Natural
History of Schizophrenia Subtypes
II. Positive and
Negative Symptoms and Long-term Course
Wayne S. Fenton, MD, Thomas natural history and long-term course of schizophrenia divided by pervasiveness of positive and negative symptoms was explored among 187 schizophrenic patients from the Chestnut Lodge follow-up study. Schizophrenia with many negative symptoms was associated with poor premorbid functioning, insidious onset, partial or no remissions during the first several years of illness, and in most cases a progressive course leading to permanent disability. Schizophrenia with few negative symptoms was associated with good premorbid functioning, acute onset, intermittent early course, and a better prognosis. Positive symptoms predicted future hospitalizations but were less powerful and specific as indicators of differential illness history, course, and long-term functional incapacity. As predictors of long-term outcome, negative symptoms were of greater value measured at index admission several years after illness onset than at first hospital admission. Multivariate analyses indicated that two negative symptoms (anhedonia and affective flattening) contribute significantly to outcome variance independent of their association with premorbid functioning or positive symptoms. Patients with the poorest long-term outcome tended to show an increase in negative symptoms during the early years of their illness. Progressive negative symptoms early in the course of schizophrenia may thus reflect or signal a process leading to long-term functional dis\s=b\ The
ability.
(Arch Gen Psychiatry. 1991;48:978-986)
distinction between positive and negative symp¬ The schizophrenia has been recognized for much of this but studied until Strauss al2 toms in
et century1 was rarely Crow3 and their poten¬ suggested explicitly hypothesized tial significance as dual underlying pathophysiologic processes. Type 1 schizophrenia was described by Crow as having many positive symptoms, good premorbid functioning, acute onset, and neuroleptic responsive symptoms. Type 2 schizophrenia had many negative symptoms, poorer premorbid functioning, insidious on¬ set, and drug-resistant symptoms. Type 1 schizophrenia
Accepted for publication March 4, 1991. From the Chestnut Lodge Research Institute, Rockville, Md (Dr Fenton); and the Yale Psychiatric Institute, New Haven, Conn (Dr
McGlashan). Reprint requests to Chestnut Lodge Research Institute, 500 Montgomery Ave, Rockville, MD 20850 (Dr Fenton).
W
H.
McGlashan,
MD
postulated to reflect reversible hyperdopaminergic activity in an anatomically intact central nervous system was
and type 2 to reflect irreversible neuronal loss.4,5 The development of reliable methods of measuring negative and positive symptoms facilitated a substantial research effort aimed at testing, validating, and refining this powerful heuristic hypothesis.6"13 Evidence of the va¬ lidity of the distinction, as outlined in recent reviews,14-19 focuses on validators from multiple domains, including neuroimaging, neurochemistry, and neuropsychological functioning. Likewise, as reviewed in our companion ar¬ ticle,20 a growing body of literature has focused on the differential clinical correlates and natural history valida¬ tors of positive and negative symptoms. As a putative reflection of an underlying structural brain abnormality, negative symptoms are often assumed to portend long-term deterioration and functional disabil¬ ity. Although there is considerable evidence linking neg¬ ative symptoms to poor premorbid functioning,12-21"27 most studies have been cross sectional28-29 or of limited duration.30"36 Prospective longitudinal studies will even¬ tually provide an understanding of the long-term course of positive and negative symptoms, but they have only begun to bear fruit.37"40 As a result, we currently lenow lit¬ tle about the natural history and prognostic significance of negative (or positive) symptoms of schizophrenia over the long term.41 The availability of detailed clinical data for a cohort of schizophrenic patients treated largely in the predrug era and followed up an average of 19 years after admission allowed us to construct a longitudinal profile of the life and illness history for patients differing in pervasiveness of negative and positive symptoms. Herein we report on the evolution of positive and negative symptoms over the early illness course, their association with the heteroge¬ neity of illness natural history, and their predictive valid¬ ity as indicators of long-term functional outcome. PATIENTS AND METHODS Data Sources A complete méthodologie outline of the Chestnut Lodge Follow-up Study has been presented elsewhere.42-43 Included in the original study were all patients discharged from a tertiarycare psychiatric hospital between 1950 and 1975 and a smaller cohort of nondischarged inpatients from a comparable period. Patients without organic brain syndrome who were between 16
