The
Positive-Negative Distinction in Schizophrenia Review of Natural
History Validators
Thomas H. McGlashan, MD, \s=b\ A
review of the interaction between the positive-negative
symptom distinction in schizophrenia and multiple
mea-
of illness natural history reveals some redundant and compelling patterns. Negative or deficit symptoms are often associated with inferior social/instrumental functioning premorbidly, more abnormal voluntary/involuntary movements at illness presentation, and poorer long-term outcome when present beyond the early phase of illness. Negative symptoms are semi-independent of positive symptoms. They are variable early in the illness but accrue in severity, stability, and prognostic weight with time. The nature of the processes that generate negative symptoms and their specificity to schizophrenia remain to be elucidated. Nevertheless, it is clear that negative symptoms are a common and valid component of schizophrenia and deserve recognition as such in our nosology. (Arch Gen Psychiatry. 1992;49:63-72) sures
the past decade, the distinction between positive and In negative symptoms in schizophrenia has captured the attention of researchers and clinicians alike.
Perhaps,
in
retrospect, the revitalized focus on this perspective was long overdue, fueled by frustrations with the vast heter¬
ogeneity of schizophrenia and by disenchantments with the preexisting (classic) subtyping schema. Furthermore, developing diagnostic systems including Research Diag¬
nostic Criteria,1 DSM-lll,2 and DSM-III-R3 leaned heavily on the more reliable positive symptoms to define schizo¬ phrenia, an emphasis that many considered psychometrically sound but scientifically suspect. Eschewing nega¬ tive symptoms or deficit psychopathologies, some postulated, weakens construct validity by ignoring pro¬ cesses that have been considered central to schizophrenia for almost 100 years. The recent reliable operationalization of negative phénoménologies, however, rendered empiric study of these issues within the purview of many investigators, and a virtual explosion of investigative ef¬ fort ensued. We will attempt to review only a segment of this effort, that dealing with the natural history validators
Accepted for publication March 4, 1991. From the Yale Psychiatric Institute, New Haven, Conn (Dr McGlashan), and the Chestnut Lodge Research Institute, Rockville, Md (Dr Fenton). Reprint requests to Yale Psychiatric Institute, PO Box 12-A, Yale Station, New Haven, CT 06520 (Dr McGlashan).
Wayne S. Fenton,
MD
positive-negative distinction. Other validating cri¬ keeping with Robins and Guze,4 such as neu¬ roimaging, neurochemistry, and neuropsychological test¬ ing, while no less important, have been reviewed elsewhere.5 We focus on the longitudinal clinical perspec¬ of the
teria, in
tive, ie, how the person and the disease present and in¬ teract
over a
significant proportion of the life span.
BRIEF HISTORY OF THE POSITIVE-NEGATIVE
DISTINCTION The distinction between positive and negative schizo¬ phrenia is ubiquitous historically, and no single person can lay claim to its original formulation. In the 19th cen¬ tury, Jackson6 used the terms negative and positive symp¬ toms to describe insanity and speculated that the former came from disease-induced loss of higher mental func¬ tioning and the latter from release phenomena secondary to this loss of higher control. The distinction was present in the thinking of our nosologie forefathers of the 20th century. Kraepelin7 described "two principal disorders that characterize the malady [dementia praecox] a emotional of those which activities weakening perma¬ nently form the mainspring of volition and the loss of the inner unity of activities of intellect, emotion, and volition."7pp74"75 Bleuler8 had a similar negative-positive dichotomy in mind with his fundamental vs accessory symptoms. The former involved loss of function (eg, of attention, volition, affective responsiveness, and associa¬ tion) and was always present. The latter involved an ab¬ erration of function (eg, hallucinations, delusions, and catatonia) and was present only during severe relapse. This distinction was carried through the 20th century in one form or another. In fact, the most anecdotal but tell¬ ing validation of the positive-negative distinction comes from the degree to which it has entered—and saturated our everyday clinical language. Schizophrenia has almost always been described in terms of a bipolarity, the valence of which suggests today's positive-negative distinction. The following dichotomies from Sass9 are illustrative: acute (positive) vs chronic (negative), reactive (positive) vs process (negative), secondary (positive) vs primary (negative), accessory (positive) vs fundamental (nega¬ tive), active (positive) vs residual (negative), florid (pos¬ itive) vs quiescent (negative), and productive (positive) vs deficit (negative).
