British Journal of Psychiatry (1992), 161, 335343

Are There More Than Two Syndromes in Schizophrenia?

A Critique

of the PositiveNegative Dichotomy

VICTOR PERALTA,JOSE de LEON and MANUEL J.CUESTA

A sample of 115 DSMlIlR
schizophrenics was studied by means of the SANS and SAPS.
A factor analysis from the nine subscalesand two symptoms (inappropriateaffect and poverty
of content) and a review of the previous factor analysessuggest that schizophrenic symptoms
cannot be appropriately classified into positive and negative syndromes. The low internal
consistency of the SAPS suggests that the positive symptoms are not a homogeneous
syndrome. Our results fit better with Liddle's model of three syndromes (negative, delusion
hallucination and disorganisation

syndromes). ft is argued that we are far from a valid classifica

tion of schizophrenic symptoms and the positive-negative dichotomy appears to be an

oversimplification.

1978). Thus, it can be used to explore whether

Since Crow (1980) formulated
his model of positive
and negative symptoms, this dichotomy has become the schizophrenic symptoms can be appropriately
widely accepted, despite the variability in definitions
summarised in a positive and negative syndrome.
Those symptoms considered to be negative should
of positive and negative symptoms by various
researchers, insufficient study of the psychometric
weigh heavily in a negative factor and those
properties of some scales, failure to account for considered positive should weigh in a positive factor.
cross-sectional confounding variables, and lack of Liddle (1987a), who conducted a factor analysis of
15 individual items included in the SANS and SAPS,
follow-up to confirm the irreversibility of negative
has proposed that the so-called positive symptoms
symptoms (Sommers, 1985; de Leon et al, 1989).
Andreasen's model of positivenegativeschizo
should be divided into at least two syndromes: a
phrenia is one of the hypothesised models which has delusion-hallucination syndrome and adisorganisation
syndrome.
been more thoroughly studied from a methodological
point of view. She developed two scales: the Scale
Two of the authors (VP and MJC) have made a
for the Assessment of Negative Symptoms (SANS) factor analysis of the SANS and SAPS in 115
(Andreasen, 1982) and the Scale for the Assessment
patients, published in Spanish (Peralta & Cuesta,
of Positive Symptoms (SAPS) (Andreasen, 1984). 1990). The first factor was loaded by negative
The SANS has 5 subscales which include 20 symptoms excluding attentional impairment, the
symptoms, and the SAPS has 4 subscales and 30 second was similar to the disorganisation syndrome
symptoms. Criteria have been proposed categorically (formal thought disturbances and attentional impair
to classify patients as exhibiting positive, mixed, or ment), the third was loaded with hallucinations and
negative syndromes (Andreasen & Olsen, 1982). The delusions, and a fourth consisted almost completely
SANS has good internal consistency (0.85 of bizarre behaviour.
Andreasen
& Olsen, 1982; Andreasen
et a!, 1990;
Following Liddle's model, we have reanalysed the
0.78 - our study), but contamination by akinesia and data of our factor analysis, taking into account
depression seems to be a problem (Dc Leon et a!, the following hypotheses. (a) Inappropriate affect
1989). The internal consistency of the SAPS appears
should not be considered a negative symptom as it
low (0.40- Andreasen&Olsen, 1982;0.48 - Andreasen was in the first version of the SANS, and if its rating
et al, 1990; 0.30
our study), suggestingthat the isextracted from theaffective flattening item, it should
positive syndrome may not be homogeneous. For merge into the disorganisation factor. (b) Poverty
our proposed scales, internal consistencies were: of content of speech should not be classified as
disorganisation syndrome scale 0.65, delusion
negative, and if its rating is extracted from the
hallucination syndrome scale 0.63, and modified
alogia subscale it should move to the disorganisation
negative syndrome scale 0.80. The numbers of factor. (c) We proposed to obtain from the SANS
subscales were three, two, and four, respectively.
SAPS conglomerate not two, but four scales. The
Factor analysis reduces a large number of four scales would include (i) the so-called negative
independent variables to a smaller, conceptually
symptom scale after extracting attentional impair
more coherent set of variables (Kim & Mueller,
ment, inappropriate affect and poverty of content,
335

336

PERALTA

ET AL

Table 1
Sociodemographic
and clinicalcharacteristicsof the sample
Variable

Mean (s.d.)