Downloaded From: by a UNIVERSITY OF SYDNEY LIBRARY User on 03/31/2018
and 55 years of age at admission and who had been treated at Chestnut Lodge Hospital, Rockville, Md, for at least 90 days were selected. Initially, two realms of data were collected: (1) demographic, premorbid, sign and symptom, and diagnostic data were retrospectively rated from abstracts of index admis¬ sion medical records and (2) long-term follow-up data were in¬ dependently collected via interviews with subjects and/or sig¬ nificant others an average of 19 years after the index admission (range, 6 to 32 years). The reliability of measures used for these assessments has been reported.42 For the current study, subtype ratings were generated for the index hospitalization for 187 patients from the follow-up who on review received a research diagnosis (mostly DSM-III)44 oí schizophrenia. As outlined previously, index admission ratings were made on the basis of extensive archival records (not abstracts), including written observations of multiple informants and verbatim transcripts of clinical case conferences.45 These were sufficient to allow positive and negative symptom ratings from the index admission with minimal missing information. Records from an earlier (usually first) psychiatric hospitalization were also available for 67% (125/187) of the patient cohort. These were less detailed, and missing ratings were coded for between 1% and 7% of patients, depending on the particular scale employed. Additional historical, manifest illness, and illness course variables were also collected. The reliability of these as¬ sessments has been detailed previously with data on the natural history of the classic schizophrenia subtypes.45
Positive and
Negative Symptoms
Because of a lack of consensus about what constitutes positive and negative symptoms or how best to measure them, patients were rated by means of several available scales and subtype systems.6"10-13*46*47 The reliability, comprehensiveness, and pre¬ dictive validity of these scales will be compared in a future report. For the current study, the Schedule for the Assessment of Negative Symptoms (SANS) and Schedule for the Assessment of Positive Symptoms (SAPS) were used to rate and classify pa¬ tients' symptoms.46,47 These emphasize observable behaviors frequently recorded in clinical records. Ratings consisted of as¬ sessments of the global severity and persistence of each of five
negative symptom complexes (affective flattening, alogia, avolition/apathy, anhedonia/asociality, and attentional impair¬ ment) and four positive symptom complexes (hallucinations, delusions, positive formal thought disorder, and bizarre behav¬ ior). Each of these nine symptom complexes consists of behav¬ ioral components that correlate highly with each other and with
the global symptom complex severity.12 This redundancy makes the SANS and SAPS particularly useful for coding archival ma¬ terial, where only a subset of multiple indicators of the symptom complex may be described. A rating of the global severity and persistence of each of the nine symptom complexes was coded, with great weight given to one or two prominent component symptoms or behaviors if they were particularly striking and enduring. Total positive and negative symptom scores were ob¬ tained by summing scores of the four positive and five negative
symptom complexes, respectively.
After a period of rule and consensus development, two raters
(W.S.F. and K. Bardenstein, PhD) independently scored 26 cases to test the reliability of the symptom scales applied to the med¬
ical record data. Reliability for total SANS and SAPS was excel¬ lent with the use of highly detailed index admission records (in¬ traclass correlation, .83 for SANS, .85 for SAPS) and adequate with the use of less detailed first admission records (intraclass correlation, .47 for SANS, .66 for SAPS). Information with ref¬ erence to the patient's condition during the 6 months leading to index and/or first hospitalization and the initial 3 months of each hospitalization was considered in assigning ratings. To be rated as severe, symptoms must have been described as marked or prominent and must have been present during much or all of the observation period. Although we intended to divide patients into those having