Downloaded From: http://archpsyc.jamanetwork.com/ by a Georgetown University Medical Center User on 07/11/2015
...
.
.
.
—
explicit hypothesis that these phénomén¬ ologies represented distinct pathophysiologies within schizophrenia came from Strauss et al.10 This notion was elaborated soon after by Crow,11 who postulated two syndromes and psychopathologic processes in schizo¬ phrenia. Type I schizophrenia had mainly positive symp¬ toms, good premorbid functioning, acute onsets, neuroleptic-responsive symptoms, and better long-term The first
and outcome without intellectual deterioration. Type schizophrenia had mainly negative symptoms, poor premorbid functioning, insidious onset, drugresistant symptoms, and poorer long-term course and outcome with intellectual deterioration or dementia. Crow12 later added behavioral deterioration and abnormal course
II
involuntary movements to type II. Type I was postulated to reflect reversible hyperdopaminergic activity in a nor¬
mal brain and type II to reflect irreversible neuronal loss in a structurally abnormal brain.
Crow's powerfully heuristic hypothesis catalyzed a chain reaction of research, once systems for operationalizing positive and negative phénoménologies became available, first from Britain13 and then from the United States.14 The 1980s witnessed the appearance of literally hundreds of articles on the topic, studying virtually every class of validation, including clinical phenomenology, lab¬
oratory findings, family history,
treatment
responsive¬
long-term course and outcome. Many of these inves¬ tigations have already become the subject of reviews.15"20 ness,
DEFINITIONS AND RELIABILITY OF POSITIVE AND NEGATIVE SYMPTOMS have been devised to define positive and Many systems negative symptoms. Seven have been used most fre¬ quently during the past decade: the Manchester Scale13,21 and Crow's modification of this scale,12 the Positive and Negative Symptom Scale of Kay et al,22 Andreasen and Olsen's Scale for the Assessment of Negative Symptoms and Scale for the Assessment of Positive Symptoms,23 Abrams and Taylor's Rating Scale for Emotional Blunt¬ ing,24 Carpenter and associates' deficit and nondeficit cat¬ egorization,25 and scales devised by Lewine et al26 and Pogue-Geile and Harrow.27 Which symptoms are considered positive or negative varies across these systems. All systems include flat affect and poverty of speech among the negative symptoms and hallucinations and delusions among the positive symp¬ toms. Most systems regard anhedonia, apathy, avolition, or abulia as negative symptoms when they are included. Thought disorder, bizarre behavior, and inappropriate affect, on the other hand, are classified as positive or negative with variability. More recently, Kulhara et al,28 Liddle,29'30 and Liddle and colleagues31 suggested that they constitute a third factor that is orthogonal to the positive-negative distinction. Lewine and Sommers32 asserted that inappropriate af¬ fect and poverty of content of speech (a form of thought disorder) fit better in a separate domain reflecting a loss of higher integrative capacity. Fenton and McGlashan33 found thought disorder and bizarre behavior, in contrast to hallucinations and delusions, to correlate with negative symptoms and with poorer long-term outcome. On the other hand, thought disorder and bizarre behavior were unlike negative symptoms but like positive symptoms in being unrelated to a variety of measures of premorbid functioning. Therefore, it appears that these domains of
psychopathology stand in ambiguous relationship to the positive-negative distinction. Nevertheless, they are oc¬ casionally included in the definitions used in the studies
as such, this may confound the interpretation of results in ways that are difficult to specify, since com¬ parisons are almost always made between samples de¬ fined by groups of symptoms, ie, by composite scores. Operationalizing the assessment of negative symptoms has had a positive effect on reliability. All of the scales specified above have generated levels of interrater reli¬ ability in the acceptable range. For example, the intraclass correlations for the Scale for the Assessment of Negative Symptoms and Scale for the Assessment of Positive Symptoms symptom complex global ratings ranged be¬ tween .70 and .88,14 and the interrater Pearson correla¬ tions for the negative symptom items of the Positive and Negative Symptom Scale averaged . .22 A more complete comparison of these assessment systems with regard to reliability and other psychometric measures will be the subject of another communication.34 For now, suffice to say that negative symptoms can be rated with as much reliability as positive symptoms. Whether they actually were rated reliably in the studies cited herein is another matter. Reliability of assessment was tested and reported in only about one of five of the investigations reviewed. Clearly, there is room for improvement.
