Prediction of Neuropsychological Performance by Neurological ...


Published on

  • Be the first to comment

  • Be the first to like this

No Downloads
Total views
On SlideShare
From Embeds
Number of Embeds
Embeds 0
No embeds

No notes for slide

Prediction of Neuropsychological Performance by Neurological ...

  1. 1. Prediction of Neuropsychological Performance by Neurological Signs in Schizophrenia Celso Arango, M.D., John J. Bartko, Ph.D., James M. Gold, Ph.D., and Robert W. Buchanan, M.D. Objective: The major purposes of this study were 1) to examine whether neurological signs predict cognitive performance in both schizophrenic patients and healthy subjects and 2) to determine the ability of neurological signs and neuropsychological tests to dis- criminate schizophrenic patients from healthy subjects. Method: Eighty-five patients with a DSM-III-R diagnosis of schizophrenia and 36 normal comparison subjects were included in the study. All subjects were administered a comprehensive neuropsychological test bat- tery, and neurological signs were assessed with the Neurological Evaluation Scale. Step- wise regression analyses were used to predict neuropsychological test performance from the subscale scores on the Neurological Evaluation Scale. Forward stepwise linear dis- criminant function analyses were used to examine the discriminative ability of neurological subscale scores, neuropsychological test scores, and the two combined. Results: Scores on the Neurological Evaluation Scale predicted the neuropsychological test performance of both patients and comparison subjects. The sensory integration subscale score was the most frequent predictor of neuropsychological test performance. In contrast, the “others” subscale, which includes frontal release signs, abnormalities in eye movements, and short-term memory, was the most highly discriminating subscale, correctly classifying 78.5% of the total study group. The best predictors from the neuropsychological battery (category fluency and Trail Making Test, part A, time test) correctly classified 81.8%. When both sets of variables were used, the Neurological Evaluation Scale “others” subscale en- tered the discriminant function first. Conclusions: Neurological signs are reliably related to measures of neuropsychological performance and also reliably discriminate between patients and healthy subjects. However, some neurological signs may be more sensitive to the presence of schizophrenia, while others may be more predictive of neuropsychological performance. (Am J Psychiatry 1999; 156:1349–1357) P atients with schizophrenia are characterized by the presence of neurological abnormalities that are not spe- able to a specific brain area, while the latter are consid- ered nonspecific and nonlocalizing. Schizophrenic pa- cific to the disease but are more prevalent in such pa- tients have been shown to have a greater prevalence of tients than in persons with other mental illnesses and both soft and hard signs when compared with normal control subjects (1). These differences do not appear to normal control subjects (1). These neurological abnor- be attributable to either medication status or the pres- malities classically have been divided into “hard” signs ence of medication-induced neurological side effects and “soft” signs. The former are signs that are localiz- (e.g., tardive dyskinesia, extrapyramidal symptoms). The significance of neurological abnormalities in Presented in an earlier version at the Winter Workshop on Schizophrenia, Davos, Switzerland, Feb. 7–13, 1998. Received schizophrenia is unclear. It has been argued that they June 12, 1998; revision received Feb. 18, 1999; accepted Feb. 23, represent a trait characteristic of the disorder, and they 1999. From the Maryland Psychiatric Research Center, University have been observed to be associated with positive fam- of Maryland School of Medicine. Address reprint requests to Dr. ily history (1). Genetic (2), environmental (3), or in- Buchanan, Maryland Psychiatric Research Center, P.O. Box 21247, trauterine or perinatal (4) etiologies have been pro- Baltimore, MD 21228; (e-mail). Supported in part by NIMH grants MH-48225 and MH-40279 and posed for these signs. It is possible, then, that they are grant 97/5091 from Fondo de Investigación Sanitaria, Ministerio de indicators of prior brain insults/injuries undetected in Sanidad y Consumo, Spain. earlier life that predispose to schizophrenia or are Am J Psychiatry 156:9, September 1999 1349
  2. 2. PREDICTION OF NEUROPSYCHOLOGICAL PERFORMANCE modifying factors for the outcome of the disease. An METHOD approach for examining their significance is to study the correlations of neurological signs with other illness All patients attending the Maryland Psychiatric Research Center Outpatient Research Program who met the DSM-III-R criteria for manifestations. In previous studies, these signs have chronic schizophrenia and had both a neurological and a neuropsy- been correlated with negative or deficit symptoms (5– chological assessment performed within 3 months of each other were 7), the disorganized subtype (8), poorer outcome (9), included in the study. Eighty-five patients were selected. All available lower psychosocial performance (10), earlier age at on- sources of information were used for making the diagnosis, including the Structured Clinical Interview for DSM-III-R—Patient Version set (11), and repeated episodes of violence (12). (23), direct assessment, family informants, and