Psychopathological Syndromes of Schizophrenia:
Evaluation of the Dimensional Structure
of the Positive and Negative Syndrome Scale
by Reinhard Mass, Thomas Schoemig, Klaus Hitschfeld,
Eleonora Wall, and Christian Haasen
and Kay 1975), the PANSS consists of 7 positive symptom items, 7 negative symptom items, and 16 general psychopathology items. All 30 PANSS items are rated on a 7point symptom severity scale, ranking from 1 (absent) to
7 (extremely severe). The PANSS provides short symptom descriptions to anchor the different degrees of severity for each symptom.
However, further investigations (e.g., Bilder et al.
1985; Liddle 1987; Andreasen et al. 1995) suggested the
positive-negative dichotomy to be an oversimplification
of schizophrenic psychopathology and gave evidence for
the existence of a third core syndrome, which is featured
by symptoms such as formal thought disorder and inappropriate affect (Liddle 1987; Liddle and Barnes 1990).
This syndrome is called the disorganized syndrome.
Accordingly, several factor analyses conducted with
PANSS data showed that the negative and positive syndromes were not sufficient to explain schizophrenic psychopathology (Kay and Sevy 1990; Kay and Sandyk
1991; Bell et al. 1994a; Lindenmayer et al. 1994; Cuesta
and Peralta 1995; von Knorring and Lindstrom 1995).
Though the important symptom of inappropriate affect is
not considered, the disorganized syndrome seems to be
represented in the PANSS as an independent cognitive
component (Bell et al. 1994ft). Moreover, depression and
excitement components also are usually extracted. Since
measurement of depression has been shown to overlap
with that of negative symptoms (Prosser et al. 1987), the
possibility of distinguishing between depressive and negative symptoms in people with schizophrenia is a special
advantage of the PANSS.
Application of the PANSS depends to some extent on
observation of the patient's behavior; from such observation the occurence of psychopathological symptoms is
inferred, and as a rule these symptoms are associated with
abnormalities of behavior. However, the validity of such
Abstract
The Positive and Negative Syndrome Scale (PANSS)
was originally designed as a rating system that provides
balanced representation of positive and negative symptom features. Evidence from recent factor-analytic studies suggests that a five-dimensional solution appears to
best fit the psychopathological data as assessed with the
PANSS. To investigate the dimensional structure, we
administered the PANSS to 253 inpatients with schizophrenia. In accordance with former studies, principal
components analyses yielded five orthogonal dimensions: hostile excitement; negative, cognitive, and positive syndrome; and depression. When compared with
questionnaires measuring subjective nonpsychotic experiences of schizophrenia, paranoid mood, and depression, the correlation pattern verifies the PANSS components. In addition, we investigated a subsample of 70
male patients with a Continuous Performance Test
(CPT), a Span of Apprehension Task, and a Modality
Shift Effect (MSE) paradigm; the CPT was significantly
associated with the cognitive syndrome, and the MSE
correlated with the negative syndrome.
Keywords: PANSS, psychopathology, factor analysis, evaluation, self-rating, neuropsychology.
Schizophrenia Bulletin, 26(1):167-177, 2000.
The Positive and Negative Syndrome Scale (PANSS; Kay
et al. 1987) is one of the most widely used methods for
standardized measurement of schizophrenic core symptoms. The theoretical frame for the construction of the
PANSS was the assumption of two distinct subtypes of
schizophrenia (Crow 1980; Andreasen and Olsen 1982).
Type I is characterized by florid (positive) symptoms such
as delusions or hallucinations, and type II is composed of
deficit (negative) symptoms such as blunted affect or
social withdrawal. Based originally on the Brief
Psychiatric Rating Scale (BPRS; Overall and Gorham
1962) and the Psychopathology Rating Schedule (Singh
Reprint requests should be sent to Dr. Reinhard Mass, University
Hospital Eppendorf, Psychiatric Clinic, Martinistr. 52, 20246 Hamburg,
Germany; e-mail: [email protected].
167
Schizophrenia Bulletin, Vol. 26, No. 1, 2000
R. Mass et al.
inferences is not always certain because a variety of different mental states might be associated with a particular
abnormality of observed behavior. Hence, for evaluation
of the validity of the PANSS assessments it is necessary
to take into account additional sources of information—
for example, subjective questionnaire data and objective
neuropsychological measures. Consideration of the pattern of relationships between PANSS ratings and both
subjective and neuropsychological measures can help verify the data obtained with the PANSS.
The objective of the present study is to evaluate the
dimensions underlying schizophrenic psychopathology as
measured with the PANSS by means of exploratory factor
analysis. For further evaluations, the resulting PANSS
components are compared with clinical questionnaires
(aimed at subjective nonpsychotic experiences of schizophrenia, paranoid mood, and depression) and with neuropsychological measures that are known to be associated
with schizophrenia.
Methods
Subjects. A total of 253 psychiatric inpatients with
ICD-10 (World Health Organization 1992) diagnoses of
schizophrenia were recruited from the Northern Clinic,
Hamburg-Ochsenzoll, and from the University Hospital,
Hamburg-Eppendorf. Table 1 presents background sample
characteristics. Diagnostic criteria according to ICD-10
were checked with the International Diagnosis Check
Lists (IDCL; Hiller et al. 1995). Patients with organic
brain disorder (e.g., dementia) or with severe somatic diseases were not included. Of the total patients, 239
(94.5%) were taking neuroleptic drugs (typical, atypical,
or both), 119 (47.0%) received benzodiazepines, 32
(12.6%) received biperiden, and 49 (19.4%) were taking
other medication (e.g., sleeping drugs, antidepressants, or
anti-epileptic medication).
