DIRECTOR: Beng Choon Ho, M.D

The Longitudinal Course of Recent Onset Psychotic Disorders and Predictors of Outcome

 

Findings Supportive of the Cognitive Dysmetria Construct

 

Findings Supportive of Neurodevelopmental Mechanisms

Development and Revision of Assessment and Diagnostic Instruments

 

References

 

Projects -- a description of current and planned projects in the Diagnosis and Phenomenology Unit

Since the central theme of this MH-CRC involves the effort to relate the signs and symptoms of the major psychoses to their underlying neurobiological mechanisms, the Diagnosis and Phenomenology Unit is conceptually and practically linked to all of the other core and research units.
As discussed in the Introduction and Overview, schizophrenia researchers now have very powerful tools such as molecular biology and neuroimaging to study this complex illness. Yet in stark contrast to the possibility of these tools helping us to understand the biological underpinnings of this illness is the fact that there are so many very basic questions about the clinical presentation of the illness yet to be answered. What are the characteristic symptoms of schizophrenia? What are the subtypes and boundaries? What is the common longitudinal course? Can we predict outcome in any meaningful way? Is the illness one of developmental etiology with a static course or is there degeneration and deterioration over time? Despite decades of study, these questions go unanswered.

One of the most perplexing aspects of the phenomenology of schizophrenia is the heterogeneity of clinical signs and symptoms. Medical students rotating through our five bed inpatient unit are routinely struck by this and are confused that such an apparently diverse group of patients are referred to by the same diagnostic label. They may see one patient lying in bed day after day, detached, apparently unconcerned about the goings on within or outside the unit, and seemingly organized only around the need to get himself up for cigarettes once an hour. Another patient spends all of her time chattering with any passers-by who will listen, in a silly, disorganized and childlike manner. Still another takes every opportunity to go out on passes so that he can get to the library and continue working on his thesis regarding "astrophysical reality," on which he has already completed several volumes. This diversity of symptoms, which suggests a diversity of disease processes, is perhaps schizophrenia's most intellectually challenging feature.

As outlined in the general Overview and Introduction to the research plan as well as in several papers in the Appendices (e.g., Andreasen et al 1993b), there are three types of competing explanatory models for the apparent heterogeneity: the multiple disease entities model, the multiple dimensions model, and the single disease entity model. Each one is described very briefly here, and in more detail in the introduction.

Multiple Disease Entities Model: This model postulates that there are a variety of disease entities leading to schizophrenia via different etiopathologic processes. This history of this construct is traced to the proposed disease entities of paranoia, hebephrenia, and catatonia, which were defined before Kreapelin joined them in the dementia praecox syndrome. Later examples include Crow's Type I/Type II schema, or our own division into positive/mixed/negative subtypes. The most recent and highly generative example of this research model is the study of the deficit syndrome. Carpenter's group has postulated that patients with primary, prominent, and enduring negative symptoms have a deficit syndrome, a form of schizophrenia with distinct and separate etiology. They have supported this with several studies that show deficit patients to have distinctly different patterns of cognitive function (Buchanan et al 1994), symptom patterns (Kirkpatrick et al 1990; Kirkpatrick et al 1994), brain morphology (Buchanan et al 1993), risk for spontaneous dyskinesia (Fenton et al 1994b), eye movement abnormality (Ross et al 1996), Borna disease virus seropositivity (Waltrip et al 1995) and course and outcome (Fenton et al 1994a).

Multiple Dimensions Model: A dimensional approach postulates that specific symptom clusters within schizophrenia reflect different disease processes that combine in different ways in different patients. This construct is fundamentally different from the multiple disease entity model, which requires the assumption of distinct groups or subgroups. Rather, this approach can incorporate the idea that specific symptoms or groups of symptoms, which may be underlain by discrete pathophysiological processes, may coexist within individuals in a variety of combinations. Thus, instead of limiting our questions to how schizophrenics differ from controls on a variable of interest, we can ask how individuals who display a given symptom(s) prominently differ from those who do not.

Factor analytic studies have shown, with a great amount of consistency, that positive symptoms subdivide into two dimensions, a psychosis dimension, and a disorganization dimension, while negative symptoms make up the third dimension (for review, see Andreasen et al 1995k). Our own group has done work validating the biological correlates of the three dimensions by demonstrating that they: 1) have different functional neural substrates as seen with PET, 2) different structural brain correlates as evaluated with MRI, and 3) and may also have different and independent longitudinal courses (Miller 1997b; Flaum 1995d; Arndt 1995b; Andreasen 1995l; Andreasen 1996k; Andreasen 1997c; Flaum 1997c; O'Leary submitted b).