reviewed;
SCOPE AND METHOD OF REVIEW This review is limited to the set of "natural history" validators, ie, those variables that can be elaborated at the clinical phénom¬ énologie level. These variables are as follows: demography, in¬ cluding age, sex, and marital status; premorbid functioning, in¬ cluding social/sexual, education/work, and IQ; illness loading, including family history of mental illness, twin concordance, age at onset, established chronicity at baseline, and neurologic abnormalities; relationship between positive and negative symp¬ toms; course of positive and negative symptoms; and prognosis and outcome. Approximately 200 potentially relevant articles from the En¬ glish psychiatric and psychological literature of the past 15 years were collected with the aid of computer searches using such key words as positive, negative, deficit, type / , anhedonia, apathy, flat or blunted affect, asociality, avolition, alogia, poverty of speech, and so on. The méthodologie standards applied were modestly rigorous. Selected were studies of schizophrenia that characterized diag¬ nosis and the positive-negative or deficit-nondeficit typology using current operational criteria. Also included were studies in which these variables could be translated easily into these con¬ structs. Reliability testing was not required for inclusion; had it been, this review would be remarkably brief. Studies without tests of significance were excluded, as were studies with samples that were probably too small for meaningful results (eg, popu¬ lations or comparison groups of less than a dozen subjects). NATURAL HISTORY VALIDATORS
Demography Age.—Eleven studies tested for different ages at the time of positive-negative assessment.23,25-33,35"42 All recorded no differ¬ ences, except that of Opler et al,39 who found patients in the negative category to be older. Overall, age differences were un¬ remarkable.
Gender.—Categoric and dimensional differences by gender in
and negative symptoms were not found in 13 positive ¡es 23,33,35,36,38-47 ]\jegatjve or deficit symptoms were found
studto be
characteristic of male patients in nine studies.25,33,48"54 Only investigation found male outpatients to have lower negative symptom scores.55 In sum, gender differences appeared at a
more one
Downloaded From: http://archpsyc.jamanetwork.com/ by a Georgetown University Medical Center User on 07/11/2015
moderate frequency. When present, they almost always were in the direction of more negative symptoms among men. Marital Status.—All seven investigations studying marital status found no difference on the positive-negative dimension.23,38"40-44,46,51 Subdividing their sample by length of ill¬ ness, however, Kay et al46 found more negative symptoms among the unmarried patients in the chronic (3 to 10 years of ill¬ ness) and long-term chronic (>10 years of illness) subgroups. Similarly, Fenton and McGlashan33 found single status corre¬ lated with negative symptoms in their largely chronic schizo¬ phrenic sample. Overall, however, this variable failed to dis¬ criminate, perhaps because of the low overall rate of marriage among schizophrenic patients generally.