past medical records. The relation between neurological signs and another Patients with an organic brain disorder, mental retardation, or a his- important disease manifestation, cognitive impair- tory of drug abuse/dependence were excluded from the study. ment, has not received much attention in the past. Thirty-six normal comparison subjects were recruited from the gen- eral population through advertisements. The comparison subjects did Most but not all previous studies have found a correla- not have past or current DSM-III-R axis I or axis II disorders, as deter- tion between general cognitive dysfunction and neuro- mined by the Structured Clinical Interview for DSM-III-R, or a history logical signs (13–16). Neurological signs have been of organic brain disorder, mental retardation, or severe head trauma. correlated with markers of organicity such as lower IQ Patients and comparison subjects gave written informed consent before examination, and the comparison subjects were reimbursed (17, 18) and impaired performance on the Mini-Men- for participation in the study. tal State examination (11, 17, 19). Cuesta et al. (20) re- The patients were interviewed while clinically stable. The major- ported that seven “frontal” neurological signs corre- ity of the patients were receiving conventional antipsychotics (either fluphenazine or haloperidol) at the time of testing; two patients were lated with poor cognitive testing performance. drug free, and one patient was taking clozapine. The Brief Psychiat- Two recent studies, however, have pointed to a more ric Rating Scale (BPRS) (24) was used to measure symptoms. Dyski- selective correlation between neurological signs and neu- netic and parkinsonian movements were assessed with the Maryland ropsychological impairment. Mohr et al. (21), using the Psychiatric Research Center Involuntary Movement Scale (25). Mean neuroleptic doses for each group were determined with the use Neurological Evaluation Scale, studied neurological of the conversion system developed by Schooler (26). Anticholin- signs in a group of 93 patients with schizophrenia. Neu- ergic medication was rated as present or not. rological Evaluation Scale total score and subscale scores Neurological signs were evaluated with the Neurological Evalua- correlated with scores on the Wisconsin Card Sorting tion Scale (27). The scale comprises 26 items designed to assess three functional areas of interest: sensory integration, motor coordination, Test and the Weigl Sorting Test (measures thought to and sequencing of complex motor acts. In addition, short-term represent frontal lobe function) after control for perfor- memory, frontal release signs, and eye movement abnormalities are mance on the Raven Standard Progressive Matrices. included in an “others” category. The items are presented in Correlations were highest for the sequencing of complex appendix 1. The total score and scores for each of the four subscales were used in the study to measure severity of neurological impairment. motor acts subscale of the Neurological Evaluation The interrater reliability (intraclass correlation coefficient) for the sub- Scale. The authors suggested that this relationship was scale scores and total score ranged from 0.63 (for motor coordination) suggestive of prefrontal cortical dysfunction. to 0.99 (for sensory integration). Total time for the administration of the Neurological Evaluation Scale was approximately 45 minutes. In the second study, Flashman et al. (22) assessed 22 The neuropsychological test battery (28–37) is presented in neurological signs, including six developmental reflexes. appendix 2. Neuropsychological tests were administered according Patients with neurological signs (N=68) and those with- to standardized testing procedures. The Mini-Mental State examina- out (N=108) were compared. The only neuropsycholog- tion (38) was administered before the Neurological Evaluation Scale. All other tests were administered in a fixed rather than a ran- ical tests that discriminated between the two groups af- dom order to ensure that more difficult tasks did not occur consecu- ter control for lifetime medication exposure, tardive tively. Frequent breaks were allowed, and if necessary, tests were ad- dyskinesia, and extrapyramidal symptoms were those ministered in two test sessions. Not all patients and comparison involved with motor speed and motor coordination subjects completed all measures in the battery, so the number of pa- tients and comparison subjects varies from test to test. (e.g., finger tapping, the Purdue pegboard task, and the In the statistical analysis, the distribution of each variable was ex- Trail Making Test, part B). Tests of general cognitive amined with the Kolmogorov-Smirnov test to determine whether it function, attention, frontal lobe function, and memory was normally distributed. Scores on the Trail Making Test, parts A failed to distinguish between the two groups. and B (measured as time to complete the test in seconds), which were not normally distributed, were log transformed to normalize them. Neurological signs have therefore been related to Derived scores were used for the Trail Making Test, part B, and the both global and selective cognitive impairments. These Stroop Color and Word Test so that measures focused on the execu- discrepant results could be attributed to variability in tive function construct (37). For the Trail Making Test, we measured the difference between each subject’s actual part