The surplus of men in our study is due to the fact that
a substantial part of the sample was recruited in a closed
psychiatric station for males (Ward 38A of the Northern
Table 1. Sample characteristics (n = 253)
35.2(10.6)
27.6 (8.0)
Age, yrs (SD)
Age of onset, yrs 1 (SD)
Duration of illness, yrs 2 (SD)
Number of hospitalizations, median (IQR)
Gender, n (%)
Male
Female
7.6 (8.6)
3(7)
214(84.6)
39(15.4)
Marital status, n (%)
Unmarried
Married
Separated
Divorced
Widowed
197(77.9)
29(11.5)
1 (0.4)
24 (9.5)
2 (0.8)
Educational level, n (%)
No school degree
Elementary school
Middle school
Secondary school 3
21 (8.3)
96 (37.9)
64 (25.3)
72 (28.5)
ICD-10 diagnoses, n (%)
Paranoid schizophrenia (code F20.0)
Hebephrenic schizophrenia (F20.1)
Catatonic schizophrenia (F20.2)
Undifferentiated schizophrenia (F20.3)
Residual schizophrenia (F20.5)
Other schizophrenia spectrum disorders
212(83.8)
13(5.1)
2 (0.8)
6 (2.4)
16(6.3)
4(1.6)
Note.—IQR = interquartile range; SD = standard deviation.
1
2
3
Age at first hospitalization.
Age minus age at first hospitalization.
Qualifying for entrance to university.
168
Psychopathological Syndromes
Schizophrenia Bulletin, Vol. 26, No. 1, 2000
Clinic Ochsenzoll), where the equipment for neuropsychological testing also has been installed.
Psychopathological Assessment. PANSS ratings were
completed for all patients by clinically trained staff (usually one of the attending psychiatrists: T.S., K.H., E.W.)
after semistructured interviews of about 30-60 minutes'
duration. Unfortunately, the limited personnel resources
of our project made it impossible to conduct the PANSS
interviews by two investigators simultaneously to allow
the calculation of interrater reliability coefficients.
However, all investigators were skilled and familiar with
the PANSS system and experienced in the assessment of
schizophrenic symptoms. Owing to the detailed definitions of the seven different levels of symptom severity in
former analyses (Bell et al. 1992), good-to-excellent reliabilities were reported for most of the items.
The PANSS interviews were conducted about 1
month after the admission (median = 31 days; quartile 1 =
15 days; quartile 3 = 62 days).
Questionnaires. All patients filled out the Paranoid
Depression Scale (PDS; von Zerssen 1976), a clinical questionnaire that is well established in German-speaking countries. It consists of two 16-item subscales, PDS-P (Paranoid
Mood) and PDS-D (Depressive Mood). The PDS includes
an 8-item Illness Denying subscale, which is a validity scale
that focuses on tendencies to deny ordinary symptoms of
indisposition (e.g., chill, headache, upset stomach). All PDS
items had to be answered on a 4-step scale (ranking from 0,
"applies not at all," to 3, "fully applicable").
In addition, the 8-item subscale FCQ-S (Mass et al.
1997) of the Frankfurt Complaint Questionnaire (FCQ;
Siillwold 1991) was employed. FCQ-S represents nonpsychotic subjective experiences ("basic symptoms") which
seem to be diagnostically characteristic for people with
schizophrenia (Mass et al. 1997, 1998) when compared with
people with alcoholism or obsessive-compulsive disorder.
Typical items of FCQ-S are "I am reluctant to read, because
I have so much trouble to grasp the meaning correctly";
"People's faces sometimes look unusual, almost distorted or
displaced to me"; and "If someone speaks with long sentences, it is particularly difficult for me to grasp the meaning." The FCQ-S items had to be answered dichotomously
with "yes" or "no." The reliability coefficient Cronbach a of
this experimental scale amounted to 0.69, indicating a satisfactory internal consistency for a short 8-item scale.
Continuous Performance. The CPT is usually considered
to be a test of sustained attention/vigilance. According to
more recent conceptions (Nuechterlein et al. 1994), CPT
variants using stimuli that are degraded through blurring may
particularly stress early perceptual stages of information processing. The CPT version used in our study agrees largely
with the recommendations of Nuechterlein et al. (1991). In a
computerized test 480 blurred digits were presented successively on a monitor, each stimulus was presented for 42 milliseconds with an interstimulus interval of 1 second. The task
was to detect the digit 0 (25% of all stimuli, randomly distributed) and to respond to it by pressing the space bar of the
keyboard. As a performance measure, the index of sensitivity
d' (Green and Swets 1966) was calculated, summarizing the
patient's discrimination of targets (the digit 0) from nontargets (other digits). The test duration was 8 minutes.
Span of Apprehension. The SAT refers to the number of
items (letters) that can be attended to at one time.
Granholm et al. (1996) discussed several hypotheses
regarding cognitive processes tapped by the SAT; especially slowed initiation of scanning and slowed encoding
operations are supposed to explain the SAT performance
deficits of people with schizophrenia. The computerized
SAT corresponds to the recommendations of Asarnow et
al. (1991). In a randomized order, either 3 (16 stimuli) or
8 (80 stimuli) consonant letters falling in a 4 x 4 matrix
were displayed (presentation time = 1 0 0 milliseconds).
Every display contained either the letter F or the letter T,
but never both. The patients were instructed to press the
left cursor key when detecting F and the right cursor key
when detecting T, and to guess (right or left cursor key)
when in doubt. The visual angle of the displays was relatively wide (about 15° x 15°). The performance measures
are the numbers of correct reactions during the 3- or 8-letter condition, respectively. Since the target letters are
quite easy to detect in the 3-letter displays, these displays
served as a control of the patient's motivation to perform
the task: Patients who performed at random or less in the
3-letter condition were excluded from further analysis.
The test duration was also about 8 minutes.
Modality Shift. Simple reaction times (RTs) were found
to be longer when the imperative stimulus reflected a shift
in modality than when a stimulus was repeated. Several
studies showed this difference to be more pronounced in
subjects with schizophrenia than in normal subjects. A
possible explanation is that people with schizophrenia
may tend to process stimuli in a more concrete code than
normals with different pathways for lights and sounds
(Cohen and Rist 1992).