Sociodemographic
Age
Education:years
Sex

Range

Median

No. (%)

35.7
(12.3)1676349.0
(3.7)8-18810.2

male

79 (69)

36 (31)

female
Marital status
single
married
divorced
widowed
Social class

98 (85)

8 (7)
7 (6)
2 (2)

I, II, Ill
IV, V

54 (47)
61(53)

Clinical
Illness duration
Onset age: years

(8.2)0-46825.3
(8.4)13-61224.7

No.of admissions
Monthsof prioradmissions
GAFforlastyear
GAF atadmission
Chiorpromazineequivalents

Biperidone:
mg/day
NI treatment:years
Schizophreniatype DSM-lll-R
paranoid
undifferentiated
disorganised
residual
catatonic
Andreasensymptoms
positive
mixed
negative
Most frequentoral NL
haloperidol
clozapine

(4.5)1-27312.7

(42.2)0-384449.5
(13.0)18-855028.7
(7.9)1555301282
(987)100720010503.2
(2.2)0-844.2
(3.3)0274
63 (55)

19 (17)
19 (16)
8 (7)
6 (5)
33 (29)

61(53)
21(18)

78 (68)
10 (9)

tioproperazine
Long-acting
NI
fluphenazine

50 (43)

NI status prior to admission

never treated
continuous for last year
c@scontiruous
for last year

12 (10)
71(62)
32 (28)

6 (5)

NI neuroleptics.
Chlorpromazine
equivalents:
maximum
dosage
duringhospital
admission.

(ii) a scale for delusions and hallucination, (iii) the

disorganisation syndrome scale and (iv) a bizarre be
haviour scale. These new scales should have higher
internal consistency than the SANS and SAPS.
Our results are compared with six previous factor
analyses which have used SANS and SAPS.
Method
The sociodemographicand clinical characteristics of the

sample are described in Table 1. Patients were obtained

from 115consecutiveadmissionsto an acute in-patientunit
who were diagnosed as suffering from DSM-III-R
schizophreniaby VP. A semistructuredinterviewdesigned
for schizophrenic patients (Landmark, 1982) and the
AMDP (Guy & Ban, 1979)were used.
The SANSand SAPSratingsweredone by VP and MJC
during the first fivedays of admissionwith the last month
taken into account to score the symptoms. Inter-rater
reliabilityfor the subscalescoresranged from 0.82 to 0.93
(interciass

correlation

coefficient);

and for inappropriate

DCO@@0

HOW MANY SCHIZOPHRENIA

ol@2
@l@E

@O
<<r@c@

Ia
.@I@u)

U)C1)

SYNDROMES?

337

@j

000

O
>.

CCC

0
>.

2
.3
a

@0

@
Cl)

a
<<0)t@

.@c')@

Cl)@

a
>

lioo
>.CC

.@
I
cn

>

0
a

O)@

o
a -@
oa

I C')Oa00

cflCl)

>.

0@0

C>.C

00

>
0

0@c0

01

>>.E

(flU)
ZOc

0(0

:2

2!

Ca

a
Ec@)CD@
N-

(flU)

a@C

0>.

@
@
@
@

>

Ca

@00

>

>.CC

SE

00)

.00
0
C

0
C

aIC

U)E

o@I

@
@
@

.5
01:
u)I@O

,-j-r--o

(I)!

N-C

000

0>.
000
>>.>.

CCC

@IC2

a'
01

@I<0

E .c
I-a

j.

cl)X@Lo
@<,...T

@
@

a
E@0@

00

El
a@

N->.

:i

>>.>.

>

a'
0I_

ia

l@C

q@
zg@u,I

C
0

*
a

a
>

aaa

ooo
>.>.>.

>

a
0

a
.C
0

c0

@.

.2

a
C

0
.@

@>
o2

C*@C
.2o,..

o
C
@a

a. @

.@

C@

a@

.:
CCa

oa@

cnb

c@a:zi-z

@nN

obi@

:.c

.2

2
C,..
ao

a2'@E't
@0a
Oa

cn

.8

U)

ii
a

@-

@
C

@
@

C@

@@@

338

.@

-@-@

@.

-@

PERALTA ET AL

ml

vaa@a@@

t@

@t@tI@

.2.!

>->.>.>->....
-

C@

.2.2.!.!.!

ti

.2

00

00000

U)u)
00

ooooti
aaaaa

ti
a

CC

CCCCC

21
a'
,

a@

U)aU)aa@@
00000@@

@ICa

>>>>>.

>>.

L@ .@
0

a@

000*0
>.>.>.