Premorbid
Functioning
Social/Sexual Functioning. Seven investigations found neg¬ ative or deficit symptoms significantly correlated with poorer social and/or sexual functioning in childhood and/or adoles¬ cence, ie, in the premorbid era.23,45,48'56'59 Two studies recorded the same finding but only for male patients.40,46 One study found the opposite, ie, a significant correlation between negative symptoms and better premorbid functioning among 19 acute schizophrenic patients.60 One study recorded nonsignificant correlations.33 Overall, the link between negative symptoms and poor premorbid social functioning appears to be robust in all ex¬ cept perhaps the most acute schizophrenic samples. Education/Work.—The profile for education follows some¬ what the same pattern as premorbid social functioning. In nine studies, or the majority, negative symptoms were correlated significantly with lower levels of education.* A similar associa¬ tion was found in two studies for males only40,63 and in one study for more chronically ill patients.46 Wagman et al64 found schizo¬ phrenic patients with deficit symptoms to be less educated than a normal control group; the same did not hold true for those without deficit symptoms. Six studies found no differences in educational level,25,30,35,36,41,44 and one found female patients with negative symptoms to have a higher level of education65 than fe¬ male patients without negative symptoms. Inferior premorbid work adjustment in patients with negative was noted in all studies rating this vari¬ symptoms able W*10*17 Intelligence.—Intelligence has been studied less frequently. Numerous assessment techniques have been applied, some premorbidly and others concurrent with the illness. Further¬ more, added to this variability is the fact that the level of psychosis is seldom controlled for between comparison groups, rendering solid inferences hazardous. Four studies found no correlation between positive-negative status and IQ.36·39·65·68 Four studies43,48,51,69 found a significant association between negative 41 symptoms and lower IQ. Pogue-Geile and Harrow found the same for negative symptoms and thought disorder, and Aylward et al70 found the same for male patients only. Wagman et al64 found that patients with deficit symptoms scored lower on IQ tests than did a normal control group, whereas patients without deficit symptoms did not. Other studies found patients with negative symptoms scoring poorly on intelligence-related assessments, such as the Mini-Mental Status Examination23 or cognitive performance tests.71 Overall, the results suggest an as¬ sociation between lower intelligence and negative symptoms, but this is not definitive. —
Illness
Loading Variables
The following dimensions constitute various inputs to or cor¬ relates of illness severity and depth. Family History of Mental Illness.—Results of intergenerational genetic linkage studies are sparse and equivocal. While Sautter et al72 found a family history of schizophrenia correlated with negative symptoms, Fenton and McGlashan,33 Johnstone et
al,62 and Pearlson et al66,67 did not. Kay et al46 found a family his¬ tory of psychosis significantly related to negative symptoms in a long-term chronic population (>10 years of illness) but not in a more acute sample (
Positive-Negative Distinction in Schizophrenia Review of Natural
History Validators
Thomas H. McGlashan, MD, \s=b\ A
review of the interaction between the positive-negative
symptom distinction in schizophrenia and multiple
mea-
of illness natural history reveals some redundant and compelling patterns. Negative or deficit symptoms are often associated with inferior social/instrumental functioning premorbidly, more abnormal voluntary/involuntary movements at illness presentation, and poorer long-term outcome when present beyond the early phase of illness. Negative symptoms are semi-independent of positive symptoms. They are variable early in the illness but accrue in severity, stability, and prognostic weight with time. The nature of the processes that generate negative symptoms and their specificity to schizophrenia remain to be elucidated. Nevertheless, it is clear that negative symptoms are a common and valid component of schizophrenia and deserve recognition as such in our nosology. (Arch Gen Psychiatry. 1992;49:63-72) sures
the past decade, the distinction between positive and In negative symptoms in schizophrenia has captured the attention of researchers and clinicians alike.
Perhaps,
in
retrospect, the revitalized focus on this perspective was long overdue, fueled by frustrations with the vast heter¬
ogeneity of schizophrenia and by disenchantments with the preexisting (classic) subtyping schema. Furthermore, developing diagnostic systems including Research Diag¬
nostic Criteria,1 DSM-lll,2 and DSM-III-R3 leaned heavily on the more reliable positive symptoms to define schizo¬ phrenia, an emphasis that many considered psychometrically sound but scientifically suspect. Eschewing nega¬ tive symptoms or deficit psychopathologies, some postulated, weakens construct validity by ignoring pro¬ cesses that have been considered central to schizophrenia for almost 100 years. The recent reliable operationalization of negative phénoménologies, however, rendered empiric study of these issues within the purview of many investigators, and a virtual explosion of investigative ef¬ fort ensued. We will attempt to review only a segment of this effort, that dealing with the natural history validators
Accepted for publication March 4, 1991. From the Yale Psychiatric Institute, New Haven, Conn (Dr McGlashan), and the Chestnut Lodge Research Institute, Rockville, Md (Dr Fenton). Reprint requests to Yale Psychiatric Institute, PO Box 12-A, Yale Station, New Haven, CT 06520 (Dr McGlashan).