B score and the part the assessment of neurological signs and neuropsycho- B score estimated by the regression of the log-transformed part A logical tests, not using validated standardized scales, time on the log-transformed part B time (39). The regression equa- differential attention to potential confounding vari- tion was computed using the data from the comparison subjects. A ables, and/or population differences among studies. similar approach was used to estimate the residual score for the Stroop color-word measure. The residual score was the difference The current study was an attempt to clarify previous between the observed color-word score and the score predicted by findings by using a validated neurological scale and a the color-word score with the color score regression in the compari- comprehensive neuropsychological test battery. A fur- son group. Raw scores were used for the individual subscales of the ther goal was to assess the ability of the Neurological Wechsler Memory Scale—Revised. Demographic variables of pa- tients and comparison subjects were compared by t tests and chi- Evaluation Scale to discriminate between normal com- square analyses. Differences in scores on the Neurological Evalua- parison subjects and patients with schizophrenia. tion Scale were also compared by t tests. 1350 Am J Psychiatry 156:9, September 1999
  3. 3. ARANGO, BARTKO, GOLD, ET AL. TABLE 1. Demographic and Clinical Characteristics of Schizophrenic Patients and Healthy Comparison Subjects Patients (N=85) Comparison Subjects (N=36) Variable N Mean SD Range N Mean SD Range Age (years) 35.89 7.70 18–61 35.13 7.40 22–48 Sex Male 60 24 Female 25 12 Race African American 21 3 White 63 32 Other 1 1 Education (years) 12.3 2.3 7–18 14.3 2.1 12–20 Socioeconomic status of head of householda 3.02 1.22 1–5 3.75 1.21 1–5 Handedness Right 73 32 Left 4 3 Ambidextrous 8 1 Age at onset of illness (years) 20.77 5.87 Duration of illness (years) 14.82 6.85 3–31 BPRS total score 36.2 10.00 19–65 Tardive dyskinesia ratingb 2.5 3.6 0–21 Parkinsonism ratingb 2.6 2.9 0–13 Neuroleptic dosec 3.0 1.0 1–4 Anticholinergic medication Yes 80 No 5 a Determined by the Hollingshead-Redlich rating of 1–5, in which 1 is the highest and 5 the lowest. Socioeconomic status could not be determined for four patients and one comparison subject. b Involuntary Movement Scale (25). c Based on a scale of 0–4, in which 0=no drug, 1=low dose, and 4=high dose. Stepwise regression analysis was used to predict neuropsychological RESULTS test performance from scores on the neurological signs subtests. Possi- ble confounding factors were based on previous studies showing that different variables relate to neurological signs (1, 14, 40–42). Separate The demographic and clinical variables of the pa- regression models were constructed for the patient and comparison tients and normal comparison subjects are presented in groups. For comparison subjects the covariates were age, sex, race, and table 1. There were no significant differences between parental socioeconomic status. To control for the potentially con- the two groups in age, sex, race, or socioeconomic sta- founding effects of ancillary variables, these variables, common to comparison subjects and patients, were entered as a block at step 1 of tus of the head of household. The comparison subjects the regression analyses. For patients the additional covariates were had significantly more years of education than the pa- BPRS total score, parkinsonism rating from the Involuntary Movement tients (t=4.46, df=119, p<0.001). Neurological Evalu- Scale, tardive dyskinesia rating from the Involuntary Movement Scale, ation Scale scores and neuropsychological test scores medication status, and anticholinergic medication. To facilitate direct of the patients and comparison subjects are reported in comparison between comparison subjects and patients, covariates for the patients were entered in two blocks, the first including the same co- table 2 and table 3, respectively. Patients had signifi- variates used for the normal comparison subjects and the second block cantly higher Neurological Evaluation Scale total including the additional patient covariates. Because the Neurological scores and scores on the sensory integration, motor co- Evaluation Scale “others” subscale includes two memory items (appen- ordination, sequencing of complex motor acts, and dix 1), the regression analysis for neuropsychological tests related to memory function were performed excluding those two items. “others” subscales. Because education significantly Finally, three forward stepwise linear discriminant function analy- differed between the groups, an analysis of covariance ses, with jackknifed classification matrices, were performed to exam- controlling for education was performed. All differ- ine the discriminant ability of neurological signs and neuropsycholog- ences remained significant. ical tests to categorize the subjects as patients or comparison subjects. Regression analyses for patients and comparison This linear discriminant function model is developed using all of the subjects minus one (N–1). The resulting linear discriminant function is subjects are reported in table 4 and table 5. For pa- used to predict the group membership of the omitted subject. The pro- tients, neurological signs subscale scores significantly cess is repeated for all subjects. This method provides for internal rep- increased the variance explained for 15 of 19 test