Sources of heterogeneity in schizophrenia: the role of neuropsychological functioning.

Although schizophrenia is often characterized as a heterogeneous disorder, efforts to validate stable and meaningful subtypes have met with limited success. Thus, the issue of whether schizophrenia reflects a continuum of severity or a number of discrete subtypes remains controversial. This review evaluates efforts to establish subtypes based upon a model that includes causes, characteristics, and course and outcomes of heterogeneity. Emphasis is placed on empirical classification studies utilizing cognitive tests or symptom rating scales, sometimes in conjunction with neuroimaging procedures. Results of recent cluster analytic studies are reviewed that produced evidence of four or five clusters, varying in level and pattern of performance. Although this research typically generated meaningful subtypes, it was often the case that there was little correspondence between subtyping systems based upon cognitive function and those based upon symptom profile. It was concluded that there may be different mechanisms for producing cognitive and symptomatic heterogeneity, and that diversity in presentations of schizophrenia reflects a combination of continuities in severity of the disorder with a number of meaningful and stable subtypes.

Differential neuropsychological patterns of frontal- and temporal-lobe dysfunction in patients with schizophrenia.

The frontal and temporal lobes have been implicated as pathogenic sites for schizophrenia, although there is a marked heterogeneity of brain function and structure between individual patients. It is currently unclear whether some patients with schizophrenia exhibit primarily frontal lobe dysfunction, while others exhibit primarily temporal-lobe dysfunction. The current investigation examined this issue in a preliminary way by using neurocognitive tests to discriminate test performances of patients with schizophrenia from patients without schizophrenia who had definitive neurological evidence of either frontal- or temporal-lobe dysfunction. Of the patients with schizophrenia, 20.7% were classified as having a frontal lobe dysfunction profile, while 19.3% had a temporal lobe dysfunction profile. Results further clarify neurobiological heterogeneity in schizophrenia by demonstrating that a substantial number of patients with schizophrenia exhibit either primarily frontal- or temporal-lobe dysfunction. Results may partially explain the inadequacy of neurobiological models for schizophrenia that do not consider these differential patterns of dysfunction.

Confirmation of a two-factor model of premorbid adjustment in males with schizophrenia.

Because schizophrenia is considered to be a neurodevelopmental disorder, premorbid adjustment is of particular interest. Premorbid adjustment is probably not a unitary construct but rather is expressed across a number of developmental domains. The current investigation examined the validity of a two-factor model that differentiated premorbid adjustment across social and academic domains and evaluated relationships between these premorbid adjustment domains and other variables of interest. Participants with schizophrenia (n = 141) underwent evaluation of premorbid adjustment (using the Premorbid Adjustment Scale), intellectual functioning, and psychiatric symptoms. Using confirmatory factor analysis, a two-factor model of premorbid adjustment was identified that included an academic domain and a social domain. The social domain was associated with symptom variables, while the academic domain was associated with measures of intelligence. Results provide evidence for at least two domains of premorbid adjustment in schizophrenia. Distinguishing between these two premorbid domains may be theoretically important because of potential differences in incidence rates and deterioration courses; some individuals with schizophrenia may exhibit adequate academic adjustment but poor social adjustment, while others may exhibit the opposite pattern.

Factor structure of neurologic examination abnormalities in unmedicated schizophrenia.

The heterogeneity and uncertain significance of neurologic exam abnormalities in schizophrenia prompted us to evaluate their factor structure. We administered a modified version of the Neurological Evaluation Scale (NES) to 103 unmedicated patients with schizophrenia. Data were distilled by combining right- and left-side scores, and by eliminating superfluous, rarely abnormal and unreliable items from the analysis. Exploratory principal components analysis yielded four factors: repetitive motor tasks (fist-ring, fist-edge-palm, alternating fist-palm, dysdiadochokinesis); cognitive-perceptual tasks (memory, audiovisual integration, right-left orientation, face-hand test, rhythm tapping reproduction); balancing tasks (Romberg, tandem gait); and the palmomental reflex. Evaluation of the relationship between these factors and clinical and demographic variables revealed a robust correlation between the cognitive-perceptual factor and full-scale IQ score. This analysis is a step toward developing empirical subscales of a modified NES, which may provide insights into the nature of neurologic impairment in schizophrenia and may prove clinically useful.