To obtain an MSE measure, 128 visual (red light)
or auditive (signal tone) stimuli were presented in a ran-
Neuropsychological Testing. A battery of three neuropsychological tests was used, consisting of a degraded
stimulus Continuous Performance Test (CPT), a forcedchoice Span of Apprehension Task (SAT), and a Modality
Shift Effect (MSE) paradigm.
169
Schizophrenia Bulletin, Vol. 26, No. 1, 2000
R. Mass et al.
5.0 years (SD = 7.4). The subsample's age (3.2 years
less than the sample) and duration of illness (3.6 years
less) differed significantly from the remaining sample.
domized order (presentation time = 1 second; interstimulus interval =1.5 seconds) using a Wiener Reaction
Device. Patients had to press a button as fast as possible
when a stimulus occured. Since most studies comparing
subjects with schizophrenia and normal controls
reported a differential MSE only for RTs to tones (Rist
and Cohen 1991), the MSE in our study was calculated
for the tone condition only. The performance measure is
the difference between the medians of the cross-modal
RTs (tone after light, 32 stimuli) and the ipsimodal RTs
(tone after tone, 32 stimuli). The test duration was 5
minutes.
In a 1991 review, Nuechterlein et al. reported subjects with schizophrenia to have significant deficits on
degraded-stimulus CPT and forced-choice SAT during
both clinical remission and acute psychotic states; these
deficits did not increase with psychotic state and appear
to fit the definition of stable vulnerability indicators
(Nuechterlein and Dawson 1984).
Since the neuropsychological test devices were
available only during the final period of the investigation, neuropsychological data were obtained from 71
male patients only. One of these patients reached only 7
hits in the SAT 3-letter condition, thus indicating an
inadequate performance during the test session (possibly because of lack of motivation), and he was therefore
excluded. The resulting subsample (n = 70) had a mean
age of 32.8 years (SD = 9.8), a mean age of onset of
27.9 years (SD = 6.8), and a mean duration of illness of
Statistical Analyses
All analyses were conducted with the Statistical Package for
the Social Sciences, Release 4.0.5, for Macintosh (SPSS
1990). To evaluate the dimensional structure of the PANSS
data, exploratory principal components analysis (PCA) with
orthogonal varimax rotation and application of the eigenvalue criterion were carried out. PANSS factors, questionnaire scores, and MSE data showed a moderately skewed
distribution; CPT and SAT were normally distributed. There
were no statistical outliers. Spearman correlations calculated
for control purposes instead of the Pearson correlations
(tables 2 and 3) led to nearly identical results.
Results
To begin with, distribution characteristics for all PANSS
items were calculated. As table 4 shows, disorientation
(G10) was a very uncommon symptom within the present
sample. More than 96 percent of all patients were rated
with 1 ("absent") or 2 ("minimal"). Hence, the distribution
of this item is extremely skewed (kurtosis = 19.6; skewness = 4.1), and it was excluded in subsequent analyses.
Table 3. Pearson correlations (rxy) between the
five PANSS syndromes and neuropsychological
performance (n = 70)
Table 2. Pearson correlations ( r x j between the
five PANSS syndromes and questionnaire
scales (n = 253)
PDS-P1
PDS-D2
FCQ-S3
-0.02
EXC
0.37 " *
0.07
NEG
0.25 " *
0.19"
COG
0.37 * "
POS
DEP
CPT (d1)
SAT (hits)1
MSE (tone)
EXC
-0.22 (*)
-0.15
0.17
0.18"
NEG
-0.10
-0.15
0.31 **
0.17*
0.22 * "
COG
-0.29*
-0.21 (*)
0.22 (*)
0.48 ***
0.10
0.15*
POS
-0.14
-0.07
-0.08
0.15*
0.35 ***
0.10
DEP
0.05
-0.07
-0.23 (*)
Note.—COG = cognitive syndrome; CPT = Continuous
Performance Test; DEP = depression; EXC = hostile excitement;
MSE = Modality Shift Effect; NEG = negative syndrome; POS =
positive syndrome; SAT = Span of Apprehension Task.
1
8-letter condition.
Note.—COG = cognitive syndrome; DEP = depression; EXC =
hostile excitement; NEG = negative syndrome; POS = positive
syndrome.
1
PDS subscale "paranoid mood."
2
PDS subscale "depressive mood."
3
FCQ subscale "schizophrenia-specific subjective experiences."
(*)p < 0.1; *p < 0.05; **p < 0.01; two-tailed.
*p < 0.05; **p < 0.01; * " p < 0.001; two-tailed.
170
Psychopathological Syndromes
Schizophrenia Bulletin, Vol. 26, No. 1, 2000
Table 4. Distribution characteristics of the PANSS items (n = 253)
Item
Mean (SD)
P1, Delusions
P2, Conceptual disorganization
P3, Hallucinatory behavior
P4, Excitement
P5, Grandiosity
P6, Suspiciousness/persecution
P7, Hostility
N1, Blunted affect
N2, Emotional withdrawal
N3, Poor rapport
N4, Passive social withdrawal
N5, Difficulty in abstract thinking
N6, Lack of spontaneity
N7, Stereotyped thinking
G1, Somatic concern
G2, Anxiety
G3, Guilt feelings
G4, Tension
G5, Mannerisms and posturing
G6, Depression
G7, Motor Retardation
G8, Uncooperativeness
G9, Unusual thought content
G10, Disorientation
G11, Poor attention
G12, Lack of judgment and insight
G13, Disturbance of volition
G14, Poor impulse control
G15, Preoccupation
G16, Active social avoidance
2.89 (1.48)
3.06(1.40)
2.30 (1.42)
2.48 (1.57)
2.02(1.42)
2.81 (1.48)
1.89(1.22)
3.14(1.37)
3.28(1.49)
3.06 (1.48)
3.09(1.46)
3.14(1.60)
2.80 (1.43)
3.26(1.73)
2.42(1.55)
2.84(1.40)
2.21 (1.40)
2.63(1.42)
2.46(1.59)
2.53(1.28)
2.43(1.28)
2.04 (1.49)
3.15(1.71)
1.16(0.53)
2.77(1.17)
3.09(1.43)
2.80(1.41) '
2.18(1.52)
3.34(1.61)
2.80(1.43)
Range
1-7
1-7
1-6
1-7
1-7
1-6
1-6
1-6
1-7
1-6
1-7
1-7
1-6
1-7
1-7
1-7
1-6
1-6
1-6
1-6
1-6
1-7
1-7
1-5
1-6
1-7
1-7
1-7
1-7
1-7
Note.—PANSS = Positive and Negative Syndrome Scale.