00

00

@C

C.) 0000
a

@.>.
QQ

C@

CC

aa

.2.2.!.!
0000

CCCC

all)

-o

U)U)U)

EE

.Sr'.

EEE

0000

0000

00000
.2.2.!.!.!

U
.2

aaaa
0000
>->.

22222

00
000
>_>.

0
0

.@

.2

>>_>.

10
0*000
a...aaafl

0) @

@.

>->.

CC

.c
C')c

(-4

02000

2?

aO
l@

e
o@

00
>.

@a@0)@@

C
E <@

@u,

0000
a

aaaaa

00000

>.

0000

CCCCC

.!.!aa

CCCC

22

CC

C
C

(0

0
@

.0

0 C@

0
U
0

ooot

.2.2.2.2.!.!.!

.2.2

.2.2.!.!.!

.2

.2.2.2.2

0000000

00

00000

0000

aaaa@a

aa

a@aa

CCCCCCC

CC

CCCCC

C
0

CCCC

C
0

.5

@E

0000*02
>.>.>.>.@>.

00
>->.

*oo2
>.@>.>.,..

0>.

.2.2.!.!

C0

0000

a
2

CCCC

al

I
@
@

0l
Ia

@u@o
00

00

-titititi

>>.>>.
0

@C

OCCCC

0
C

C.)

.@Co

..5@
a

@
@

-2@

2@@-4-

@ti-@

@o

-@

-4-@I-@

-@..o

oa

._

--

.@E
0

i@

.8
--

0,>!a

@Q@<<

0
.2
Z

HOW MANY SCHIZOPHRENIA SYNDROMES?

affect and povertyof content it was, respectively,0.74and

affective flattening (0.57, P<0.00l);

0.85. All patients were taking neuroleptics. Most patients

(70%) were treated with an anticholinergic, biperiden hydro

P<0.00l).

chloride, to exclude, as far as possible, akinesia as a

confounder in the assessment of negative symptoms (de
Leon et a!, 1989). A further 9% of patients were treated
with clozapine,
symptoms.

which

induces

very few Parkinsonian

The Statistical Package for Social Sciences(Nie el a!,

factor

analysis

whose

rotation.

Only those

factors

eigenvalues were greater than I .0 were retained. Those

subscales or items that have a weight higher than 0.40 in
each factor were chosen to construct syndromes. The
internal consistency of these syndromes was calculated by

Cronbach's alpha.

Internal consistency of the scales

Cronbach's alpha is influenced by the number of items
included in the scale (Green et a!, 1977; HattIe, 1985). The
0.78, respectively. The new subscales based on the factor
analysis have greater internal consistency even with

fewer items. The delusionhallucinationsyndrome scale

(hallucinations

attentional

Factor analysis using inappropriate affect and poverty of

factor, the delusionhallucinationfactor and the bizarre

behaviour factor. Respectively, they explained 33%, 16lo,
12% and 9% of the variance, making a total of 70%.
Four of the five negative symptoms described by

Andreasen weighted on the negative factor (affective

flattening 0.83, alogia 0.75, avolition 0.75, and anhedoma
0.76). The disorganisation syndrome comprised formal
thought disturbance (0.80), inappropriate affect (0.68)
and attentional impairment (0.84). The delusion-hallucina
tion factor was only constituted by these two subscales

(0.84 and 0.85 respectively),and the bizarre behaviour

explained

almost

completely

by this

sub

scale (0.96).

Reclassification of some negative symptoms

Following Liddle's model, our hypotheses predicted that
attentional impairment, inappropriate affect and poverty

of contentare not part of the negativefactor. As predicted,

attentional impairment was related to the disorganisation

factor (Table3). It wassignificantlycorrelatedwithapathy

0.25,

P<0.01),

poverty

of content

(0.26,

P<0.0l),

inappropriate affect (0.36, P< 0.001), positive formal

thought disturbance (0.41, P< 0.001), and alogia (0.43,
P< 0.001). Inappropriate affect weighted heavily (0.68) on
the disorganisation factor (Table 3) and was correlated with
positive formal thought disturbance (0.40, P< 0.001) and
attentional impairment (0.36, P<0.00l). Poverty of content
did not follow our prediction

of being more related to the

disorganisation factor (Table 3); rather, the weight

in the negative factor (0.74) was clearly higher than
that

in the disorganisation

factor

(0.20).

Poverty

of

content wassignificantlycorrelated with severalsubscales

such as delusions

impairment

and inappropriate

affect) was 0.65.