Neurologic examination abnormalities in schizophrenia with and without a history of alcoholism.

OBJECTIVE: The current investigation examines the impact of a past history of alcoholism on neurologic examination abnormalities in schizophrenia (SZ). BACKGROUND: Individuals with SZ have a high rate of comorbid alcohol use disorders (AUDs), but relatively little is known about the potential adverse consequences of alcoholism for neuropsychological and neurologic functioning in SZ. Recent evidence suggests consistent but subtle neurocognitive differences between groups, with more prominent differences in neurologic examination abnormalities. METHOD: Thirty-three male patients with SZ or SZ/AUDs were evaluated using a modified Neurologic Evaluation Scale (NES) and ratings for positive and negative symptoms. RESULTS: The SZ/AUD group exhibited a greater impairment in the Cognitive-Perceptual factor of the Neurologic Evaluation Scale. Greater impairment in the tandem-Romberg factor or in motor items was not found, nor were groups different based on positive or negative symptoms. CONCLUSIONS: A history of alcoholism in SZ is associated with greater overall neurologic impairment, particularly in the area of cognitive-perceptual dysfunction, an area often found to be impaired in patients with schizophrenia without alcoholism.

Influence of haloperidol on the relationship of frontal lobe function to psychomotor poverty and disorganization syndromes.

The aim of the study was to examine effects of haloperidol on the relationships between neuropsychological measures of frontal lobe functioning and the schizophrenia syndromes of psychomotor poverty and disorganization. Twenty-one participants with schizophrenia were initially evaluated when clinically stable and chronically treated with haloperidol, and 19 were evaluated again after a 3-week haloperidol-free period. Participants were evaluated with the Trail Making Test, the Wisconsin Card Sorting Test, the Purdue Pegboard, and psychiatric rating scales at each evaluation. There were significant correlations between schizophrenia syndromes and the tests sensitive to frontal lobe function when participants were medicated but not when drug-free. No significant changes in symptom severity or motor function occurred from the medication to the medication-free evaluation. The results indicate that haloperidol mediates the relationship between tests sensitive to frontal lobe function and the schizophrenia syndromes of psychomotor poverty and disorganization. This mediation effect was not attributable to changes in overall symptom severity or motor function.

Is the Halstead Category Test a multidimensional instrument?

Factor structure of the Halstead Category Test was evaluated in patients with schizophrenia, heterogeneous forms of brain damage, and patient controls using confirmatory factor analysis. Analyses were performed including and excluding subtests 1 and 2. In the first analysis, a three-factor model was optimal, with subtests 1 and 2 loading on one factor (Counting), 3, 4, and 7 loading on a second factor (Spatial Positional Reasoning), and subtests 5 and 6 loading on a third factor (Proportional Reasoning). Excluding subtests 1 and 2, a two-factor solution was optimal consisting of the Spatial Positional (subtests 3 and 4) and Proportional Reasoning (subtests 5 and 6) factors, with subtest 7 loading on both factors. Optimal factor structures for the three groups were identical. Correlations between factor scores were similar among groups. Factor scores also correlated significantly (p <.01 ) with all of the other cognitive measures. It was concluded that the Category Test is a multidimensional procedure with factors associated in a general way with other cognitive abilities.

Relations between cognitive and symptom profile heterogeneity in schizophrenia.