• Factor 1, hostile excitement (EXC): P7, P4, G14, G4,
G8, P5, G5.
• Factor 2, negative syndrome (NEG): N4, N2, Nl, N3,
G16, N6.
• Factor 3, cognitive syndrome (COG): N5, Gl 1, P2.
• Factor 4, positive syndrome (POS): PI, P3, G9.
• Factor 5, depression (DEP): G6, G2, G3.
PC A extracted five components with eigenvalues > 1,
representing altogether 68.9 percent of the variance.
However, three items reached only small communalities,
thus being hardly represented by a five-factor model: P6,
suspiciousness/persecution (h2 [communality] = 0.481),
Gl, somatic concern (h2 = 0.516), and G7, motor retardation (h2 = 0.512). Therefore, to avoid misleading results,
in the next step these items were excluded. PCA again
extracted five factors with eigenvalues less than 1, now
representing 72.3 percent of the variance; communalities
of all items were greater than 0.6.
When considering only loadings greater than or equal
to 0.55 as substantial (corresponding to a common variance between factor and variable of 30%), the following
factor solution results (see also table 5):
Cronbach a coefficients for the factors are 0.91
(EXC), 0.92 (NEG), 0.81 (COG), 0.76 (POS), and 0.69
(DEP). It is important to keep in mind that syndromes as
represented by the PANSS components are not subtypes:
Although they are statistically independent, they can (and
usually do) overlap within given patients. To evaluate the
subtype approach in the present sample, we parted the
171
Schizophrenia Bulletin, Vol. 26, No. 1, 2000
R. Mass et al.
Table 5. Final PANSS factor solution (edged: all loadings > 0.55)
Variable
P1
P2
P3
P4
P5
P6
P7
N1
N2
N3
N4
N5
N6
N7 1
G1
G2
G3
G4
G5
G6
G7
G8
G9
G10
G11
G12 1
G13 1
G14
G15 1
G16
1
h2
0.79
0.72
0.60
0.82
0.63
F1
(42.8%)
0.06
0.33
0.10
0.85
0.64
0.80
0.74
0.80
0.80
0.77
0.72
0.72
0.78
0.87
0.19
0.24
0.34
-0.01
0.15
-0.01
0.48
0.69
0.66
0.74
0.63
0.75
0.13
0.19
0.75
0.61
-O.08
0.69
0.77
0.74
0.52
0.61
0.63
0.74
0.81
0.67
0.73
0.28
0.48
0.52
0.83
0.47
0.15
F2
F3
(7.7%)
(11.4%)
0.04
0.11
I 0.58 I
0.38
0.33
-0.01
0.24
-0.03
-0.18
-0.01
(item excluded, see text)
0.17
0.06
0.23
0.80
0.12
0.85
0.79
0.23
0.85
0.05
| 0.66 |
0.51
0.71
0.46
0.50
0.50
(item excluded, see text)
0.12
0.33
0.09
-0.11
0.29
0.06
0.17
0.44
0.12
-0.12
(tern excluded, see text)
0.10
0.36
0.24
0.14
(tern excluded, see text)
I 0.65 |
0.30
0.47
0.25
0.48
0.46
0.21
0.23
0.40
0.46
0.12
0.79
F4
(6.4%)
I 0.88
0.33
I
I
0.11
0.42
F5
(4.0%)
0.07
-0.13
0.14
0.15
-0.06
0.08
0.05
0.02
0.03
0.01
0.09
0.04
0.22
0.00
0.00
0.08
-0.04
0.18
-0.02
0.09
-0.07
0.20
0.31
0.06
0.21
-0.22
0.72
0.71
0.31
-0.07
0.81
0.04
0.65
-0.10
0.01
0.68
0.13
0.25
0.15
0.00
0.21
0.19
|
I
|
0.03
-0.22
0.08
0.15
0.16
0.20
No loading > 0.55.
NEG, COG, and POS scores into lower, middle, and
upper thirds. The negative subtype was defined as a
patient with a NEG score in a higher third than his or her
COG and POS scores; cognitive and positive subtypes
were defined correspondingly. When applying these criteria, we found 37 negative subtypes (14.6%), 45 positive
subtypes (17.8%), and 20 cognitive subtypes (7.9%); 151
patients (59.7%) showed a mixed psychopathology and
could not be attached to any subtype.
Subsidiary analyses of the effects of age, gender, and
neuroleptic (typical vs. atypical) medication yielded that
males were significantly more impaired in all of the
PANSS dimensions; NEG (r xy = 0.22, p < 0.001) and
COG scores (rxy = 0.17, p < 0.01) were correlated with
age; and patients receiving typical neuroleptics (n = 148)
reached higher EXC, COG, and POS ratings than patients
receiving atypical neuroleptic drugs (n = 77).
In table 2 the correlations between PANSS syndromes and questionnaires are presented. In general, the
correlation pattern corresponds with the differences of the
five syndromes. The paranoid subscale of the PDS
showed the strongest correlation with the positive syndrome (rxy = 0.47). The PDS depression subscale showed
the strongest correlation with the depression syndrome
(rxy = 0.35). The experimental scale FCQ-S showed the
strongest correlation with the cognitive syndrome (rxy =
0.22).