The modified negative symptom scale, i.e. the SANS

excluding attentional impairment, had an internal
of 0.80, which was greater than that of the

SANS, in spite of having one item less.

content as independent items showed results similar to our

previous analysis and that of other authors (Tables 2 and
3), with four factors: the negative factor, the disorganisation

was

and delusions) had an internal consistency

of 0.63. The Cronbach's alpha of the disorganisation

syndrome scale (positive formal thought disturbance,

consistency

Results

factor

and alogia (0.63,

internalconsistenciesof the SAPSand SANSwere0.30and

1975) was used to carry out a principal-component

with varimax

339

(0.42, P<0.001);

positive

formal

thought disturbance (0.29, P<0.01); attention impair

ment (0.26, P<0.Ol); inappropriate affect (0.30, P<0.0l);
apathy (0.45, P<0.001); anhedonia (0.51, P<0.001);

Discussion

The principal limitation of factor analysis is the need

to have a large sample size. It has been suggested
that an absolute minimum ratio is five individuals per

variable and not fewer than 100 individuals for any

analysis (Gorsuch, 1983). The conglomerate SAPS
SANS has 9 subscales and 50 items. Thus, the sample

size for most of the previous work seems too small

to use even subscales scores (Table 2). Although all

authors used principal-component analysis, differences

in rotation techniques and numbers of factors
obtained should be kept in mind when comparing
articles (Kim & Mueller, 1978). Additionally, the
changing characteristics of schizophrenic symptoms
and differences in the time course of the illness could
confuse the results because of the cross-sectional

nature of factor analysis (Barnes & Liddle, 1990).

Before we compare our results with those previously
published, a review of their differences in method is
needed. The first analysis published by Andreasen
& Olsen (1982) probably included too few patients,
and a preliminary version of the SAPS that included
catatonic symptoms was used. Unfortunately, in the

second factor analysis the sample size was not

specified, although it was defined as a somewhat
larger sample' than the sample studied in 1982.
Although patients were diagnosed using DSM-III
(American Psychiatric Association, 1980), the criteria

are only slightly broader than the DSM-III-R criteria

(American Psychiatric Association, 1987)used by us.
An important problem in the interpretation of
Andreasen's results is that the four factors explained
only 47% of the variance, and in addition, positive
formal thought disorder did not carry significant
weight in any factor.
Bilder et a! (1985) carried out the first factor
analysis that showed positive symptoms weighed in

340

PERALTA

two factors. Their results are limited by the small

sample size. They used the Schedule for Affective
Disorders and Schizophrenia (SADS) (Endicott &
Spitzer, 1978) as a way of measuring positive
symptoms, and the Research Diagnostic Criteria
(RDC) (Spitzer et a!, 1978) for diagnosing
schizophrenia. Although the SAPS is based on the
SADS, and the DSM.-III on the RDC, their
differences

should

be taken

into

account.

Liddle

confirmed the results of Bilder; attention was not

related to other negative symptoms and was weighted
in the same factor as positive formal thought
disturbancesand inappropriateaffect, while delusions
and hallucinations comprised another factor. The
main difference with the rest of the analysis is that
Liddle used the Comprehensive Assessment of
Symptoms and History (CASH) (Andreasen, 1983),
which contains all the items included in the SAPS
and SANS without the artefact of the subscale scores.
He included 15 clinical items in the factor analysis
and made correlations between the three factors and
items which he considered to be psychosocial. These
latter items are included in the SANS within the
anhedonia, apathy and attentional impairment
subscales, and were correlated with the negative or
disorganisation factor. He did not include bizarre
symptoms. Despite their different methods, there is
a remarkablesimilarity between his results and those
of Bilder. However, Liddle's results have two main
caveats, the sample size and lack of information
about percentage of variance explained by the factors
The other two works described in Tables 2 and 3
are even more problematic. Moscarelli et a! (1985)
gave conflicting results about their sample size (59
and 23 patients), and the three factors oddly
accounted for 100lo
of the variance. Kulahara &
Chandiramani (1990) studied a bigger sample, 98
patients at admission and 79 at follow-up. In contrast
to our sample, they used ICD9as diagnostic
criteria, and only 55% of their patients met DSM
III criteria.