Although numerous studies have consistently revealed cognitive heterogeneity in schizophrenia, the relationships between such heterogeneity and clinical phenomenology are not clear. Clusters derived from cognitive heterogeneity studies may or may not be associated with symptom profile or severity of illness. The purpose of this study was to examine the relationship between cognitive heterogeneity and demographic and clinical phenomenological measures. We examined cognitive heterogeneity in schizophrenia by empirically deriving clusters of patients based upon WAIS-R subtest scores and then analyzed the way in which these clusters related to demographic and symptom variables and to DSM-III-R diagnostic subtypes. Four cognitive clusters were identified that were consistent with previous research. These clusters were differentiated on the basis of educational level and occupational status but not on the basis of symptom profile, severity, or DSM-III-R subtypes. Results suggest that cognitive measures are independent of severity of the disorder and phenomenological symptom presentation in these subgroups of schizophrenic patients.

Neurocognitive dysfunction in patients diagnosed with schizophrenia and alcoholism.

The authors administered the Halstead-Reitan Neuropsychological Test Battery to schizophrenic groups with (n = 54) and without (n = 217) coexisting alcoholism, nonschizophrenic groups with alcoholism (n = 231), and a patient comparison group (n = 145) to determine the extent of additive cognitive impairment in schizophrenia associated with alcoholism and to compare cognitive function in alcoholism and schizophrenia. The additive effects of alcoholism on cognitive dysfunction in schizophrenia were subtle but were consistently identifiable. Cognitive dysfunction in alcoholism was less severe than in schizophrenia with or without alcoholism. The magnitude of additive effects of alcoholism on cognitive dysfunction in schizophrenia was age related with a significant interaction between age and presence or absence of alcoholism on a global index of cognitive dysfunction.

Empirical validation of primary negative symptoms: independence from effects of medication and psychosis.

OBJECTIVE: Recent studies of negative symptoms in schizophrenia-specifically, those involving the deficit syndrome-have focused on uncovering the symptoms that are primary to the disease rather than secondary to the psychotic process. One of the foremost concerns in this effort is establishing whether the negative symptoms observed are the result of medication effects. METHOD: This study used negative symptom ratings obtained in a drug withdrawal paradigm to compare symptom profiles in the same schizophrenic patients when they were on and off antipsychotic drug treatment. The study group consisted of 93 physically healthy male patients with DSM-III-R-defined schizophrenia. Principal components analysis was performed on negative symptom data obtained separately during haloperidol treatment and again when the patients were drug free to determine whether there were meaningful factor scores that were consistent across medication conditions. Drug withdrawal effects on negative symptom factors were then tested for associations with secondary sources of variance including extrapyramidal side effects, anxiety/depression, and psychosis. RESULTS: Two factors, termed affective flattening and diminished motivation, exhibited similar loadings when the patients were both on and off medication. Changes in motivation were associated with changes in anxiety/depression and psychosis, while changes in affective flattening were associated with changes in extrapyramidal side effects. CONCLUSIONS: The documented secondary sources of negative symptoms are related to different and distinct aspects of negative symptoms; this finding will aid in the identification of primary negative symptoms.

Lateralized brain dysfunction in schizophrenia: a comparison with patients with lateralized structural lesions.

A comparison was made among participants with schizophrenia and those with structural lateralized or diffuse brain damage in order to determine the extent to which the cognitive profile of the schizophrenia sample resembled the profiles obtained from patients with left-hemisphere, right-hemisphere, and diffuse brain damage. The Halstead-Reitan Neuropsychological Battery was used as the testing procedure. The data were subjected to discriminant analysis in order to obtain frequencies of predicted classification of the participants with schizophrenia into schizophrenia, left-hemisphere, right-hemisphere, and diffuse groups. Half of the participants with schizophrenia were classified into the schizophrenia group. The other half was evenly distributed across the left-hemisphere, right-hemisphere, and diffuse brain damage groups. There was not a disproportionately large number of participants classified into the left-hemisphere group. Comparisons among these four predicted groups were accomplished for each of the Halstead-Reitan Battery measures using one-way analysis of variance. The comparison of the subtest scores among the predicted groups indicated that the patients classified into the left-hemisphere group were characterized by a pattern of language dysfunction thought to be developmental in nature, and an abnormal lack of asymmetry in tapping speed favoring the right hand.