PDS-P (rxy = -0.41, p < 0.001), PDS-D (rxy = -0.58,
p < 0.001), and FCQ-S (rxy = -0.31, p < 0.001) all were
correlated with the Illness Denying control scale. To eval-
172
Psychopathological Syndromes
Schizophrenia Bulletin, Vol. 26, No. 1, 2000
uate the possible falsification of the questionnaire data by
denying attitudes, the correlations shown in table 2 were
recalculated (1) with exclusion of Illness Denying high
scorers (i.e., patients with a Denying score beyond 2 SD
of the mean) and (2) as partial correlations with the Illness
Denying scale as control variable. Both methods did not
yield substantial changes of the presented correlations.
The PANSS item G12 (lack of judgment and insight)
showed no significant relationship with FCQ-S and PDSD. Interestingly, it correlated positive with PDS-P (Rho
[Spearman's nonparametric correlation coefficient] =
0.35, p < 0.001). It seems that a patient's lack of insight to
suffer from a psychotic illness does not prevent him or her
automatically from admitting psychotic symptoms in
questionnaires.
In the CPT, the neuropsychologically tested subsample reached a mean d' of 1.54 (SD = 0.85, min = 0.06,
max = 3.93). The mean number of hits in the SAT was
14.4 (SD = 1.4, min = 9, max = 16) in the 3-letter control
condition and 59.3 (SD = 7.1, min = 41, max = 77) in the
8-letter condition. The mean MSE was 23.8 milliseconds
(SD = 36.3, min = -30, max = 139).
Table 3 shows the correlations between the PANSS
syndromes and the CPT, SAT, and MSE measures, respectively. The COG syndrome was significantly correlated
with CPT, and the NEG syndrome was significantly correlated with the MSE.
drome: for NEG, the variables Nl, N2, N3, N4, N6; for
EXC, the variables P4, P7, G8, G14; for COG, the variables P2, N5, G i l ; for DEP, the variables G2, G3, G6; for
POS, the variables PI, P3, G9.
As in former analyses (Peralta and Cuesta 1994), the
PANSS disorientation item (G10) had to be excluded in
the present study, since it was unsuitable for the description of the sample. This unsuitableness obviously resulted
from the a priori exclusion of patients with organic brain
syndromes. In a factor-analytical study on a gerontopsychiatric population, Overall and Beller (1984) found the
disorientation item of the BPRS to be the marker variable
of an organic brain deficit factor mainly constituted by the
Pfeiffer Mental Status Questionnaire (Pfeiffer 1975).
The correlations with questionnaire measures in general support the differences among the PANSS dimensions (table 2). However, statistical significances aside,
effect sizes obtained in these analyses are not great.
Furthermore, only a part of the experiences included in
the PANSS were covered by the questionnaires. Hence, all
interpretations have to be made with reservations. The
paranoid subscale of the PDS shows significant correlations with all of the PANSS syndromes; on the other hand,
the strongest relationship was found with the POS component (rxy = 0.47, p < 0.001). Similarly, the PDS depression
subscale was most distinctly correlated with the DEP
component (rxy = 0.35, p < 0.001). Relationships between
the FCQ and schizophrenic symptoms have already been
reported in former studies. Using the Scale for the
Assessment of Negative Symptoms (SANS; Andreasen
1982) and the Scale for the Assessment of Positive
Symptoms (SAPS; Andreasen 1984) for the evaluation of
psychopathological symptoms, Peralta et al. (1992) found
moderate correlations between FCQ scores and hallucinations, alogia, and formal thought disturbances. Since these
symptoms represent the three schizophrenic core syndromes (Andreasen et al. 1995), the present significant
correlations between FCQ-S and the NEG, COG, and
POS component of the PANSS (table 5) fit well with the
Peralta et al. (1992) results.
As mentioned above, in the present sample male
patients showed greater impairment in all psychopathological dimensions than females did. Corresponding to these
results, studies by Goldstein et al. (1990) and Josiassen et
al. (1990) showed that men with schizophrenia expressed
more negative symptoms (blunted/flat affect, emotional
withdrawal) than women. In contrast to our findings,
depressive symptoms are said to be more common in
females with schizophrenia (McGlashan and Bardenstein
1990). Probably the main reason for the psychopathological gender differences in the present study is the fact that
more than 50 percent of the males with schizophrenia were recruited on a closed ward where patients
with more severe symptoms are treated; only about 10
Discussion
The analyses of the present study indicate a high degree of
concordance with the results of former investigations. This
concordance is illustrated by the synopsis given in table 6.
Differences between the present findings and those of former studies may be influenced by (1) sample differences
(e.g., age), (2) differences in the cultural expression of illness, and (3) the fact that our raters were not trained by
experts from the Kay and Sevy group. Given these potential
sources of variance, the similarity of findings is quite
remarkable, indicating that the PANSS is a reliable instrument in the hands of experienced clinicians.
The present factorial solution shows highest accordance with the von Knorring and Lindstrom (1995)
results, since there are 22 corresponding PANSS item
attachments. The PANSS items N7 (stereotyped thinking),
G12 (lack of judgment and insight), G13 (disturbance of
volition), and G15 (preoccupation) could not be attached
to any of the factors when considering only loadings
greater than or equal to 0.55; von Knorring and Lindstrom
(1995) also excluded N7, G12, and G13.