Single Disease Entity (Unitary) Model: : This model was first used by Bleuler, who believed that fundamental flaws (e.g., loosening of associations) explained the disorder and that the various accessory symptoms of schizophrenia were secondary to a basic change in cognitive/emotional processes. Therefore, each subject is thought to have a central pathologic process despite differences in symptom manifestation. One of the most recent incarnations of this model is the one of "cognitive dysmetria" put forward by our group (Andreasen 1997d; Andreasen et al in press a). This fundamental cognitive abnormality is the consequence of an underlying disruption in neural functions, which leads to a disruption of the fluid, coordinated sequences of thought and action that are the hallmark of normal cognition.

All three of the above constructs have their pros and cons. Therefore, our strategy for conducting research on the diagnosis and phenomenology of schizophrenia has involved an ongoing cross-walk between them, as outlined in the projects below. For instance, Project 1 outlines our current longitudinal prospective study which has used the dimensional approach for many years to evaluate the course of psychotic, disorganized, and negative symptom dimensions and their phenomenologic correlates. Project 2 is a proposal to prospectively follow a new group of patients, focusing more on the unitary disease model with focused assessment of measures of "cognitive dysmetria" and the brain regions subserving the circuit involved. Finally, Project 3 explores the reliability and validity of the primary (deficit) vs. secondary distinction.

This unit also has an additional focus on investigation into the role of developmental processes as causative or contributing factors to the occurrence of the illness. Several studies have shown that children who later develop schizophrenia manifest a variety of abnormalities, including speech problems, disturbances in affect, poor cognitive function and academic achievement, motor disturbance and social difficulties (Rosen et al 1971; Aylward 1984; Jones 1994; Done 1994; Walker 1990; Walker 1994; Murray 1992). Within the prospective longitudinal studies (Projects 1 and 2), there are a variety of assessments which can be used to explore possible developmental mechanisms. For example, in collaboration with Elaine Walker, we have begun the study of childhood home movies to assess affect and/or motor abnormalities that may have been present in our subjects premorbidly. In addition, Project 4 is directed at specifically evaluating premorbid cognitive function. The goal of the project is to retrospectively establish a data base of standardized childhood test scores in order to examine not only premorbid cognitive function, but also to track the premorbid course of cognitive function (scores are obtained at grades 4, 8, and 11), and to evaluate other phenomenological correlates of premorbid cognitive function such as age of onset, neurologic soft signs, and structural brain abnormalities.
Both Projects 1 and 2 described above follow patients over a period of time -- up to 10 years or more. However, this illness lasts a lifetime. Study of the first 10 years of the illness (after onset of psychosis) is probably still not nearly long enough to understand the full course of the illness. Studies have suggested that there are significant changes in severity and type of symptoms during the decades beyond 50 years of age (Pfohl et al 1982; Schultz et al 1997a). Moreover, some patients with schizophrenia become demented, and little is known about why this occurs, in whom it occurs, or the biological correlates of the dementia (e.g., is the dementia related to types of symptoms, types of treatment, or movement disorder?). Project 5 is designed to investigate some of these issues in a sample of elderly patients with schizophrenia.
This unit is also concerned with the building blocks that make this type of effort meaningful, i.e., the development of reliable and valid methods for assessing clinical signs and symptoms, and the use of these measures in large samples of suitable individuals. Project 6 describes the ongoing development and testing of a new assessment instrument called the "5-STEPS" (an acronym for "5-dimensional Scale To Evaluate Psychopathology in Schizophrenia"). This is the product of a collaboration with investigators from several other groups (including four of the five schizophrenia-oriented MH-CRC's) which we are hoping will become the new "industry standard." It is intended to combine the efficiency of the BPRS and PANSS with the coverage of the SANS and SAPS, and to incorporate our current understanding of the core symptom complexes in schizophrenia.

The Longitudinal Course of Recent Onset Psychotic Disorders and Predictors of Outcome

Our prospective longitudinal study has been active now for ten years. We have enrolled a total of 325 subjects and are currently following nearly 260. The details of the study are outlined in Project 1.

Clinical Presentation

• The symptoms of schizophrenia consistently cluster in three dimensions, as determined by several factor analytic studies: psychoticism, disorganization, and negative symptoms.
• Negative symptoms are present early in the course of illness and, once present, tend to persist.
• The psychotic and disorganized dimensions, although prominent at intake, tend to decline over the course of the follow-up period and to be less stable.
• A longitudinal factor analysis of the three dimensions indicated that all three groups of symptoms tended to change in unison and independently from one another, suggesting that the three dimensions have different patterns of exacerbation and remission.
• Depressive symptoms are common early in the course of schizophrenia, and the clinical distinctions between schizophrenia and schizoaffective disorder are difficult; an algorithmic method for making a DSM diagnosis of MDD was developed and applied to CASH data, and this approach (which bypasses clinical judgment and ignores hierarchies) classified approximately 30% of patients as having Major Depressive Disorder (MDD); depressive symptoms at intake did not predict outcome, however.
• Substance abuse is uncommon in our sample; at intake only 7.7% had a substance use disorder (primarily alcohol or marijuana).