ET AL

The negative factor is the most consistent across

studies (Table 3). The only difference is the inclusion
or exclusion of attentional impairment. In Liddle's
results, as well as ours, this item has a very low
weight in the negative factor. In Bilder's study,
attentional

impairment

has a certain weight in this

factor, but it is clearly lower than in the factor related

to positive formal thought disorder (Table 3). In the
three other studies attentional impairment seems to
be a part of the negative factor. In Andreasen's
study, there are two negative factors, one which is
related to attention and another which is not. It must

be remembered that positive formal thought disorder

does not weigh in any factor in this study.
The delusionhallucinationfactor is also quite
persistent across studies (Table 3), suggesting that
it forms a rather homogeneous syndrome.
The number of factors extracted seem to influence
the results. When four factors are obtained, as in
our sample or Andreasen's sample, bizarre behaviour
alone makes a factor. When three factors are used,
bizarre behaviour is grouped with positive formal
thought disorder (Table 3).
Reclassification of some negative symptoms

In our sample, attentional impairment was related

to the disorganisation factor, and Bilder & Liddle's
results support this. The results of other studies,
although problematic,suggestthat attentionalimpair
ment could be a negative symptom. Attention is
probably a heterogeneous and multifactor concept,
and its classification seems difficult to resolve.
Walker et a! (1988), using the SANSSAPS,found
significant correlations with positive formal thought
disorder, alogia, anhedonia and apathy. When other
instruments were used to measure positive and
negative symptoms, attentional impairment was
associated with positive symptoms (Cornblatt et a!,
1985; Kay et a!, 1986). Most negative scales do not
include attention impairment as a negative symptom
(de Leon et a!, 1989).
Inappropriate affect was considered to be a
Comparison of factors across studies
positive symptom by Crow. Andreasen (1990) has
Despite these problems and differences, there are reversed her opinion, and now believes it to be a
certain consistencies across studies (Table 2): a factor positive symptom. When this symptom was extracted
weighted by most negative symptoms which we will from the affective flattening, we found it weighted
call the negative factor, a delusionalhallucination in the disorganisation factor, as Liddle suggested
factor (described

by Bilder & Liddle),

and two other

factors, the disorganisation factor (described by

Liddie) and the bizarre behaviour factor. In some
studies the last-named comprise a single factor, the
disorganisation-bizarre behaviour factor. In three
studies, factors made by the mix of previously
described factors were found.

(Table 3). Other studies using different

scales also

found a correlation with positive formal thought

disturbance (Gibbons et a!, 1985), and lack of
correlation

with

flattening

of affect

(Craig

et a!,

1985; Lindenmayer & Kay, 1987).

Our results do not support Liddle's claim that
poverty of content is part of the disorganisation

341

HOW MANYSCHIZOPHRENIASYNDROMES?
syndrome;

rather,

our

results,

like

Andreasen's,

classify it as part of the alogia subscale.

Proposal for a new classIfIcatIon
of schizophrenic symptoms
Rather than confining themselves to their original
model, Andreasen et a!(1990) have recently advocated
the need to explore other approaches, such as
changing thresholds or altering the configuration of
the symptoms required to classify a patient in a
particular category. The review of the factor analyses
does not support the dichotomy of positive and
negative syndromes. In fact, it supports the existence
of three or even four syndromes when the SAPS
SANS is used. This is also indicated

by the low

internal consistency of the SAPS.

We propose an alternative way of using the SANS
SAPS, by constructing four scales. The modified
negative syndrome scale includes all SANS subscales
except attentional impairment. Indeed, the suggested
changes lead to increased internal consistency in spite
of having one item less than the original SANS. The
disorganisation syndrome (positive formal thought
disturbance, attentional impairment and inappropriate
affect) and the delusionhallucination syndrome
scale clearly have higher internal consistency than
the SAPS, despite having fewer items. The bizarre
behaviour subscale could be used as an additional
scale. The internal consistency of this four-item
subscale was lower in our sample (0.45) than (0.79)
found by Andreasen

et a! (1990).