In the studies considered in table 6, 18 PANSS items
show highest loadings throughout on the same factor, thus
representing stable components of the respective syn-
173
Schizophrenia Bulletin, Vol. 26, No. 1, 2000
R. Mass et al.
Table 6. Synopsis of factor-analytic studies with the PANSS1
Variable
P1
P2
P3
P4
P5
P6
P7
N1
N2
N3
N4
N5
N6
N7
G1
G2
G3
G4
G5
G6
G7
G8
G9
G10
G11
G12
G13
G14
G15
G16
present study
(n = 253)
POS
COG
POS
EXC
EXC
EXC
NEG
NEG
NEG
NEG
COG
NEG
DEP
DEP
EXC
EXC
DEP
EXC
POS
COG
EXC
NEG
Bell et al. 1994a
(Bronx sample)
(n = 240)
POS
COG
POS
EXC
POS
POS
EXC
NEG
NEG
NEG
NEG
COG
NEG
EXC
DEP
DEP
DEP
EXC
COG
DEP
NEG
EXC
POS
COG
COG
POS
COG
EXC
DEP
NEG
Bell etal. 1994a
(WHVA sample)
(n=146)
POS
COG
POS
EXC
POS
POS
EXC
NEG
NEG
NEG
NEG
COG
NEG
COG
POS
DEP
DEP
COG
COG
DEP
NEG
EXC
POS
COG
COG
NEG
EXC
NEG
DEP
von Knorring and
Lindstrom 1995
(n = 88)
POS
COG
POS
EXC
POS
EXC
NEG
NEG
NEG
NEG
COG
NEG
DEP
DEP
DEP
DEP
NEG
EXC
POS
COG
COG
EXC
COG
Note.—COG = cognitive syndrome; DEP = depression; EXC = hostile excitement; NEG = negative syndrome; POS = positive syndrome.
1
Because Kay and Sevy (1990), Kay and Sandyk (1991), Bell et al. (1994a, Bronx sample), and Lindenmayer et al. (1994) analyzed the
same sample of n = 240 schizophrenia subjects, only the Bell et al. (1994a) results were considered.
percent of the females were recruited on a closed station.
It is a well-known phenomenon that negative symptoms tend to increase with the duration of illness
(McGlashan and Fenton 1992); this phenomenon is probably reflected by the correlation of the NEG and COG
syndromes with age.
The observed differences between typical and atypical neuroleptic drugs should be interpreted only with
strong reservations. Since our study was not designed as a
test of psychopharmacological effects, medication was not
assigned randomly to the patients. Therefore, atypical
neuroleptics may have been administered preferably to
patients with less severe symptoms. Moreover, there were
significant differences in the administration of additional
drugs: patients receiving typical neuroleptics were given
benzodiazepines and biperiden more often than patients
receiving atypical neuroleptics were.
Goldberg et al. (1995) contrasted neuropsychological
performances of male and female subjects with schizophrenia on a wide range of tests over several independent
samples. They found only six significant gender differences (out of 96 t test comparisons); the tests that yielded
such differences were usually spatial or attentional in
nature, and men performed better than women. It has to
be noted that the number of significant results reported by
Goldberg et al. (1995) merely exceeded a error. Hence,
there is no evidence that gender has a substantial effect on
neuropsychological performance of people with schizo-
174
Psychopathological Syndromes
Schizophrenia Bulletin, Vol. 26, No. 1, 2000
phrenia. From this point of view, the fact that in the present study only male patients were neuropsychologically
tested should have no serious influence on the results.
The cognitive syndrome was the only PANSS component that was related to all of the neuropsychological
assessments (table 3) by showing a significant negative
correlation with the sensitivity index of the CPT (rxy =
-0.29, p < 0.05) and tendencies to correlate negatively
with the SAT performance (8-letter condition, rxy = -0.21,
p < 0.10) and positively with an enlarged MSE (rxy =
0.22, p < 0.10). These results confirm earlier findings of
Bilder et al. (1985) and Bell et al. (1994b), who reported
the disorganization to be associated with neuropsychological impairment. Hence, the COG syndrome of the
PANSS seems to be more closely connected to schizophrenic cognitive deficits than the other syndromes are.
However, considering the small effect sizes, interpretations should be made cautiously.
In earlier studies (Nuechterlein et al. 1986; Hain at al.
1993; Strauss 1993), inconsistent results have been
reported concerning the relations between CPT and psychopathological symptoms. Therefore, further investigations are necessary to value the present correlation
between CPT and the COG syndrome (rxy = -0.29, p <
0.05). Braff (1993) suggested that the difficulties of subjects with schizophrenia shifting attention from one
modality to another (MSE) could be considered as perseverations and thus be connected with hypofrontality of
subjects with schizophrenia on the Wisconsin Card
Sorting Test (Weinberger et al. 1986). Furthermore, there
is evidence that negative symptoms are related to
hypofrontality (Andreasen et al. 1992). This relation may
explain the correlation between the PANSS NEG syndrome and the MSE (table 3).
Surprisingly, the SAT (hit rate at the 8-letter condition) shows only weak correlations with the PANSS syndromes; usually SAT performance is associated with negative symptoms (e.g., Nuechterlein et al. 1986; Strauss et
al. 1993). However, in the present study the 3-letter condition of the SAT was significantly associated with the
NEG syndrome (rxy = -0.26, p < 0.05).
The subtype analysis presented above suggests that further research should emphasize the value of the dimensional
view on schizophrenic psychopathology. A first step in this
direction has been made in the DSM-IV (American
Psychiatric Association 1994, p. 71 Of) by proposing a syndromal alternative to the traditional subtype diagnosis.
Andreasen, N.C. The Scale for the Assessment of Negative
Symptoms. Iowa City, IA: University of Iowa, 1982.
Andreasen, N.C. The Scale for the Assessment of Positive
Symptoms. Iowa City, IA: University of Iowa, 1984.
Andreasen, N.C; Arndt, S.; Alliger, R.; Miller, D.; and
Flaum, M. Symptoms of schizophrenia. Methods, meanings, and mechanisms. Archives of General Psychiatry,
52:341-351, 1995.
Andreasen, N.C, and Olsen, S. Negative v. positive schizophrenia. Archives of General Psychiatry, 39:789-794, 1982.