Cognition

• Patients assessed during their first episode already display significant cognitive deficit compared to controls
• Cognitive function, as assessed by a standard battery of cognitive and neuropsychological tests, either remains stable over a two year follow-up period or shows a modest improvement in some tests involving response inhibition or attention.
• When the follow-up period is extended to five years, the same pattern of stability or modest improvement remains, suggesting that cognitive deterioration does not occur during the period after the initial onset of schizophrenia.

Course and Outcome

• Various approaches to assessing outcome indicate that the majority of patients remain significantly impaired; despite significant symptomatic improvement in some patients, the majority have marked impairment of quality of life throughout the follow-up period, in spite of generally state-of-the-art treatment with typical neuroleptics.
• Symptom severity at discharge (for all three dimensions) tends to be predictive of symptom severity at follow-up, although overall severity does tend to decline over time (arguing against a global deteriorative process).
• Severity of negative symptoms at intake and at discharge is the most powerful predictor of symptomatic outcome; patients with prominent negative symptoms at intake continued to have more severe negative symptoms over the follow-up period, while psychotic and disorganized symptoms tended to decline.
• Severity of negative symptoms at intake is also a powerful predictor of psychosocial outcome, having high correlations with poor quality of life measures (occupational impairment, recreation impairment, relationship impairment), overall psychosocial functioning, and Global Assessment Scale (GAS).
• Suicide is a serious risk in this sample of young patients, with twelve subjects completing suicide and more than 20% making serious attempts; predictors of high suicide risk include depressive symptoms (and specifically hopelessness and worthlessness rather than depressed mood) and more severe neuroleptic side effects (specifically akathisia); this clearly has important treatment implications, particularly for the use of atypical neuroleptics.
• The rate of active substance abuse declined over the 60 month follow-up interval, and was typically in the 1% range; substance abuse at intake was not predictive of outcome.
• Although patients improved symptomatically over the follow-up interval, nonetheless the majority were in need of public assistance; by one year after first hospitalization, 56.3% were supported by social service agencies; once support was initiated, it was maintained over the follow-up period.
• While 34% of patients with clinically diagnosed schizophrenia also met algorithmic criteria for MDD, this percentage dropped to 6% at five-year follow-up, and having a depressive syndrome at intake did not make any meaningful predictions concerning outcome.

Brain Morphology

• Structural brain abnormalities are present shortly after the time of onset in first episode patients, suggesting that the brain injury antedates the onset of symptoms.
• Structural brain abnormalities observed in first episode patients are similar to those that are seen in more chronic patients, including increased surface and ventricular CSF, decreased frontal size, and changes in the thalamus.
• Symptom dimensions have some association with abnormalities in brain morphology; psychoticism (and specifically severity of hallucinations) is associated with decreased temporal size (and specifically decreased size of superior temporal gyrus), while both negative symptoms and psychoticism were associated with increased third ventricular CSF.
• Decreased cerebellar size is the most robust predictor of long-term symptomatic outcome, particularly severity of negative symptoms.
• Basal ganglia volume increases in size over a two-year follow-up, and the increase in size is correlated with extent of neuroleptic exposure, as measured by "dose years."
• This effect may be limited to "typical neuroleptics," perhaps as a consequence of their potent D2 blockade, since atypical neuroleptics are correlated with a modest decrease in basal ganglia size over a two year period.
• Based on a subsample of 127 subjects for whom we have obtained both baseline and at least one repeat MR scan at intervals ranging from two to nine years, there is no structural evidence for a neurodegenerative process; correlations for ventricular volume between time 1 and time 2 are >.9.

Functional Studies

• SPECT studies using the Tower of London have shown that "hypofrontality" is present in first episode patients.
• PET studies in a sample of neuroleptic-naive patients studied during REST (random episodic silent thought) have shown decreased frontal flow (as compared to healthy volunteers) in three separate frontal regions: dorsolateral, medial, and orbital; flow is increased in thalamic and cerebellar regions.
• PET studies that examine the correlation between rCBF and severity of symptoms in medication free patients assessed at the time of scanning find a pattern of association between negative symptoms and decreased flow in the fronto-pontine-cerebellar network.

Back to Top

Findings Supportive of the Cognitive Dysmetria Construct

One of our current working models of schizophrenia posits that the disease is a single illness with a single phenotype and that the phenotype is defined by the fundamental cognitive abnormality "cognitive dysmetria." Furthermore, we hypothesize that the substrate of cognitive coordination is the cortico-cerebellar-thalamic-cortical circuit (CCTCC). Therefore, schizophrenia may be conceptualized as a disease that is characterized by poor coordination, or dysmetria, in all domains of functioning, including both motor and cognitive.