positive, negative, or mixed syndromes in accordance with the original classification scheme of Andreasen and Olsen,12 this did not prove feasible. Because of our long window of observation—9 months—the majority of patients at some time demonstrated both positive and negative symptoms and hence classified as mixed. Considerable variation, however, was observed in the relative pervasiveness and severity of positive and negative symptoms during the observation period. To explore the relationship between this symptomatic heterogene¬ ity and natural history of the illness in the most straightforward manner, the sample was divided at the median SANS and SAPS scores into overlapping comparison groups rated high and low on negative and positive symptoms. were
Natural
History
previously, 26 clinical elements coded from three independent sources were selected as representing significant components of illness natural history.45 These elements were grouped into eight domains: genetic predisposition, develop¬ mental problems, premorbid functioning, illness onset features, early illness course, subtype stability, long-term functional out¬ come, and suicide risk. Patients classified as showing high and As detailed
low levels of negative symptoms at index admission were com¬ pared across dependent variables from each domain; this proce¬ dure was repeated with the sample divided by severity of pos¬ itive symptoms. The significance of differences among groups was assessed by the 2 x 2 or 2 x 3 2 test for dichotomous (or trichotomous) dependent variables and t test for continuous dependent variables. To offset the probability of chance find¬ ings, individual levels were set for variables in each natural history domain by dividing 0.05 by the number of variables in the domain.48 Correlation coefficients and the statistic, respec¬ tively, were used to assess the dimensional and categorical sta¬ bility of positive and negative phenomenology across first and index admissions, which were separated by an average of 4.5 years. Where significant associations were ascertained between glo¬ bal measures of positive and negative symptoms and features of the natural history of the illness, correlation analyses were used to assess the independent contribution of individual SANS and SAPS symptom complexes to the scales' overall predictive util¬ ity. Finally, multivariate analyses were employed to explore the long-term prognostic significance of positive and negative symp¬ toms independent of their association with premorbid function¬
ing.
Medication Status
Nearly all ratings were based on observations of patients who had never used neuroleptics and/or who were in an unmedicated
(mean year, 1954), 117 (98%) of 120 pa¬ tients rated were taking no neuroleptics. Up to index admission, 100 (53.1%) of 187 had never been treated with antipsychotics. At index admission (mean year, 1959), 141 (75%) of 187 patients state. At first admission
the majority of those who were had all medications discontinued within several weeks after admission. Comparisons of patients with and with¬ out previous neuroleptic exposure at each admission revealed no significant differences in age, gender, or mean SANS and SAPS scores. Consequently, these groups were collapsed in subse¬ were
not
taking neuroleptics;
(40/46; 87%)
quent analyses.
RESULTS
Prevalence of Positive and
Negative Symptoms
Index admission SANS and SAPS ratings were obtained for 187 patients. Both positive and negative symptoms were com¬ mon among the schizophrenic patients studied. The proportion of patients rated as showing at least some level of flattened af¬ fect, alogia, avolition/apathy, anhedonia, and attentional im¬ pairment was 72%, 48%, 71%, 76%, and 46%, respectively. The proportion rated as having marked or severe levels of these neg¬ ative symptoms was 17%, 16%, 25%, 23%, and 17%, respec-
Downloaded From: by a UNIVERSITY OF SYDNEY LIBRARY User on 03/31/2018
SANS
Total SANS total
1.0
Flat affect (1 )
12
.78t 1.0
Alogia (2)
.67+
SAPS Total .37+
.21+
-.13
.47+
.38+
.55+
.43+
.25+
.10
-.12
.32+
.34+
.70+
.50+
.31+
.22+
-.14
.42+
.26+
.74+
.49+
.29+
.18+
-.17+
.45+
.25+
.40+
.22+
.09
-.10
.32+
.24+
.47+
.30+
.02
.41+
.43+
.75+
.56+
.62+
.53+
.31+
.25+
.17+
.07
.00
3
4
.87+
.88+
.85+
.61+
.59+
.69+ 1.0
1.0
Avolition(3) Anhedonia (4)
5
1.0
Impaired attention (5)
1.0
SAPS total
6
1.0
Hallucinations (6) Delusions (7) Bizarre behavior (8) Thought disorder (9)
8
7
1.0
1.0
1.0
.22+
1.0
"SANS indicates Schedule for the Assessment of P>.01 except as indicated. +P-C001. +P-C.01.