Andreasen, N.C; Rezai, K.; Alliger, R.; Swayze, V.W. II;
Flaum, M.; Kirchner, P.; Cohen, G.; and O'Leary, D.S.
Hypofrontality in neuroleptic-naive patients and in
patients with chronic schizophrenia: Assessment with
xenon 133 single-photon emission computed tomography
and the Tower of London. Archives of General Psychiatry,
49:943-958, 1992.
Asarnow, R.F.; Granholm, E.; and Sherman, T. Span of
apprehension in schizophrenia. In: Steinhauer, S.R.;
Gruzelier, J.H.; and Zubin, J., eds. Handbook of
Schizophrenia: Neuropsychology, Psychophysiology and
Information Processing. Amsterdam, The Netherlands;
London, England; New York, NY; Tokyo, Japan: Elsevier
Science Publishers B.V., 1991. pp. 335-370.
Bell, M.D.; Lysaker, P.H.; Beam-Goulet, J.L.; Milstein,
R.M.; and Lindenmayer, J.-P. Five-component model of
schizophrenia: Assessing the factorial invariance of the
Positive and Negative Syndrome Scale. Psychiatry
Research, 52:295-303, 1994a.
Bell, M.D.; Lysaker, PH.; Milstein, R.M.; and BeamGoulet, J.L. Concurrent validity of the cognitive component of schizophrenia: Relationship of PANSS scores to
neuropsychological assessments. Psychiatry Research,
54:51-58, 1994&.
Bell, M.; Milstein, R.; Beam-Goulet, J.; Lysaker, P.; and
Cicchetti, D. The Positive and Negative Syndrome Scale
and the Brief Psychiatric Rating Scale. Reliability, comparability, and predictive validity. Journal of Nervous and
Mental Disease, 180:723-728, 1992.
Bilder, R.M.; Mukherjee, S.; Rieder, R.O.; and
Pandurangi, A.K. Symptomatic and neuropsychological
components of defect states. Schizophrenia Bulletin,
ll(3):409-419, 1985.
Braff, D.L. Information processing and attention in schizophrenia. Schizophrenia Bulletin, 19(2):233-259, 1993.
Cohen, R., and Rist, F. The modality shift effect: Further
explorations at the crossroads. In: Friedman, D., and
Bruder, G., eds. Annals of the New York Academy of
Sciences:
Psychophysiology
and
Experimental
Psychopathology.
New York, NY: The New York
Academy of Sciences, 1992. pp. 163-181.
References
American Psychiatric Association. DSM-lV-Diagnostic
and Statistical Manual of Mental Disorders. 4th ed.
Washington, DC: The Association, 1994.
175
Schizophrenia Bulletin, Vol. 26, No. 1, 2000
R. Mass et al.
Crow, T.J. Molecular pathology of schizophrenia: More
than one disease process? British Medical Journal,
January 12, 1980, pp. 66-68.
Mass, R.; Haasen, C ; and Krausz, M. Dimensional structure and diagnostic specifity of the Frankfurt Complaint
Questionnaire. European Psychiatry, 12:117-123, 1997.
Cuesta, M.J., and Peralta, V. Psychopathological dimen-
Mass, R.; Weigel, S.; Schneider, S.; and Klepsch, R.
Schizophrenia-specific basic symptoms: A successful
replication. Psychopathology, 31:113-119, 1998.
sions in schizophrenia. Schizophrenia
Bulletin,
21(3):473-482, 1995.
Goldberg, T.E.; Gold, J.M.; Torrey, E.F.; and Weinberger,
D.R. Lack of sex differences in the neuropsychological
performance of patients with schizophrenia. American
Journal of Psychiatry, 152:883-888, 1995.
McGlashan, T.H., and Bardenstein, K.K. Gender differences in affective, schizoaffective, and schizophrenic disorders. Schizophrenia Bulletin, 16(2):319-329, 1990.
McGlashan, T.H., and Fenton, W.S. The positive-negative
distinction in schizophrenia. Archives of General
Psychiatry, 49:63-72, 1992.
Goldstein, J.M.; Santangelo, S.L.; Simpson, J.C.; and
Tsuang, M.T. The role of gender in identifying subtypes
of schizophrenia: A latent class analytic approach.
Schizophrenia Bulletin, 16(2):263-275, 1990.
Nuechterlein, K.H., and Dawson, M.E. A heuristic vulnerability/stress model of schizophrenic episodes.
Schizophrenia Bulletin, 10(2):300-312, 1984.
Granholm, E.; Asarnow, R.F.; and Marder, S.R. Display
visual angle and attentional scanpaths on the Span of
Apprehension Task in schizophrenia. Journal of
Abnormal Psychology, 105:17-24, 1996.
Green, D.M., and Swets, J.A. Signal Detection Theory
and Psychophysics. New York, NY: John Wiley, 1966.
Nuechterlein, K.H.; Dawson, M.E.; and Green, M.F.
Information-processing abnormalities as neuropsychological vulnerability indicators for schizophrenia. Acta
Psychiatrica Scandinavica, 90(Suppl 384):71-79, 1994.
Nuechterlein, K.H.; Dawson, M.E.; Ventura, J.; Fogelson,
D.; Gitlin, M.; and Mintz, J. Testing vulnerability models:
Stability of potential vulnerability indicators across clinical state. In: Hafner, H., and Gattaz, W.F., eds. Search for
the Causes of Schizophrenia. Vol. 2. Berlin, Germany:
Springer-Verlag, 1991, pp. 177-191.
Hain, C ; Maier, W.; Klingler, T.; and Franke, P.
Positive/negative symptomatology and experimental measures of attention in schizophrenic patients.
Psychopathology, 26:62-68, 1993.