Validity of this new classification

Before reviewingthe small numberof data to support
Liddle's classification, we must consider the limited
data supporting Andreasen's dichotomy of positive
negative symptoms, and the division of schizophrenia
into positive, mixed and negative. The data obtained
by using other scales and models, such as those
described by Crow (1980) and Kay & Opler (1987),
should not be used to support Andreasen's model,
because they are substantially different from those
proposed by Andreasen (de Leon et a!, 1989).
The external data to validate her model, such as
structural brain damage and cognitive deficits
measured by the Mini-Mental State Examination, are
not without problems. (a) The model is designed to
explain differences between positive and negative
schizophrenia. Mixed schizophrenia remains a grey
area, although closer to the positive type. Mixed
schizophrenia is clearly the most frequent type in
large samples, such as ours or Kulahara's, and the
percentageappearsto increasewith the time course of

illness (Kulahara & Chandiramani, 1990). This has

recently led Andreasen et a! (1990) to change the
criteria to classify the mixed and negative syndromes.
0') Andreasen's static model does not explain the
frequent changes of patients from the negative to
the other syndrome and vice versa (Tantam et a!,
1989). (c) The varying stability of data used to
validate the model externally is a problem.
Ventricular dilatation is thought to be irreversible
and its association with the negative syndrome was
not confirmed in a more recent sample (Andreasen
et a!, 1990). The results of the Mini-Mental State are
not very stable in some schizophrenic patients.
Liddle (1987b) has provided some neuropsych
ological data to support the validity of the three
syndromes in his 40 patients. The negative factor was
associated with poor performance in tests of
conceptual thinking, object naming and long-term
memory, and cortical neurological signs. The
disorgamsation factor was associated with poor
performance

in tests of concentration,

immediate

recall and word learning, and with cortical neur

ological signs. The delusion-hallucination factor was
associated with poor figure perception when the
education influence was controlled. Liddle suggested
a dysfunction at two different sites within the frontal
lobe (dorsolateral and orbital prefrontal areas) for
the first two factors and a possible impairment of
the temporal lobe for the last-named. Andreasen et
a! (1990) have also recently suggested that negative
symptoms could be related to frontal lobe abnor
malities, and positive symptoms to temporo-limbic
abnormalities.
The classical classification of schizophrenic
types also gives some support to Liddle's model.
Hebephrenic patients are characterised by the
presence of the disorganisation syndrome. The
delusionhallucination syndrome is prominent in
the paranoid type. The modified negative syndrome
(Liddle's psychomotor retardation syndrome) is
characteristic of residual schizophrenia. Catatonic
patients would show the modified negative syndrome
and catatonic symptoms, which are not included in
the SANS-SAPS. The possibilityof patientschanging
type

and

having

mixed

types

suggests

that

these

syndromes are independent and could reflect distur

bances in different areas of the brain.
Conclusion
While this article proposes criteria to classify patients
with the modified negative, delusionhallucination
and disorganisation syndromes, we have not
operationalised the criteria because we are uncertain
of their validity. In our sample, Liddle's model

342

PERALTA

showed more robust psychometric properties than

Andreasen's

dichotomic model, but we do not know

if other models would have even better psychometric

properties. Repeat factor analysis using items rather
than subscales and the replication of the results in
other samples are in progress.
The construction of a valid model of schizophrenic
symptoms is certainly more difficult than simplistic
attempts to validate the positivenegativemodel. It
requires larger samples, longer follow-ups and an
attempt to associate the changes of these hypothetical
syndromes with changes in the abnormalities found
in different

areas of the brain, measured

by brain

imaging (single photon emission computerised tomo

graphy (SPECT) or positron emission tomography
(PET)) or neuropsychological tests.
The complexity of this task is further complicated
since not all symptoms found in schizophrenic
patients are produced by schizophrenia. For
example, the negative syndrome in a cross-section
can be confused with akinesia or depression, and
even if it is stable for a long period of time, the
possibility of a previous schizoid personality disorder
or institutionalisation should be considered (Sommers,
1985; de Leon eta!, 1989). There has been very little
research on those secondary'
negative syndromes.
When studied, the SANS was shown to be influenced
by akinesia (Hoffman et a!, 1987); there are also
other indirect results that suggest the same associ
ation (Mayer et a!, 1985; Walker et a!, 1988). The
contamination of the SANS by depression in
schizophrenic patients has also received little attention.
According to Kulahara et a! (1989), negative
symptoms can be distinguished from depression.
Even if other concurrent problems are excluded,
as Bleuler (1950) suggested, primary and secondary
schizophrenic symptoms are likely. The former are
probably associated with brain disturbances, and the
latter are probably not. Another caveat difficult to
resolve is the accuracy of our assessment of
symptoms. For example, if Liddle's syndromes prove
to be valid, it would be difficult to assess with
certainty the presence and severity of the delusion
hallucination syndrome in patients with severe
disorganisation. Similarly, in patients with severe
psychomotor retardation syndrome, poverty of
speech could preclude the expression of the
disorganisation and delusion-hallucination
syn
dromes. Thus, we are some way from a valid
classification of schizophrenic symptoms.