Wayne S. Fenton,
MD
positive-negative distinction. Other validating cri¬ keeping with Robins and Guze,4 such as neu¬ roimaging, neurochemistry, and neuropsychological test¬ ing, while no less important, have been reviewed elsewhere.5 We focus on the longitudinal clinical perspec¬ of the
teria, in
tive, ie, how the person and the disease present and in¬ teract
over a
significant proportion of the life span.
BRIEF HISTORY OF THE POSITIVE-NEGATIVE
DISTINCTION The distinction between positive and negative schizo¬ phrenia is ubiquitous historically, and no single person can lay claim to its original formulation. In the 19th cen¬ tury, Jackson6 used the terms negative and positive symp¬ toms to describe insanity and speculated that the former came from disease-induced loss of higher mental func¬ tioning and the latter from release phenomena secondary to this loss of higher control. The distinction was present in the thinking of our nosologie forefathers of the 20th century. Kraepelin7 described "two principal disorders that characterize the malady [dementia praecox] a emotional of those which activities weakening perma¬ nently form the mainspring of volition and the loss of the inner unity of activities of intellect, emotion, and volition."7pp74"75 Bleuler8 had a similar negative-positive dichotomy in mind with his fundamental vs accessory symptoms. The former involved loss of function (eg, of attention, volition, affective responsiveness, and associa¬ tion) and was always present. The latter involved an ab¬ erration of function (eg, hallucinations, delusions, and catatonia) and was present only during severe relapse. This distinction was carried through the 20th century in one form or another. In fact, the most anecdotal but tell¬ ing validation of the positive-negative distinction comes from the degree to which it has entered—and saturated our everyday clinical language. Schizophrenia has almost always been described in terms of a bipolarity, the valence of which suggests today's positive-negative distinction. The following dichotomies from Sass9 are illustrative: acute (positive) vs chronic (negative), reactive (positive) vs process (negative), secondary (positive) vs primary (negative), accessory (positive) vs fundamental (nega¬ tive), active (positive) vs residual (negative), florid (pos¬ itive) vs quiescent (negative), and productive (positive) vs deficit (negative).
Downloaded From: http://archpsyc.jamanetwork.com/ by a Georgetown University Medical Center User on 07/11/2015
...
.
.
.
—
explicit hypothesis that these phénomén¬ ologies represented distinct pathophysiologies within schizophrenia came from Strauss et al.10 This notion was elaborated soon after by Crow,11 who postulated two syndromes and psychopathologic processes in schizo¬ phrenia. Type I schizophrenia had mainly positive symp¬ toms, good premorbid functioning, acute onsets, neuroleptic-responsive symptoms, and better long-term The first
and outcome without intellectual deterioration. Type schizophrenia had mainly negative symptoms, poor premorbid functioning, insidious onset, drugresistant symptoms, and poorer long-term course and outcome with intellectual deterioration or dementia. Crow12 later added behavioral deterioration and abnormal course
II
involuntary movements to type II. Type I was postulated to reflect reversible hyperdopaminergic activity in a nor¬
mal brain and type II to reflect irreversible neuronal loss in a structurally abnormal brain.