Construct validity of neuropsychological tests in schizophrenia.

Validity studies of neuropsychological tests have typically examined individuals with neurological disorders. The present study was designed to investigate the construct validity of neuropsychological measures in patients with schizophrenia. We used Wechsler Adult Intelligence Scale Revised (WAIS-R) factor scores that were generated from the population of interest as marker variables in the present analysis. The current study included 39 patients with schizophrenia who were evaluated with a battery of neuropsychological tests assessing attention, memory, and abstract reasoning abilities. Pearson correlations indicated significant relationships between (a) WAIS-R Verbal Comprehension factor and tests of sustained attention, verbal memory and remote memory; (b) WAIS-R Perceptual Organization factor and tests of visual memory and abstraction and problem solving; and (c) WAIS-R Freedom From Distractibility factor and neuropsychological measures of attention and concentration. These results provide support for the construct validity of the neuropsychological tests in patients with schizophrenia, and indicate that these tests evaluate essentially the same constructs in patients with schizophrenia as they do for patients with structural neurological disorders.

Confirmatory factor analysis of the WAIS-R in patients with schizophrenia.

Although factor scores are commonly used to interpret the Weschsler Adult Intelligence Scale--Revised (WAIS-R), the WAIS-R factor structure has not been investigated in patients with schizophrenia. We used confirmatory factor analysis (CFA) to examine five latent construct models in 169 males with schizophrenia. The WAIS-R standardization sample (ages 35-44; n = 250) was used as a comparison group. For both groups, all model fit indexes used to determine model adequacy supported models composed of Verbal Comprehension (VC), Perceptual Organization (PO) and Freedom from Distractibility (FFD) factors. However, the Digit Symbol subtest loaded on both the PO and FFD factors for patients with schizophrenia but only on the FFD factor for the WAIS-R standardization sample. Patients with schizophrenia performed significantly worse on the FFD and PO factors compared to the VC factor, reflecting the well-characterized attention and problem solving deficits associated with schizophrenia. Also, patients with schizophrenia performed significantly worse than the WAIS-R sample on all factors. These results provide support for the validity of the WAIS-R factors in patients with schizophrenia.

Individual differences in cognitive decline in schizophrenia.

OBJECTIVE: The authors' goal was to determine whether cognitively impaired patients with schizophrenia exhibit age-related cognitive declines similar to those of patients with schizophrenia who do not have substantial cognitive impairment. METHOD: Correlation coefficients were computed between age and the Average Impairment Rating, a summary index of cognitive ability, in a group of 77 patients with schizophrenia. These patients were clustered into two groups: one with near-normal cognitive function (N=51) and one with severely impaired cognitive function (N=26). A group of patients with senile dementia (N=21) and another comparison group of nonschizophrenic patients (N=299) were used as reference groups. RESULTS: There were significant correlations between age and the Average Impairment Rating in all groups except the cognitively impaired patients with schizophrenia, in which a zero-order correlation was obtained. CONCLUSIONS: Patients with schizophrenia who have substantial cognitive impairment do not have the significant correlation between age and cognitive function found in patients with schizophrenia who have mildly impaired or normal cognitive abilities, suggesting earlier onset of cognitive deficit in the cognitively impaired patients with schizophrenia.

A comparison of clustering solutions for cognitive heterogeneity in schizophrenia.

A cluster analytic solution based upon a battery of tests consisting of the Halstead Category and Tactual Performance Tests, the Trail Making Test, and the Wisconsin Card Sorting Test was compared with a solution based on the subtests of the Wechsler intelligence scales, utilizing a sample of 221 schizophrenic patients. Both analyses permitted four-cluster solutions, and we found a weak but significant degree of association between solutions. Examination of external validity of the two solutions revealed stronger associations with clinical variables for the Wechsler-scale-based solution. The major conclusions were that the existence of cognitive heterogeneity in schizophrenia exists across a broad range of abilities, and appears to reflect a combination of continuity of ability level and existence of possible subtypes requiring further neuropsychological and neurobiological verification.