ET AL
Pennsylvania, Eastern Pennsylvania Psychiatric Institute,
Philadelphia, USA, for their critiques and suggestions.

References
AMERiCAN PSYGIIATRIC Ansocwnoia

(1980) Diagnostic

and Stat&ku!

ManualofMenfalDisorders(3rd
ada) (DSM-III). Washington,
DC: APA.
(1987) Diagnostic and Statistical Manual of Mental
L@wideir(3rdedn, revised)(DSM-ItI-R). Washington, DC: APA.
ANDREASEN, N. C. (1982)

Negative

symptoms

in schizophrenia:

definition and reliability. Archives of General Psychiatry, 39,

784788.

(1983)

Comprehensive

Assessment

of

Symptoms

and

Hirtory (CASH). Iowa: University oflowa College of Medicine.

(1984).

Scale

for

the

Azressment

of

Positive

Symptoms

(SAPS). Iowa City, IA: University of Iowa.

(1990)

Methods

for

assessing

positive

and

negative

symptoms. Modern Pivblemsin Pharmacopsychiatry, 24, 73-88.

&

OI@sEN,

S.

(1982)

Negative

and

positive

schizophrenia:

definition and validation. Archives of General Psychiatry, 39,

789794.
,

Fi@ut@i,

M.

&

SWAYZE,

V.

W.

(1990)

Positive

and

negative

symptoms in schizophrenia. Archives ofGenera!Psychiat,y,

615621.
B@iu@ias,1. R. E. & Lmoia,

P. F. (1990) Evidence for the validity

of negative symptoms. Modern Problems in Pharmaco

psychiatry, 24, 43-72.
Bnjea, R. M., MUICHERGEE,
S. & RIEDER,R. 0. (1985)
Symptomatic
and neuropsychological
components
state. Schizophrenia Bulletin, 11, 409419.
BUtULER,

E.

(1950)

Dementia

Praecox

and

the

of defect
Group

information processing.SchizophreniaBulletin, 11, 397-408.

CRAIG, 1. J., R1cH4'uwsoN, M. A. & PASs, R. (1985) Measurement

of mood and affect in schizophrenic inpatients. American

Journal of Psychiatry, 142, 1272-1277.
CROW, T. J. (1980)

Molecular

pathology

of schizophrenia:

more

than a disease process? British Medical Journal, 280, 66-68.

Da

LEON, J.,

WILSON, W.

H.

& SIMPSON, 0.

M.

(1989)

Measurement of negative symptoms in schizophrenia. Psychiatric

Developments,7, 211234.
E@icorr,

J. & SPITZER, R. L. (1978) A diagnostic

interview:

the

Schedule for Affective Disorders and Schizophrenia. Archives

of GeneralPsychiatry,35, 837866.
GIBBONS, R. D., LEWINE, R. R. J. & DAVIS, 3. M. (1985) An

empirical test of a Kraepelinian versus a Bleulerian view of

negative symptoms. Schizophrenia Bulletin, 11, 390-396.
GoasucH, R. L. (1983) Factor Analysis. Hillsdale, NJ: Lawrence
Erlbaum.
GREEN, S. B., LIssrrz, R. W. & Mu@ix,

S. A. (1977) Limitations

of coefficient alpha as an index of test unidimensionality.

Education and Psychological Measurement, 37, 827838.
Guy, W. & B*ri, T. A. (1979) The AMDP-System. New York:
Springer-Verlag.
HA1ThE, 3. (1985) Methodology

review:

assessing

unidimensionality

of test and items.AppliedPsychologicalMecnw'ement,9,139-164.

Hom@w@,W. F., L@BS,S. M. & CASEY,0. E. (1987) Neuroleptic
induced

Parkinsonism

in older

schizophrenics.

Biological

Psychiatry, 22, 427439.

K,@y,S. R., OPLER,L. A. & FlSzaar@4,
A. (1986) Significance of
British

Journal of Psychiatry, 149, 439-448.

The authors are grateful to George M. Simpson, MD, Richard C.

Josiassen, PhD, and Kim Mueser, PhD, Medical College of

of

Schizophrenias (trans. J. Zinkin). New York: International

University Press.
CoaNat@rr, B. A., LENZENWEGER,
M. F. & DwoRiull, R. H. (1985)
Positive and negative schizophrenic symptoms, attention and

positive and negative syndromes in chronic schizophrenia.