re- lication, minimizes bias, is conservative, and ensures maximal inter- sults after control for covariates. For comparison sub- pretability (43). Separate discriminant functions were derived for the jects, one-half of the neuropsychological test results Neurological Evaluation Scale subscales and the neuropsychological tests. In the third discriminant analysis, scores on both the Neurologi- were also predicted by neurological subtest scores. cal Evaluation Scale subscales and the neuropsychological tests were Sensory integration was the Neurological Evaluation entered to find the variables that maximized correct classification. All Scale subscale score most frequently found to increase neuropsychological tests included in appendix 2 were available for en- the variance explaining the neuropsychological test re- tering in the discriminant analyses, with the exception of three (Con- tinuous Performance Test, Span of Apprehension, and Mini-Mental sults. As shown in table 4, sensory integration signifi- State examination), which were not available for some patients, and cantly predicted the results of 12 of the 19 neuropsy- the WAIS-R, which was not available for the comparison subjects. chological tests for the patients. For the comparison Am J Psychiatry 156:9, September 1999 1351
  4. 4. PREDICTION OF NEUROPSYCHOLOGICAL PERFORMANCE TABLE 2. Neurological Evaluation Scale Scores of Schizophrenic Patients and Healthy Comparison Subjects Comparison Subjects Patients (N=85) (N=36) Analysis Neurological Evaluation Scale Item Mean SD Range Mean SD Range t (df=119) p Total score 17.65 8.20 4–45 6.77 3.90 2–17 7.06 <0.001 Motor coordination subscale score 1.35 1.40 0–7 0.81 0.83 0–6 2.20 0.008 Sensory integration subscale score 3.72 2.72 0–10 1.50 1.59 0–7 4.76 <0.001 Sequencing of complex motor acts subscale score 1.88 2.70 0–12 0.53 0.88 0–6 3.47 0.001 “Others” subscale score 9.43 3.92 2–21 3.94 3.01 0–14 7.51 <0.001 TABLE 3. Neuropsychological Test Scores of Schizophrenic Patients and Healthy Comparison Subjects Score Patients Comparison Subjects Analysis Measure N Mean SD Range N Mean SD Range t df p Wisconsin Card Sorting Test perseverative errors 85 27.7 18.4 5.6 to 73.4 36 12.9 9.4 5.4 to 41.4 –4.50 119 <0.001 Verbal fluency 85 33.2 11.2 8 to 56 36 41.6 11.1 22 to 75 3.69 119 <0.001 Stroop Color and Word Test 85 –5.7 8.5 –30.7 to 9.8 36 0.00 10.3 –19.7 to 27.2 3.19 119 0.002 Trail Making Test, part A 85 44.4 18.8 19 to 122 36 24.7 8.8 11 to 52 –6.00 119 <0.001 Trail Making Test, part B 85 0.14 0.18 –0.16 to 0.64 36 0.00 0.13 –0.18 to 0.24 –4.30 119 <0.001 Category fluency 85 42.6 14.1 16 to 85 36 62.8 13.7 31 to 100 7.28 119 <0.001 Mooney faces closure 85 19.2 2.2 14 to 23 36 20.5 2.0 13 to 23 3.53 119 0.003 Benton judgment of lines 85 21.6 6.2 6 to 30 36 25.3 3.7 12 to 30 3.35 119 0.001 WAIS-R block design 85 8.7 2.6 4 to 15 36 12.1 2.6 6 to 18 6.49 119 <0.001 Wechsler Memory Scale–Revised Logical memory subscale 85 17.0 9.4 1 to 43 36 29.1 7.0 12 to 39 6.92 119 <0.001 Figural memory subscale 85 6.1 1.6 3 to 10 36 8.1 1.5 4 to 10 6.11 119 <0.001 Visual pairs subscale 85 9.6 4.5 0 to 18 36 14.6 3.9 5 to 18 5.80 119 <0.001 Verbal pairs subscale 85 16.4 4.5 4 to 24 36 20.2 2.8 16 to 24 4.73 119 <0.001 Visual reproduction subscale 85 28.0 8.2 9 to 41 36 36.5 3.3 29 to 41 6.48 119 <0.001 Continuous Performance Test 60 2.2 1.0 0.06 to 4.3 29 2.9 0.87 0.92 to 4.7 3.21 87 0.002 Span of Apprehension 74 69.6 9.9 45 to 87 35 77.5 9.1 56.3 to 91.3 4.00 107 <0.001 Mini-Mental State 82 27.6 2.1 21 to 30 21 28.9 1.1 24 to 30 3.59 101 0.008 WAIS-R verbal scale 82 14.9 5.0 6 to 28 WAIS-R performance scale 82 15.1 4.7 7 to 26 subjects (table 5) fewer neuropsychological test results analysis, with the neuropsychological tests, at step 1, were predicted by neurological subscale scores: sen- category fluency met the entry condition (F=53.3, df= sory integration scores significantly predicted the re- 1, 199, p<0.00001). This variable accounted for an sults of three of the 17 neuropsychological tests, and overall 78.5% correct jackknife classification. At step scores on sequencing of complex motor acts predicted 2, the Trail Making Test, part A (time), entered (F= the results of four of the 17 tests. Although some neu- 13.3, df=1, 119, p<0.0001). These two variables ac- ropsychological test results were predicted by more counted for an overall 81.8% correct jackknife classi- than one neurological subscale score, in most cases a fication. Individually, 80.0% of the schizophrenic pa- neuropsychological test result was predicted by perfor- tients were correctly classified, and 86.1% of the mance on only one of the neurological subtests. comparison subjects were correctly classified. None of Three forward stepwise linear discriminant function the remaining neuropsychological variables met the analyses with jackknifed classification matrices were entry criteria for the linear discriminant function. In run to examine the discriminant ability of neurological the third analysis, when both neurological and neurop- signs, the neuropsychological test battery, and both as- sychological variables were available to enter, the Neu- sessments together. In the first analysis, with neurolog- rological Evaluation Scale “others” subscale score en- ical signs, the scores on the four Neurological Evalua- tered first (F=56.5, df=1, 199, p<0.00001). At step 2 tion Scale subscales were available for entering. Only category fluency entered, at step 3 the logical memory one