Negative Symptoms; SAPS, Schedule for the
Assessment of Positive
Symptoms. All correlations
are
Negative Symptomst Low
Negative
High Negative
Positive Symptoms* Low
High
Positive
Positive
(N 87)
(N 100)
ANOVA
1.9±1.2
2.1 ±1.2
NS
(28)
19 (19)
NS
3.5±1.1
NS
(N 98)
ANOVA
2.0 ±1.2
2.0 ±1.3
NS§
33(37)
10(10)
3.8±0.5
3.5 ±1.1
NS
3.8 ±0.5
2.6±1.1
2.1 ±1.3
.003
2.3 + 1.3
2.4 ±1.2
NS
117±12
111±14
.002
115±13
113±13
NS
(N 89) =
=
X2
=
=
Quality social contact (0, no
meets with friends at least once/wk) Marital status, No. (%) married Quantity of work (0, none; 4, full time) Skills and interests (0, no skills; 4, good skills and
friends; 4,
interests)
IQ
.0000
24
NS
*ANOVA indicates analysis of variance; NS, not significant. tLow negative, less than or equal to 6 on Schedule for the Assessment of Negative Symptoms; high negative, greater than or equal to 7. +Low positive, less than or equal to 8 on Schedule for the Assessment of Positive Symptoms; high positive, greater than or equal to 9.
§P>.01.
tively. The proportion of patients with some level of hallucina¬ tions, delusions, bizarre behavior, and thought disorder was 66%, 91%, 87%, and 66%, respectively. The proportion rated as having marked or severe levels for these positive symptoms was 28%, 47%, 40%, and 8%, respectively. Correlation of Positive and
Negative Symptoms
Correlations among symptom clusters at index admission shown in Table 1. All of the negative symptoms were highly correlated with each other, as were some positive symptoms, albeit to a lesser extent. Total SANS and SAPS scores were positively correlated due to a positive association between bizarre behavior, thought disorder, and negative symptoms. Delusions and hallucinations, in contrast, were inversely or only weakly associated with negative symptoms. Because of the overall positive correlation between index ad¬ mission SANS and SAPS scores, 57% (51/89) of patients with low negative symptoms (eg, below the median) also had low positive symptoms, and 63% (62/98) of patients with high negative symptoms were also high in positive symptoms
Age, Chronicity, and Gender Patients with high and low negative symptoms at index admission did not differ in age, duration of illness, or number of previous hospitalizations. Male patients accounted for 60% (59/98) of patients with high negative symptoms and female pa¬ tients for 56% (50/89) of those with low negative symptoms ( 2 4.38, d/=l, P
History of Schizophrenia Subtypes
II. Positive and
Negative Symptoms and Long-term Course
Wayne S. Fenton, MD, Thomas natural history and long-term course of schizophrenia divided by pervasiveness of positive and negative symptoms was explored among 187 schizophrenic patients from the Chestnut Lodge follow-up study. Schizophrenia with many negative symptoms was associated with poor premorbid functioning, insidious onset, partial or no remissions during the first several years of illness, and in most cases a progressive course leading to permanent disability. Schizophrenia with few negative symptoms was associated with good premorbid functioning, acute onset, intermittent early course, and a better prognosis. Positive symptoms predicted future hospitalizations but were less powerful and specific as indicators of differential illness history, course, and long-term functional incapacity. As predictors of long-term outcome, negative symptoms were of greater value measured at index admission several years after illness onset than at first hospital admission. Multivariate analyses indicated that two negative symptoms (anhedonia and affective flattening) contribute significantly to outcome variance independent of their association with premorbid functioning or positive symptoms. Patients with the poorest long-term outcome tended to show an increase in negative symptoms during the early years of their illness. Progressive negative symptoms early in the course of schizophrenia may thus reflect or signal a process leading to long-term functional dis\s=b\ The
ability.
(Arch Gen Psychiatry. 1991;48:978-986)
distinction between positive and negative symp¬ The schizophrenia has been recognized for much of this but studied until Strauss al2 toms in
et century1 was rarely Crow3 and their poten¬ suggested explicitly hypothesized tial significance as dual underlying pathophysiologic processes. Type 1 schizophrenia was described by Crow as having many positive symptoms, good premorbid functioning, acute onset, and neuroleptic responsive symptoms. Type 2 schizophrenia had many negative symptoms, poorer premorbid functioning, insidious on¬ set, and drug-resistant symptoms. Type 1 schizophrenia
Accepted for publication March 4, 1991. From the Chestnut Lodge Research Institute, Rockville, Md (Dr Fenton); and the Yale Psychiatric Institute, New Haven, Conn (Dr
McGlashan). Reprint requests to Chestnut Lodge Research Institute, 500 Montgomery Ave, Rockville, MD 20850 (Dr Fenton).