Our sample of first episode patients have been utilized to evaluate several aspects of cognitive dysmetria including a structural study of some regions of the involved network (i.e., the frontal lobe), evaluation of motor abnormalities, and functional studies with PET of the CCTCC. Because neuroleptic medications have a significant effect on motor function, our sample of first-episode neuroleptic naive patients have been an informative group to study dysmetria in. Our findings with this sample have shown:

• First episode patients do indeed have structural abnormalities in regions of the CCTCC, in particular the frontal lobe.
• First episode patients also have significant motor abnormalities (neurological soft signs) and abnormal developmental reflexes.
• A PET study of our first episode patients showed abnormal blood flow during rest to not only the frontal cortex, but also in the thalamus and cerebellum, three nodes of significant importance in the CCTCC .

We have recently developed a battery of cognitive tests designed specifically to evaluate more thoroughly abnormalities in cognitive dysmetria. In particular, Project 2 outlines a proposed extension of our prospective longitudinal study whereby we enroll and follow first episode patients with a specific focus on assessment of both the structure and function of the circuitry we believe to be disrupted in schizophrenia.

Back to Top

Findings Supportive of Neurodevelopmental Mechanisms

Our investigation into the developmental processes involved in schizophrenia includes a variety of approaches: study of premorbid abnormalities suggesting manifestation of illness in childhood; evaluation of brain morphology and function in first-episode patients; study of incidence of brain anomalies that are clearly developmental in origin; and longitudinal evaluation of brain morphology over time to assess possible "degenerative" processes. Summary of our work is as follows:

• Patients with schizophrenia have an increased incidence of midline neural abnormalities, including cavum septi pellucidi, partial callosal agenesis, and gray matter heterotopias.
• Never medicated first episode patients have an increased rate of neurological soft signs and abnormal developmental reflexes.
• Never medicated and first episode patients display abnormalities assessed with both MR and PET (increased surface and ventricular CSF, decreased frontal size and perfusion).
• Patients with schizophrenia are shorter in stature compared to controls, and those patients who are smallest in stature have the poorest premorbid function, both cognitively and behaviorally. This suggests a global deficit in development.
• The male:female ratio is this sample is approximately 4:1, suggesting that males are more frequently affected when the disorder is assessed by recruiting a clinically-based sample; in this predominantly male sample, the males tend to have more severe indicators of neurodevelopmental injury (e.g., MR abnormalities, soft signs).
• Over time, the vast majority of patients show no significant change in ventricular volume suggesting the absence of any "degenerative" brain processes early on in the course of the illness.

Back to Top

Development and Revision of Assessment and Diagnostic Instruments

Our work in instrument development over the past few years has included: the development of a new instrument to assess psychopathology in schizophrenia; evaluation of the comparative validity of information from different sources of information; development and validation of computer algorithms to generate diagnoses directly from CASH items according to a variety of diagnostic criteria; and studies of the reliability and validity of distinguishing primary vs. secondary negative symptoms. A detailed summary of the new instrument project is provided in Project 6: Findings from the other efforts include:

• Patients with psychotic disorders significantly under-report both positive and negative symptoms relative to objective ratings.
• Family members appear to be more valid informants regarding patients' negative symptomatology than the patients themselves, but like the patients, under-report positive symptoms.
• The DSM-IV diagnosis of schizophrenia is concordant with DSM-III and III-R in ~90% of cases. Approximately half of the discordance (5%) in diagnosis is accounted for by more limited inclusion in DSM-IV of patients with relatively brief psychotic symptoms, and the remaining half, by greater inclusion of patients with less prominent psychotic symptoms, but persistent disorganized and negative symptoms.
• The DSM-IV diagnosis of schizophrenia is concordant with ICD-10 diagnosis in ~ 75% of cases. Much of the discordance is accounted for by differences in the boundary with schizoaffective disorder, which is diagnosed more in ICD-10, than in any of the DSM criteria.
• Reliability of distinguishing primary vs. secondary negative symptoms in the absence of specific training and a structured instrument appears to be problematic, with median Kappas of .50 and .38 for inter-rater and test-retest reliability respectively.
• Raters indicated that they felt they had adequate information to make the primary vs. secondary negative distinction based only on cross sectional evaluation in all but 10% of the cases.
• Among patients with schizophrenia, negative symptoms were rated as "primary" about twice as often as "secondary."

Back to Top

© The University of Iowa 2005. All rights reserved.

Latest update May 7, 2005 Webpages maintained by Hans J. Johnson. E-mail the webmaster