variable, the Neurological Evaluation Scale “oth- subscale (Wechsler Memory Scale—Revised) entered, ers” subscale score, met the inclusion criterion (F=4.0 and at step 4 the Trail Making Test, part A, entered. to enter) (F=56.5, df=1, 120, p<0.00001). This vari- Together these last three variables added only 3.3% to able accounted for an overall 78.5% correct jackknife the classification matrix. Overall, the four variables ac- classification. Individually, 75.3% of the schizophrenic counted for an 81.8% correct jackknife classification patients were correctly classified (empirical sensitiv- matrix. Individually, 91.5% of the schizophrenic pa- ity), and 86.1% of the comparison subjects were cor- tients were correctly classified, and 78.0% of the com- rectly classified (empirical specificity). In the second parison subjects were correctly classified. 1352 Am J Psychiatry 156:9, September 1999
  5. 5. ARANGO, BARTKO, GOLD, ET AL. TABLE 4. Results of Multiple Regression Analysis of Schizophrenic Patients’ Neuropsychological Test Performance and Neuro- logical Evaluation Scale Scores R2 Step Step Neurological Evaluation Final Measure 1a 2b Totalc Scale Subscaled ∆R2e t df p R2f Wisconsin Card Sorting Test perseverative errors 0.04 0.05 0.09 — — — — — — Verbal fluency 0.01 0.15 0.16 Sensory integration 0.08 –2.46 74 0.02 0.24 Stroop Color and Word Test (residualized) 0.04 0.11 0.15 Sequencing of complex motor acts 0.07 –2.38 74 0.02 0.23 Trail Making Test, part A 0.02 0.18* 0.20 Sensory integration 0.11 3.02 74 0.004 0.31 Trail Making Test, part B 0.13 0.06 0.19 Sensory integration 0.11 3.02 74 0.004 0.30 Category fluency 0.04 0.07 0.11 Sensory integration 0.16 –3.65 74 0.001 0.27 Mooney faces closure 0.14* 0.04 0.18 — — — — — — Benton judgment of lines 0.36** 0.03 0.39 “Others” items 0.09 –3.27 74 0.002 0.52 WAIS-R block design 0.23** 0.06 0.29 Sensory integration 0.13 –3.56 74 0.001 0.42 Wechsler Memory Scale–Revised Logical memory subscale 0.11 0.05 0.16 Sensory integration 0.17 –3.90 74 <0.001 0.33 Figural memory subscale 0.06 0.08 0.14 Sequencing of complex motor acts 0.09 –2.63 74 0.01 0.23 Visual pairs subscale, immediate recall 0.04 0.12 0.16 Sensory integration 0.16 –3.76 74 <0.001 0.32 Verbal pairs subscale, immediate recall 0.05 0.09 0.14 Sensory integration 0.06 –2.14 74 0.04 Motor coordination 0.06 1.76 73 0.04 0.26 Visual reproduction subscale 0.08 0.10 0.18 Sensory integration 0.15 –3.67 74 0.001 0.33 Continuous Performance Test 0.11 0.11 0.22 — — — — — — Span of Apprehension 0.09 0.10 0.19 — — — — — — Mini-Mental State 0.09 0.04 0.14 “Others” items 0.13 –3.21 71 0.002 Sensory integration 0.05 –3.08 70 0.04 0.32 WAIS-R performance scale 0.09 0.09 0.18 Sensory integration 0.29 –5.56 71 <0.001 0.47 WAIS-R verbal scale 0.31** 0.03 0.34 Sensory integration 0.11 –3.32 71 0.002 0.45 a Variance explained by the covariates age, sex, race, and head of household’s socioeconomic status (same covariates used for the comparison group). b Variance explained by the covariates BPRS total score, medication dose, anticholinergic medication, dyskinetic movements, and parkinsonism. c Variance explained by all covariates. d Subscale that contributed significantly to the model after the covariates were entered. e Increment in variance explained by the neurological subscales that entered into the model after the covariates. f Total variance explained by the model. *p<0.05. **p<0.01. DISCUSSION function or that global neurological impairment re- lates to specific neuropsychological function (frontal Patients with schizophrenia exhibit more neurolog- function, motor speed, and motor coordination). In ical signs than normal comparison subjects; in the the present study, while the first case seems to be true, present study the difference was significant for the to- a third possibility arises: specific neurological impair- tal score and the scores on each of the subscales de- ments are related to a broad array of neuropsycholog- rived from the standardized scale used for this study ical test results. In the regression analyses, and after (27). This replicates our previous finding (27). In the control for all theoretical confounding factors, neuro- examination of the relation between neurological logical subscores increased the variance explained for signs and neuropsychological testing, the major find- most neuropsychological test results among the pa- ings were that 1) for both comparison subjects and tients and for eight of the 17 neuropsychological tests patients, Neurological Evaluation Scale scores signif- performed by the comparison subjects. The lesser as- icantly predicted most neuropsychological test re- sociation of neuropsychological test results and neu- sults; 2) sensory integration predicted most of the rological signs among the normal comparison subjects neuropsychological test results; and 3) neurological could be due to a reduced variance in the comparison items grouped in the Neurological Evaluation Scale subjects or to a lesser degree of association between “others” subscale discriminated patients from com- those variables in persons not affected by schizophre- parison subjects better than any other neurological nia. The correlation between neurological signs and subscale. When neurological scores were entered in neuropsychological test performance suggests that the the linear discriminant function with all of the neu- two types of measures share common variance. How- ropsychological test results, the “others” subscale ever, most of the variance of the latter is not explained scores entered into the analysis first. by neurological signs, suggesting that both measures Previous studies have shown that global neurologi- also capture unique variance. The finding that neuro- cal impairment relates to global neuropsychological logical impairment correlates with general cognitive Am J Psychiatry 156:9, September 1999 1353