W
H.
McGlashan,
MD
postulated to reflect reversible hyperdopaminergic activity in an anatomically intact central nervous system was
and type 2 to reflect irreversible neuronal loss.4,5 The development of reliable methods of measuring negative and positive symptoms facilitated a substantial research effort aimed at testing, validating, and refining this powerful heuristic hypothesis.6"13 Evidence of the va¬ lidity of the distinction, as outlined in recent reviews,14-19 focuses on validators from multiple domains, including neuroimaging, neurochemistry, and neuropsychological functioning. Likewise, as reviewed in our companion ar¬ ticle,20 a growing body of literature has focused on the differential clinical correlates and natural history valida¬ tors of positive and negative symptoms. As a putative reflection of an underlying structural brain abnormality, negative symptoms are often assumed to portend long-term deterioration and functional disabil¬ ity. Although there is considerable evidence linking neg¬ ative symptoms to poor premorbid functioning,12-21"27 most studies have been cross sectional28-29 or of limited duration.30"36 Prospective longitudinal studies will even¬ tually provide an understanding of the long-term course of positive and negative symptoms, but they have only begun to bear fruit.37"40 As a result, we currently lenow lit¬ tle about the natural history and prognostic significance of negative (or positive) symptoms of schizophrenia over the long term.41 The availability of detailed clinical data for a cohort of schizophrenic patients treated largely in the predrug era and followed up an average of 19 years after admission allowed us to construct a longitudinal profile of the life and illness history for patients differing in pervasiveness of negative and positive symptoms. Herein we report on the evolution of positive and negative symptoms over the early illness course, their association with the heteroge¬ neity of illness natural history, and their predictive valid¬ ity as indicators of long-term functional outcome. PATIENTS AND METHODS Data Sources A complete méthodologie outline of the Chestnut Lodge Follow-up Study has been presented elsewhere.42-43 Included in the original study were all patients discharged from a tertiarycare psychiatric hospital between 1950 and 1975 and a smaller cohort of nondischarged inpatients from a comparable period. Patients without organic brain syndrome who were between 16
Downloaded From: by a UNIVERSITY OF SYDNEY LIBRARY User on 03/31/2018
and 55 years of age at admission and who had been treated at Chestnut Lodge Hospital, Rockville, Md, for at least 90 days were selected. Initially, two realms of data were collected: (1) demographic, premorbid, sign and symptom, and diagnostic data were retrospectively rated from abstracts of index admis¬ sion medical records and (2) long-term follow-up data were in¬ dependently collected via interviews with subjects and/or sig¬ nificant others an average of 19 years after the index admission (range, 6 to 32 years). The reliability of measures used for these assessments has been reported.42 For the current study, subtype ratings were generated for the index hospitalization for 187 patients from the follow-up who on review received a research diagnosis (mostly DSM-III)44 oí schizophrenia. As outlined previously, index admission ratings were made on the basis of extensive archival records (not abstracts), including written observations of multiple informants and verbatim transcripts of clinical case conferences.45 These were sufficient to allow positive and negative symptom ratings from the index admission with minimal missing information. Records from an earlier (usually first) psychiatric hospitalization were also available for 67% (125/187) of the patient cohort. These were less detailed, and missing ratings were coded for between 1% and 7% of patients, depending on the particular scale employed. Additional historical, manifest illness, and illness course variables were also collected. The reliability of these as¬ sessments has been detailed previously with data on the natural history of the classic schizophrenia subtypes.45
Positive and
Negative Symptoms
Because of a lack of consensus about what constitutes positive and negative symptoms or how best to measure them, patients were rated by means of several available scales and subtype systems.6"10-13*46*47 The reliability, comprehensiveness, and pre¬ dictive validity of these scales will be compared in a future report. For the current study, the Schedule for the Assessment of Negative Symptoms (SANS) and Schedule for the Assessment of Positive Symptoms (SAPS) were used to rate and classify pa¬ tients' symptoms.46,47 These emphasize observable behaviors frequently recorded in clinical records. Ratings consisted of as¬ sessments of the global severity and persistence of each of five
negative symptom complexes (affective flattening, alogia, avolition/apathy, anhedonia/asociality, and attentional impair¬ ment) and four positive symptom complexes (hallucinations, delusions, positive formal thought disorder, and bizarre behav¬ ior). Each of these nine symptom complexes consists of behav¬ ioral components that correlate highly with each other and with
the global symptom complex severity.12 This redundancy makes the SANS and SAPS particularly useful for coding archival ma¬ terial, where only a subset of multiple indicators of the symptom complex may be described. A rating of the global severity and persistence of each of the nine symptom complexes was coded, with great weight given to one or two prominent component symptoms or behaviors if they were particularly striking and enduring. Total positive and negative symptom scores were ob¬ tained by summing scores of the four positive and five negative
symptom complexes, respectively.