Acknowladg.ments

47,

&

schizophrenia:

(1987)

The

its

validity

positivenegative

Developments,5, 79103.

and

significance.

dimension

in

Psychiatric

HOWMANYSCHIZOPHRENIASYNDROMES?
Kmi, J. & Muauiit, C. W. (1978)Introduction to Factor Analysis.
Beverly Hills, CA: Sage Publications.
KULAHARA, P., AvASmI, A., CHANDIRAMANI,K., et al (1989)

Negative and depressive symptoms in schizophrenia. British

Journal of Psychiatry, 154, 207211.

&

CHANDIRAMANI,

K.

(1990)

Positive

and

negative

LANDMARK,1. (1982) A manual

for the assessment

of schizo

phrenia. Acta Psychiatrica Scandinavica,65 (suppi. 298).

LIDDLE, P. F. (1987a) The symptoms

(1987b)

Schizophrenic

syndromes,

dichotomy. British

cognitive

LINDENMAYER,J. P. & KAY, S. R. (1987) Affective impairment

in

young acute schizophrenics: its structure, course and prognostic

significance. Acta Psychiatrica Scandinavica, 75, 287-296.
P., et al (1985) Multiple

contributions of clinical presentation of fiat affect in

schizophrenia. Schizophrenia Bulletin, 11, 420-426.
MOSCARELLI,M., MArvEl, L., CESANA, B. M., et al (1987) An

international perspectiveon assessment of negative and positive

(1975) Statistical

Package for the Social Sciences (2nd edn). New York:

McGraw-Hill.
V. & CUESTA ZORITA,

M.

J. (1990)

Sintomas

esquizofrenicos positivos y negativos: un reanalisis del modelo

dicotomico de Ia esquizofrenia. Actas Luso-Erpanolas de
Neurologia y Psiquiatria, 18, 266-273.
methodological
366380.

problems.

SPITZER, R. L., ENDIcOrr,

performance

and neurological dysfunction. Psychological Medicine, 17,

4957-

MAYER, M., ALPERT, M., SmslNY,

NIE, N. H., HULL, N. C. & JENKINS,J. 0.

SOMMERS, A. A. (1985) Negative

of chronic schizophrenia:

reexamination of the positive-negative

Journal of Psychiatry, 151, 145151.

symptoms in schizophrenia. American Journal of Psychiatry,

144, 15951598.

PERALTA MARTIN,

subtypes

of schizophrenia, a follow-up study from India. Schizophrenia

Research, 3, 107-116.

343

symptoms:

Schizophrenia

conceptual

Bulletin,

and

11,

J. & ROBINS, E. (1978) Research

Diagnostic Criteria (RDC) for a SelectedGroup of Functional

Disorders (3rd edn). New York: New York State Psychiatric
Institute.
TANTAM, D., MONTAGUE, L., RING, N., et al (1988) Negative

symptoms in the first episode, and their subsequent evolution.

Abstracts of Proceedings of Meetings of the Royal College of
Psychiatrists, Psychiatric Bulletin (suppl. 1), 44.
WALKER,

E.

F.,

positive/negative

HARvEY,

P.

D.

& PERLMAN,

D.

(1988)

The

symptom distinction in psychoses. A replication

and extension of previous findings. Journal of Nervous and

Mental Diseases, 176, 359363.

Victor Peralta, MD; Manuel J. Cuesta, MD, Psychiatric Unit of the Virgen
del CaminoHospital,
Pamplona, Spain; 5Jose de Leon, MD, Medical College of Pennsylvania/EPPI, 2300 Henry Avenue,
Philadelphia, PA 19129, USA
aCorvespondence

Are there more than two syndromes in schizophrenia? A critique of

the positive-negative dichotomy.
V Peralta, J de Leon and M J Cuesta
BJP 1992, 161:335-343.
Access the most recent version at DOI: 10.1192/bjp.161.3.335

References
Reprints/
permissions
You can respond
to this article at
Downloaded
from

This article cites 0 articles, 0 of which you can access for free at:
http://bjp.rcpsych.org/content/161/3/335#BIBL
To obtain reprints or permission to reproduce material from this paper, please write
to permissions@rcpsych.ac.uk
/letters/submit/bjprcpsych;161/3/335
http://bjp.rcpsych.org/ on October 1, 2015
Published by The Royal College of Psychiatrists

To subscribe to The British Journal of Psychiatry go to:

http://bjp.rcpsych.org/site/subscriptions/