  6. 6. PREDICTION OF NEUROPSYCHOLOGICAL PERFORMANCE TABLE 5. Results of Multiple Regression Analysis of Healthy Comparison Subjects’ Neuropsychological Test Performance and Neurological Evaluation Scale Scores Neurological Evaluation Scale Final Measure R2a Subscaleb ∆R2c t df p R2d Wisconsin Card Sorting Test perseverative errors 0.13 — — — — — — Verbal fluency 0.21 “Others” items 0.14 –2.48 30 0.02 0.35 Stroop Color and Word Test (residualized) 0.13 — — — — — — Trail Making Test, part A (time) 0.15 Sensory integration 0.09 2.35 30 0.02 0.19 Trail Making Test, part B (residualized) 0.34* Sequencing of complex motor acts 0.11 2.38 30 0.02 0.45 Category fluency 0.20 — — — — — — Mooney faces closure 0.12 Sensory integration 0.25 –3.37 30 0.002 0.37 Benton judgment of lines 0.27 Sensory integration 0.13 –2.51 30 0.02 0.40 WAIS-R block design 0.15 — — — — — — Wechsler Memory Scale—Revised Logical memory subscale 0.07 — — — — — — Figural memory subscale 0.24 “Others” itemse 0.21 –3.31 30 0.003 Sequencing of complex motor acts 0.09 –1.78 29 0.03 0.54 Visual pairs subscale, immediate recall 0.16 Sequencing of complex motor acts 0.24 –3.37 30 0.002 “Others” items 0.24 –3.23 29 0.001 Motor coordination 0.09 –2.29 28 0.006 0.73 Verbal pairs subscale, immediate recall 0.16 — — — — — — Visual reproduction subscale 0.22 Sequencing of complex motor acts 0.26 –3.78 30 0.001 “Others” itemse 0.16 –2.51 29 0.002 0.64 Continuous Performance Test 0.41* — — — — — — Span of Apprehension 0.37* — — — — — — Mini-Mental State 0.25 — — — — — — a Variance explained by the same covariates as for normal subjects: age, sex, race, and head of household’s socioeconomic status. b Subscale that contributed significantly to the model after the covariates were entered. c Increment in variance explained by the neurological subscales that entered into the model after the covariates. d Total variance explained by the model. e Excluding the two memory items. *p<0.05. impairment and not with frontal function or motor schizophrenia, associated with the WAIS-R perfor- speed differs from the findings of studies by Flashman mance. WAIS-R data were not available for the nor- et al. (22) and Mohr et al. (21) but is in agreement mal comparison subjects, so it is not possible to deter- with those of most previous studies (13, 15, 40). Un- mine whether the notable variance explained by fortunately, our battery did not include neuropsycho- neurological signs for the WAIS-R in patients can be logical tests that assess motor coordination. The dif- extrapolated to normal subjects. The Neurological ferences among studies could be attributable to the Evaluation Scale sequencing of complex motor acts neurological signs and neuropsychological tests as- and “others” subscale scores were related to perfor- sessed, including the use of structured neurological mance on only two neuropsychological tests, and the sign assessments, different statistical approaches, dif- Neurological Evaluation Scale motor coordination ferences in samples (patients in the Flashman et al. signs predicted performance on only one of the neu- and Mohr et al. studies were inpatients), or the influ- ropsychological tests. These results suggest that Neu- ence of possible confounding factors. In the study by rological Evaluation Scale sensory integration signs, Flashman et al., the neurological signs used, although which are associated with parietal lobe injury (44), similar to those included in the Neurological Evalua- may be a more sensitive measure of general cognitive tion Scale, did not include items of complex motor act integrity than the remaining neurological signs, and sequencing. The disparity in the prevalence of neuro- that if patients (or normal comparison subjects) ex- logical signs between the present study and the Flash- hibit impairments on these signs, then they are more man et al. study suggests that a different threshold likely to exhibit impaired neuropsychological test per- was used to rate neurological signs as present. formance. Neurological signs were related to a broad range of There are several possible explanations for this ap- cognitive processes, including attention, executive parently broad relation between sensory integration function, memory, motor tasks, general cognitive im- and neuropsychological test performance. First, this pairment (Mini-Mental State examination), and intel- subscale of the Neurological Evaluation Scale has the ligence (WAIS-R). Thus, there was little evidence for highest reliability of any of the subscales. Therefore, any relation between neurological signs and specific the evidence of covariance with neuropsychological cognitive impairments. However, the Neurological performance may simply reflect the superior psycho- Evaluation Scale sensory integration items were asso- metric properties of this scale. Although we cannot ciated with more neuropsychological measures than rule out this “artifactual” explanation, it appears any other subset of neurological signs, with the largest highly unlikely. If the Neurological Evaluation Scale increment in variance explained, in patients with sensory integration subscale has the greatest propor- 1354 Am J Psychiatry 156:9, September 1999