After a period of rule and consensus development, two raters
(W.S.F. and K. Bardenstein, PhD) independently scored 26 cases to test the reliability of the symptom scales applied to the med¬
ical record data. Reliability for total SANS and SAPS was excel¬ lent with the use of highly detailed index admission records (in¬ traclass correlation, .83 for SANS, .85 for SAPS) and adequate with the use of less detailed first admission records (intraclass correlation, .47 for SANS, .66 for SAPS). Information with ref¬ erence to the patient's condition during the 6 months leading to index and/or first hospitalization and the initial 3 months of each hospitalization was considered in assigning ratings. To be rated as severe, symptoms must have been described as marked or prominent and must have been present during much or all of the observation period. Although we intended to divide patients into those having
positive, negative, or mixed syndromes in accordance with the original classification scheme of Andreasen and Olsen,12 this did not prove feasible. Because of our long window of observation—9 months—the majority of patients at some time demonstrated both positive and negative symptoms and hence classified as mixed. Considerable variation, however, was observed in the relative pervasiveness and severity of positive and negative symptoms during the observation period. To explore the relationship between this symptomatic heterogene¬ ity and natural history of the illness in the most straightforward manner, the sample was divided at the median SANS and SAPS scores into overlapping comparison groups rated high and low on negative and positive symptoms. were
Natural
History
previously, 26 clinical elements coded from three independent sources were selected as representing significant components of illness natural history.45 These elements were grouped into eight domains: genetic predisposition, develop¬ mental problems, premorbid functioning, illness onset features, early illness course, subtype stability, long-term functional out¬ come, and suicide risk. Patients classified as showing high and As detailed
low levels of negative symptoms at index admission were com¬ pared across dependent variables from each domain; this proce¬ dure was repeated with the sample divided by severity of pos¬ itive symptoms. The significance of differences among groups was assessed by the 2 x 2 or 2 x 3 2 test for dichotomous (or trichotomous) dependent variables and t test for continuous dependent variables. To offset the probability of chance find¬ ings, individual levels were set for variables in each natural history domain by dividing 0.05 by the number of variables in the domain.48 Correlation coefficients and the statistic, respec¬ tively, were used to assess the dimensional and categorical sta¬ bility of positive and negative phenomenology across first and index admissions, which were separated by an average of 4.5 years. Where significant associations were ascertained between glo¬ bal measures of positive and negative symptoms and features of the natural history of the illness, correlation analyses were used to assess the independent contribution of individual SANS and SAPS symptom complexes to the scales' overall predictive util¬ ity. Finally, multivariate analyses were employed to explore the long-term prognostic significance of positive and negative symp¬ toms independent of their association with premorbid function¬
ing.
Medication Status
Nearly all ratings were based on observations of patients who had never used neuroleptics and/or who were in an unmedicated
(mean year, 1954), 117 (98%) of 120 pa¬ tients rated were taking no neuroleptics. Up to index admission, 100 (53.1%) of 187 had never been treated with antipsychotics. At index admission (mean year, 1959), 141 (75%) of 187 patients state. At first admission
the majority of those who were had all medications discontinued within several weeks after admission. Comparisons of patients with and with¬ out previous neuroleptic exposure at each admission revealed no significant differences in age, gender, or mean SANS and SAPS scores. Consequently, these groups were collapsed in subse¬ were
not
taking neuroleptics;
(40/46; 87%)
quent analyses.