Crow's powerfully heuristic hypothesis catalyzed a chain reaction of research, once systems for operationalizing positive and negative phénoménologies became available, first from Britain13 and then from the United States.14 The 1980s witnessed the appearance of literally hundreds of articles on the topic, studying virtually every class of validation, including clinical phenomenology, lab¬
oratory findings, family history,
treatment
responsive¬
long-term course and outcome. Many of these inves¬ tigations have already become the subject of reviews.15"20 ness,
DEFINITIONS AND RELIABILITY OF POSITIVE AND NEGATIVE SYMPTOMS have been devised to define positive and Many systems negative symptoms. Seven have been used most fre¬ quently during the past decade: the Manchester Scale13,21 and Crow's modification of this scale,12 the Positive and Negative Symptom Scale of Kay et al,22 Andreasen and Olsen's Scale for the Assessment of Negative Symptoms and Scale for the Assessment of Positive Symptoms,23 Abrams and Taylor's Rating Scale for Emotional Blunt¬ ing,24 Carpenter and associates' deficit and nondeficit cat¬ egorization,25 and scales devised by Lewine et al26 and Pogue-Geile and Harrow.27 Which symptoms are considered positive or negative varies across these systems. All systems include flat affect and poverty of speech among the negative symptoms and hallucinations and delusions among the positive symp¬ toms. Most systems regard anhedonia, apathy, avolition, or abulia as negative symptoms when they are included. Thought disorder, bizarre behavior, and inappropriate affect, on the other hand, are classified as positive or negative with variability. More recently, Kulhara et al,28 Liddle,29'30 and Liddle and colleagues31 suggested that they constitute a third factor that is orthogonal to the positive-negative distinction. Lewine and Sommers32 asserted that inappropriate af¬ fect and poverty of content of speech (a form of thought disorder) fit better in a separate domain reflecting a loss of higher integrative capacity. Fenton and McGlashan33 found thought disorder and bizarre behavior, in contrast to hallucinations and delusions, to correlate with negative symptoms and with poorer long-term outcome. On the other hand, thought disorder and bizarre behavior were unlike negative symptoms but like positive symptoms in being unrelated to a variety of measures of premorbid functioning. Therefore, it appears that these domains of
psychopathology stand in ambiguous relationship to the positive-negative distinction. Nevertheless, they are oc¬ casionally included in the definitions used in the studies
as such, this may confound the interpretation of results in ways that are difficult to specify, since com¬ parisons are almost always made between samples de¬ fined by groups of symptoms, ie, by composite scores. Operationalizing the assessment of negative symptoms has had a positive effect on reliability. All of the scales specified above have generated levels of interrater reli¬ ability in the acceptable range. For example, the intraclass correlations for the Scale for the Assessment of Negative Symptoms and Scale for the Assessment of Positive Symptoms symptom complex global ratings ranged be¬ tween .70 and .88,14 and the interrater Pearson correla¬ tions for the negative symptom items of the Positive and Negative Symptom Scale averaged . .22 A more complete comparison of these assessment systems with regard to reliability and other psychometric measures will be the subject of another communication.34 For now, suffice to say that negative symptoms can be rated with as much reliability as positive symptoms. Whether they actually were rated reliably in the studies cited herein is another matter. Reliability of assessment was tested and reported in only about one of five of the investigations reviewed. Clearly, there is room for improvement.
reviewed;
SCOPE AND METHOD OF REVIEW This review is limited to the set of "natural history" validators, ie, those variables that can be elaborated at the clinical phénom¬ énologie level. These variables are as follows: demography, in¬ cluding age, sex, and marital status; premorbid functioning, in¬ cluding social/sexual, education/work, and IQ; illness loading, including family history of mental illness, twin concordance, age at onset, established chronicity at baseline, and neurologic abnormalities; relationship between positive and negative symp¬ toms; course of positive and negative symptoms; and prognosis and outcome. Approximately 200 potentially relevant articles from the En¬ glish psychiatric and psychological literature of the past 15 years were collected with the aid of computer searches using such key words as positive, negative, deficit, type / , anhedonia, apathy, flat or blunted affect, asociality, avolition, alogia, poverty of speech, and so on. The méthodologie standards applied were modestly rigorous. Selected were studies of schizophrenia that characterized diag¬ nosis and the positive-negative or deficit-nondeficit typology using current operational criteria. Also included were studies in which these variables could be translated easily into these con¬ structs. Reliability testing was not required for inclusion; had it been, this review would be remarkably brief. Studies without tests of significance were excluded, as were studies with samples that were probably too small for meaningful results (eg, popu¬ lations or comparison groups of less than a dozen subjects). NATURAL HISTORY VALIDATORS
Demography Age.—Eleven studies tested for different ages at the time of positive-negative assessment.23,25-33,35"42 All recorded no differ¬ ences, except that of Opler et al,39 who found patients in the negative category to be older. Overall, age differences were un¬ remarkable.