  7. 7. ARANGO, BARTKO, GOLD, ET AL. tion of true score variance, it might also be expected ment with recent studies, which show that develop- to be the test best able to discriminate patients from mental reflexes (e.g., the snout reflex) are more dis- control subjects. However, that is not what was ob- criminating than soft signs (46). The fact that the served, since it was the “others” subscale score that Neurological Evaluation Scale was basically as dis- was the best discriminator and, in fact, the only Neu- criminating as a fairly comprehensive neuropsycholog- rological Evaluation Scale variable that entered into ical battery is somewhat surprising: one might expect the linear discriminant function equation. Two other that the more refined measurement of memory, atten- explanatory frameworks can be considered. First, al- tion, and executive function available in the cognitive though abnormalities of the parietal cortex have not test battery would more completely separate the diag- been highlighted in the schizophrenia literature on nostic groups. This was clearly not what we observed. postmortem findings or those from structural and We suggest that the Neurological Evaluation Scale is a functional imaging studies, a number of recent stud- good indicator of the presence of an “abnormality”: ies have suggested that schizophrenia involves a dif- something that occurs with low frequency among com- fuse gray matter abnormality or an abnormality of parison subjects but is present with high frequency in the heteromodal association cortex, including the in- ill patients, including patients who demonstrate gener- ferior parietal cortex (45). It is plausible that evidence ally good intellectual performance. Alternatively, neu- of functional compromise of the parietal cortex pro- ropsychological performance is more broadly distrib- vides an index of the “extent” of cortical pathology. uted (in both groups) and is likely influenced by a Such a marker of compromised cortical function variety of demographic and educational factors (that would be expected to be related to multiple cognitive may be independent of the illness), which may decrease functions, including those not primarily dependent the sensitivity and specificity of neuropsychological on the parietal lobe. In essence, that is what we ob- test scores. In the current patient series, we examined served in our patient group. This argument, however, the Neurological Evaluation Scale performance of the does not provide any explanation for the fact that the 15 patients with the highest IQs (the mean of this sensory integration subscale score of the Neurologi- group across four WAIS-R subtests was 109, slightly cal Evaluation Scale also contributed to the regres- above the normal population mean). These patients sions in the normal comparison group. An alternative scored significantly worse than the comparison sub- explanation is that the correlations reflect the fact jects on the Neurological Evaluation Scale “others” that the parietal lobe plays a critical role in the sup- subscale (p=0.005) but did not differ on the other three port of more complex, widely distributed cognitive scales. This finding complements the other results of functions as the principal site of polysensory, poly- the study and provides further evidence that the “oth- modal integration. Thus, an impairment (or, possibly, ers” subscale is sensitive to the presence of schizophre- normal individual differences) in these functions nia, even among patients with relatively intact cogni- might be expected to be related to fairly broad cogni- tive performance. The high discriminant strength of tive consequences. the “others” subscale supports the inclusion of these These three explanations (psychometric artifact, dis- items in the study of neurological signs as a phenotypic ease-specific implications of parietal function, and cog- marker. We suggest that studies designed to address the nitive implications of parietal function) are not neces- familial transmission of the disease should include the sarily mutually exclusive, and it is difficult to evaluate items from the Neurological Evaluation Scale “others” their relative merits in this data set with confidence. subscale. However, the present data suggest that further studies of There are some limitations to this study. One is the parietal lobe structure and function in schizophrenia small number of subjects, especially in the comparison may prove to be fruitful. If a high correlation between group. Another limitation refers to generalizability; be- Neurological Evaluation Scale sensory integration items cause patients attending our clinic are a group with and neuropsychological impairment is confirmed in fur- chronic illness, the present findings may not be gener- ther studies, this simple, inexpensive, and easy-to-ad- alizable to other settings. minister subscale could be used to screen patients for In conclusion, neurological signs seem to be a neuropsychological impairment. marker of global cognitive impairment. Although Neurological signs included in the “others” subscale neurological signs have been previously reported as of the Neurological Evaluation Scale discriminated be- soft signs because of a lack of localizing ability, some tween normal subjects and patients as accurately as the subsets of these signs are specifically related to