RESULTS
Prevalence of Positive and
Negative Symptoms
Index admission SANS and SAPS ratings were obtained for 187 patients. Both positive and negative symptoms were com¬ mon among the schizophrenic patients studied. The proportion of patients rated as showing at least some level of flattened af¬ fect, alogia, avolition/apathy, anhedonia, and attentional im¬ pairment was 72%, 48%, 71%, 76%, and 46%, respectively. The proportion rated as having marked or severe levels of these neg¬ ative symptoms was 17%, 16%, 25%, 23%, and 17%, respec-
Downloaded From: by a UNIVERSITY OF SYDNEY LIBRARY User on 03/31/2018
SANS
Total SANS total
1.0
Flat affect (1 )
12
.78t 1.0
Alogia (2)
.67+
SAPS Total .37+
.21+
-.13
.47+
.38+
.55+
.43+
.25+
.10
-.12
.32+
.34+
.70+
.50+
.31+
.22+
-.14
.42+
.26+
.74+
.49+
.29+
.18+
-.17+
.45+
.25+
.40+
.22+
.09
-.10
.32+
.24+
.47+
.30+
.02
.41+
.43+
.75+
.56+
.62+
.53+
.31+
.25+
.17+
.07
.00
3
4
.87+
.88+
.85+
.61+
.59+
.69+ 1.0
1.0
Avolition(3) Anhedonia (4)
5
1.0
Impaired attention (5)
1.0
SAPS total
6
1.0
Hallucinations (6) Delusions (7) Bizarre behavior (8) Thought disorder (9)
8
7
1.0
1.0
1.0
.22+
1.0
"SANS indicates Schedule for the Assessment of P>.01 except as indicated. +P-C001. +P-C.01.
Negative Symptoms; SAPS, Schedule for the
Assessment of Positive
Symptoms. All correlations
are
Negative Symptomst Low
Negative
High Negative
Positive Symptoms* Low
High
Positive
Positive
(N 87)
(N 100)
ANOVA
1.9±1.2
2.1 ±1.2
NS
(28)
19 (19)
NS
3.5±1.1
NS
(N 98)
ANOVA
2.0 ±1.2
2.0 ±1.3
NS§
33(37)
10(10)
3.8±0.5
3.5 ±1.1
NS
3.8 ±0.5
2.6±1.1
2.1 ±1.3
.003
2.3 + 1.3
2.4 ±1.2
NS
117±12
111±14
.002
115±13
113±13
NS
(N 89) =
=
X2
=
=
Quality social contact (0, no
meets with friends at least once/wk) Marital status, No. (%) married Quantity of work (0, none; 4, full time) Skills and interests (0, no skills; 4, good skills and
friends; 4,
interests)
IQ
.0000
24
NS
*ANOVA indicates analysis of variance; NS, not significant. tLow negative, less than or equal to 6 on Schedule for the Assessment of Negative Symptoms; high negative, greater than or equal to 7. +Low positive, less than or equal to 8 on Schedule for the Assessment of Positive Symptoms; high positive, greater than or equal to 9.
§P>.01.
tively. The proportion of patients with some level of hallucina¬ tions, delusions, bizarre behavior, and thought disorder was 66%, 91%, 87%, and 66%, respectively. The proportion rated as having marked or severe levels for these positive symptoms was 28%, 47%, 40%, and 8%, respectively. Correlation of Positive and
Negative Symptoms
Correlations among symptom clusters at index admission shown in Table 1. All of the negative symptoms were highly correlated with each other, as were some positive symptoms, albeit to a lesser extent. Total SANS and SAPS scores were positively correlated due to a positive association between bizarre behavior, thought disorder, and negative symptoms. Delusions and hallucinations, in contrast, were inversely or only weakly associated with negative symptoms. Because of the overall positive correlation between index ad¬ mission SANS and SAPS scores, 57% (51/89) of patients with low negative symptoms (eg, below the median) also had low positive symptoms, and 63% (62/98) of patients with high negative symptoms were also high in positive symptoms
Age, Chronicity, and Gender Patients with high and low negative symptoms at index admission did not differ in age, duration of illness, or number of previous hospitalizations. Male patients accounted for 60% (59/98) of patients with high negative symptoms and female pa¬ tients for 56% (50/89) of those with low negative symptoms ( 2 4.38, d/=l, P