Gender.—Categoric and dimensional differences by gender in
and negative symptoms were not found in 13 positive ¡es 23,33,35,36,38-47 ]\jegatjve or deficit symptoms were found
studto be
characteristic of male patients in nine studies.25,33,48"54 Only investigation found male outpatients to have lower negative symptom scores.55 In sum, gender differences appeared at a
more one
Downloaded From: http://archpsyc.jamanetwork.com/ by a Georgetown University Medical Center User on 07/11/2015
moderate frequency. When present, they almost always were in the direction of more negative symptoms among men. Marital Status.—All seven investigations studying marital status found no difference on the positive-negative dimension.23,38"40-44,46,51 Subdividing their sample by length of ill¬ ness, however, Kay et al46 found more negative symptoms among the unmarried patients in the chronic (3 to 10 years of ill¬ ness) and long-term chronic (>10 years of illness) subgroups. Similarly, Fenton and McGlashan33 found single status corre¬ lated with negative symptoms in their largely chronic schizo¬ phrenic sample. Overall, however, this variable failed to dis¬ criminate, perhaps because of the low overall rate of marriage among schizophrenic patients generally.
Premorbid
Functioning
Social/Sexual Functioning. Seven investigations found neg¬ ative or deficit symptoms significantly correlated with poorer social and/or sexual functioning in childhood and/or adoles¬ cence, ie, in the premorbid era.23,45,48'56'59 Two studies recorded the same finding but only for male patients.40,46 One study found the opposite, ie, a significant correlation between negative symptoms and better premorbid functioning among 19 acute schizophrenic patients.60 One study recorded nonsignificant correlations.33 Overall, the link between negative symptoms and poor premorbid social functioning appears to be robust in all ex¬ cept perhaps the most acute schizophrenic samples. Education/Work.—The profile for education follows some¬ what the same pattern as premorbid social functioning. In nine studies, or the majority, negative symptoms were correlated significantly with lower levels of education.* A similar associa¬ tion was found in two studies for males only40,63 and in one study for more chronically ill patients.46 Wagman et al64 found schizo¬ phrenic patients with deficit symptoms to be less educated than a normal control group; the same did not hold true for those without deficit symptoms. Six studies found no differences in educational level,25,30,35,36,41,44 and one found female patients with negative symptoms to have a higher level of education65 than fe¬ male patients without negative symptoms. Inferior premorbid work adjustment in patients with negative was noted in all studies rating this vari¬ symptoms able W*10*17 Intelligence.—Intelligence has been studied less frequently. Numerous assessment techniques have been applied, some premorbidly and others concurrent with the illness. Further¬ more, added to this variability is the fact that the level of psychosis is seldom controlled for between comparison groups, rendering solid inferences hazardous. Four studies found no correlation between positive-negative status and IQ.36·39·65·68 Four studies43,48,51,69 found a significant association between negative 41 symptoms and lower IQ. Pogue-Geile and Harrow found the same for negative symptoms and thought disorder, and Aylward et al70 found the same for male patients only. Wagman et al64 found that patients with deficit symptoms scored lower on IQ tests than did a normal control group, whereas patients without deficit symptoms did not. Other studies found patients with negative symptoms scoring poorly on intelligence-related assessments, such as the Mini-Mental Status Examination23 or cognitive performance tests.71 Overall, the results suggest an as¬ sociation between lower intelligence and negative symptoms, but this is not definitive. —
Illness
Loading Variables
The following dimensions constitute various inputs to or cor¬ relates of illness severity and depth. Family History of Mental Illness.—Results of intergenerational genetic linkage studies are sparse and equivocal. While Sautter et al72 found a family history of schizophrenia correlated with negative symptoms, Fenton and McGlashan,33 Johnstone et
al,62 and Pearlson et al66,67 did not. Kay et al46 found a family his¬ tory of psychosis significantly related to negative symptoms in a long-term chronic population (>10 years of illness) but not in a more acute sample (
