:31–49. schizoaffective disorder and schizophrenia among Medi-
96. Dell’osso L, Akiskal HS, Freer P, Barberi M, Placidi caid patients. Psychiatr Serv 2009;60:210–216.
GF, Cassano GB. Psychotic and nonpsychotic bipolar 106. Schumann G, Binder EB, Holte A et al. Stratified medi-
mixed states: comparisons with manic and schizoaffective cine for mental disorders. Eur Neuropsychopharmacol
disorders. Eur Arch Psychiatry Clin Neurosci 2014;24:5–50.
1993;243:75–81. 107. Vieta E. Individualizing treatment for patients with
97. Laursen TM, Labouriau R, Licht RW, Bertelsen A, schizoaffective disorder. J Clin Psychiatry 2010;71:e26.
Munk-Olsen T, Mortensen PB. Family history of psychi-
atric illness as a risk factor for schizoaffective disorder: a
Danish register-based cohort study. Arch Gen Psychiatry Supporting Information
2005;62:841–848.
98. Cheniaux E, Landeira-Fernandez J, Lessa Telles L et al. Additional Supporting Information may be found in the online
Does schizoaffective disorder really exist? A systematic version of this article:
review of the studies that compared schizoaffective disor- Figure S1. Inclusion of neuropsychological and neuroimaging
der with schizophrenia or mood disorders J Affect Disord studies of schizoaffective subjects in this systematic review.
2008;106:209–217.

30