Teaching memory strategies to persons with multiple sclerosis.

A memory-training program previously used effectively upon persons with head-injury (HI) was conducted upon eight subjects with multiple sclerosis (MS). The program involved computer-assisted teaching of imagery-based mnemonic strategies for recall of lengthy lists of words, and for associating names with faces. Results were similar to those found in individuals with HI, but the MS subjects learned the strategies quickly, and did not appear to require the lengthy training needed by persons with HI. It was concluded that memory training of those with MS may sometimes only require teaching of mnemonic strategies without extensive practice.

Neuropsychological similarities and differences among Huntington's disease, multiple sclerosis, and cortical dementia.

A comparison of cognitive function was made among patients with Huntington's disease, multiple sclerosis, and cortical dementia. Utilizing indexes from the Wechsler Adult Intelligence Scale and the Halstead-Reitan Battery, it was found that there was substantially more severe cognitive deficit in the Huntington's disease patients than in the multiple sclerosis patients, and the level of impairment was similar between the Huntington's disease and cortical dementia groups. Qualitative differences, particularly involving amount and type of perseveration, were noted among the three groups. It was concluded that subcortical dementia is not necessarily characterized by mild cognitive impairment, and there appear to be important qualitative differences between cortical and subcortical dementia. Results are discussed in terms of the usefulness of the presently conceptualized distinction between cortical and subcortical dementia.

Utility of WAIS-R short forms in schizophrenia.

Recently, short forms of the Wechsler Adult Intelligence Scale-Revised (Wechsler, 1981) have received increasing attention because of their ability to provide estimated IQ scores with substantial time savings (in some cases 85-90% savings). These short forms may have particular utility for individuals with schizophrenia because they require less time to administer and, as a result, are less taxing for these patients who often exhibit impaired attention and deficient motivation. In this study, we examine the psychometric properties of nine popular WAIS-R short forms in a group of 143 patients diagnosed with schizophrenia. Our results indicated that Kaufman's four subtest short form was the best overall estimator of Full Scale IQ (FSIQ) when a combination of administration time and psychometric properties were considered. However, Ward's seven subtest short form provided the closest estimation of FSIQ and had the lowest misclassification rate, while also providing estimates of Verbal and Performance IQs and yielding 46.5-49.7% time savings. All short forms had substantial misclassification rates, indicating that caution is warranted when using these forms to classify individuals according to standard levels of intellectual functioning (e.g., Average, Low Average, High Average). Clearly, the main consideration in selecting a short form is whether time savings or accuracy have priority.

Chronic haloperidol treatment does not affect structure of attention in schizophrenia.

Results of a number of investigations indicate attention is a multifactorial construct composed of four distinct cognitive factors including focus-execute, sustain, encode and shift abilities. While investigators have partially or fully replicated this attentional structure in a number of clinical and nonclinical populations, no study has adequately examined the structure of attention in patients with schizophrenia who are not treated with antipsychotics. In this study, we examined the four-factor theory of attention in patients with schizophrenia while they were stabilized on haloperidol (with no adjunctive antiparkinsonian/anticholinergic medications) and again when they were approximately 3 weeks drug free. Standard neuropsychological measures were used to assess attentional functions. Principal components analyses (varimax rotation) of neuropsychological test scores in medicated and drug-free conditions indicated that four factors accounted for 84.2 and 91.8 of total variance in medicated and unmedicated conditions, respectively. Based on these results, it appears that: (1) haloperidol does not appreciably affect structure of the attentional system in patients with schizophrenia; (2) unmedicated patients with schizophrenia exhibit a similar structure of attention as both medicated patients and controls, suggesting that attentional structure is 'normal' in schizophrenia; and (3) the four-factor attention theory is a useful and valid paradigm for evaluating attention in patients with schizophrenia, regardless of medication status.