Hiller, W.; Zaudig, M.; and Mombour, W. IDCL:
Internationale Diagnosen-Checklisten fur ICD-10 und
DSM-IV. Bern, Switzerland; Gottingen, Germany;
Toronto, Canada; Seattle, WA: Verlag Hans Huber, 1995.
Nuechterlein, K.H.; Edell, W.S.; Norris, M.; and Dawson,
M.E. Attentional vulnerability indicators, thought disorder, and negative symptoms. Schizophrenia Bulletin,
12(3):408^26, 1986.
Josiassen, R.C.; Roemer, R.A.; Johnson, M.M.; and
Shagass, C. Are gender differences in schizophrenia
reflected in brain event-related potentials? Schizophrenia
Bulletin, 16(2): 229-246, 1990.
Overall, J.E., and Beller, S.A. The Brief Psychiatric
Rating Scale (BPRS) in gerontopsychiatric research: I.
Factor structure on an inpatient unit. Journal of
Gerontology, 39:187-193, 1984.
Kay, S.R.; Fiszbein, A.; and Opler, L.A. The Positive and
Negative Syndrome Scale (PANSS) for schizophrenia.
Schizophrenia Bulletin, 13(2):261-276, 1987.
Overall, J.E., and Gorham, D.R. The Brief Psychiatric
Rating Scale. Psychological Reports, 10:799-812, 1962.
Kay, S.R., and Sandyk, R. Experimental models of schiz-
Peralta, V., and Cuesta, M.J. Psychometric properties of
the Positive and Negative Syndrome Scale (PANSS) in
schizophrenia. Psychiatry Research, 53:31-40, 1994.
ophrenia. International Journal of Neuroscience,
58:69-82, 1991.
Kay, S.R., and Sevy, S. Pyramidical model of schizophrenia. Schizophrenia Bulletin, 16(3):537-545, 1990.
Peralta, V.; Cuesta, M.J.; and de Leon, J. Positive versus
negative schizophrenia and basic symptoms.
Comprehensive Psychiatry, 33:202-206, 1992.
Liddle, P.F. Schizophrenic syndromes, cognitive performance and neurological dysfunction. Psychological
Medicine, 17:49-57, 1987.
Psychiatry,
Pfeiffer, E.A. A short portable mental status questionnaire
for the assessment of organic brain deficit in elderly
patients. Journal of the American Geriatric Society,
23:433-441, 1975.
Lindenmayer, J.-P.; Bernstein-Hyman, R.; and
Grochowski, S. Five-factor model of schizophrenia:
Initial validation. Journal of Nervous and Mental Disease,
182:631-638, 1994.
Prosser, E.S.; Csemansky, J.G.; Kaplan, J.; Thiemann, S.;
Becker, T.J.; and Hollister, L.E. Depression, parkinsonian
symptoms, and negative symptoms in schizophrenics
treated with neuroleptics. Journal of Nervous and Mental
Disease, 175:100-105, 1987.
Liddle, P.F., and Barnes, T.R.E. Syndromes of chronic
schizophrenia. British
Journal
of
157:558-561, 1990.
176
Psychopathological Syndromes
Schizophrenia Bulletin, Vol. 26, No. 1, 2000
Rist, F., and Cohen, R. Sequential effects in the reaction
times of schizophrenics: Crossover and modality shift
effects. In: Steinhauer, S.R.; Gruzelier, J.H.; and Zubin, J.,
eds. Handbook of Schizophrenia:
Neuropsychology,
Psychophysiology and Information Processing. Amsterdam,
The Netherlands; London, U.K.; New York, NY; Tokyo,
Japan: Elsevier Science Publishers B.V., 1991. pp. 241-271.
von Zerssen, D.
Paranoid-Depressivitats-Skala.
Depressivitats-Skala. Manual. Weinheim: Beltz, 1976.
Weinberger, D.R.; Berman, K.F.; and Zee, R.F.
Physiological dysfunction of dorsolateral prefrontal cortex in schizophrenia. Archives of General Psychiatry,
43:114-124, 1986.
World Health Organization. The ICD-10 Classification of
Mental and Behavioral Disorders: Clinical Descriptions
and Diagnostic Guidelines. Geneva, Switzerland: WHO,
1992.
Singh, M.M., and Kay, S.R. A comparative study of
haloperidol and chlorpromazine in terms of clinical
effects and therapeutic reversal with benztropine in schizophrenia: Theoretical implications for potency differences
among neuroleptics. Psychopharmacologia, 43:103-113,
1975.
Acknowledgments
Statistical Package for the Social Sciences Inc. Statistical
Package for the Social Sciences Reference
Guide.
Chicago, IL, 1990.
This research was supported by a grant from the German
Research Society, Project Ma 1765/1: "SchizophreniaSpecific Subjective Basic Symptoms." The authors wish
to thank Dr. Burghard Andresen, Psychiatric Clinic of the
University Hospital Hamburg-Eppendorf, for his helpful
comments.
Strauss, M.E. Relations of symptoms to cognitive deficits
in schizophrenia. Schizophrenia Bulletin, 19(2):215-231,
1993.
Strauss, M.E.; Buchanan, R.W.; and Hale, J. Relations
between attentional deficits and clinical symptoms in
schizophrenic out-patients. Psychiatry
Research,
47:205-213, 1993.
The Authors
Reinhard Mass, Ph.D., and Christian Haasen, M.D., are
Clinical Research Assistants, Psychiatric Clinic,
University Hospital, Hamburg-Eppendorf, Germany.
Thomas Schoemig, M.D., Klaus Hitschfeld, M.D., and
Eleonora Wall, M.D., are Psychiatrists, Klinikum Nord,
Hamburg-Ochsenzoll, Germany.
Siillwold, L. Manual zum Frankfurter
BeschwerdeFragebogen. Berlin, Germany: Springer-Verlag, 1991.
von Knorring, L., and Lindstrom, E. Principal components and further possibilities with the PANSS. Ada
Psychiatrica Scandinavica, 91:5-10, 1995.
177