differ- comprehensive neuropsychological test battery ent neuropsychological tests thought to represent (78.5% versus 81.8%). In addition, when both assess- function in different brain areas. In addition, a bat- ments were included in the discriminant function anal- tery of neurological signs, such as the Neurological ysis, the overall percent of total classification was Evaluation Scale in the present study, has shown the 81.8%, and the variable that entered first was the Neu- same discriminating power as a comprehensive neu- rological Evaluation Scale “others” subscale score. ropsychological test battery. Integration of knowl- The “others” subscale includes frontal release signs edge about the different neurological impairments in (primitive reflexes), abnormalities in eye movements, schizophrenia should illuminate the pathophysiology and short-term memory. These findings are in agree- of the disease. Am J Psychiatry 156:9, September 1999 1355
  8. 8. PREDICTION OF NEUROPSYCHOLOGICAL PERFORMANCE APPENDIX 1. Items on the Neurological Evaluation Scalea REFERENCES Sensory integration subscale 1. Heinrichs DW, Buchanan RW: Significance and meaning of Audio-visual integration neurological signs in schizophrenia. Am J Psychiatry 1988; Stereognosisb 145:11–18 Graphesthesiab 2. Fish B: Neurobiologic antecedents of schizophrenia in chil- Extinction dren: evidence for an inherited, congenital neurointegrative Right/left confusion defect. Arch Gen Psychiatry 1977; 34:1297–1313 Motor coordination subscale 3. Cooper SJ, King DJ: Is schizophrenia a neurodevelopment Tandem walk disorder? (letter). BMJ 1987; 295:1068 Rapid alternating movementsb 4. Parnas J, Schulsinger F, Teasdale TW, Schulsinger H, Feld- Finger-thumb oppositionb man PM, Mednick SA: Perinatal complications and clinical Finger-nose testb outcome within the schizophrenia spectrum. Br J Psychiatry Sequencing of complex motor acts subscale 1982; 140:416–420 Fist-ringb 5. Merriam AE, Kay SR, Opler LA, Kushner SF, van Praag HM: Fist-edge-palmb Neurological signs and the positive-negative dimension in Ozeretski test schizophrenia. Biol Psychiatry 1990; 28:181–192 Rhythm tapping test, B 6. Buchanan RW, Kirkpatrick B, Heinrichs DW, Carpenter WT Jr: “Others” subscale Clinical correlates of the deficit syndrome of schizophrenia. Adventitious overflowb Am J Psychiatry 1990; 147:290–294 Romberg test 7. Smith RC, Kadewari RP: Neurological soft signs and re- Tremorb sponse to risperidone in chronic schizophrenia. Biol Psychia- Memory, 5 minutes try 1996; 40:1056–1059 Memory, 10 minutes 8. Schroder J, Geider FJ, Binkert M, Reitz C, Jauss M, Sauer H: Rhythm tapping test, A Subsyndromes in chronic schizophrenia: do their psycho- Mirror movementsb pathological characteristics correspond to cerebral alter- Synkinesisb ations? Psychiatry Res 1992; 42:209–220 Convergenceb 9. Johnstone EC, Macmillan JF, Frith CD, Benn DK, Crow TJ: Fur- Gaze impersistenceb ther investigation of the predictors of outcome following first Glabellar reflex schizophrenic episodes. Br J Psychiatry 1990; 157:182–189 Snout reflex 10. Wong AH, Voruganti LN, Heslegrave RJ, Awad AG: Neu- Grasp reflexb rocognitive deficits and neurological signs in schizophrenia. Suck reflex Schizophr Res 1997; 23:139–146 a For a description of the items, see Buchanan and Heinrichs (27). 11. Quitkin F, Rifkin A, Klein DF: Neurologic soft signs in schizo- b Assessed bilaterally. phrenia and character disorders: organicity in schizophrenia with premorbid asociality and emotionally unstable character APPENDIX 2. Neuropsychological Test Battery disorders. Arch Gen Psychiatry 1976; 33:845–853 12. Krakowski MI, Convit A, Jaeger J, Lin S, Volavka J: Neurolog- Test Variable ical impairment in violent schizophrenic patients. Am J Psy- Wisconsin Card Sorting Test Percent of perseverative errors chiatry 1989; 146:849–853 Controlled Oral Word Associa- Total number of words 13. Tucker GJ, Campion EW, Silberfarb PM: Sensorimotor func- tion (verbal fluency) produced tions and cognitive disturbance in psychiatric patients. Am J Stroop Color and Word Test Total number of color-word Psychiatry 1975; 132:17–21 items correctly completeda 14. Kolakowska T, Williams AO, Jambor K, Ardern M: Schizophre- Trail Making Test, part A Total number of seconds nia with good and poor outcome, III: neurological “soft” signs, Trail Making Test, part B Total number of secondsa cognitive impairment and their clinical significance. Br J Psy- Category fluency Number of words chiatry 1985; 146:348–357 Mooney faces closure Total number of faces correctly 15. Liddle PF: Schizophrenic syndromes, cognitive performance identified and neurological dysfunction. Psychol Med 1987; 17:49–57 Benton judgment of lines Number of items correct 16. Chen EYH, Lam LCW, Chen RYL, Nguyen DGH: Cognitive WAIS-R block design Total number of points, age- correlates of soft signs in schizophrenia (abstract). Schizophr corrected Res 1997; 24:100 Wechsler Memory Scale— 17. Kennard MA: Value of equivocal signs in neurologic diagno- Revised sis. Neurology 1960; 10:753–764 Logical memory subscale Total number of items correctly 18. Mosher LR, Pollin W, Stabenau JR: Identical twins discordant repeated for schizophrenia: neurologic findings. Arch Gen Psychiatry Figural memory subscale Total number of designs 1971; 24:422–430 correctly identified 19. Manschreck TC, Ames D: Neurologic features and psychopa- Visual pairs subscale Total number of color-figure thology in schizophrenic disorders. Biol Psychiatry 1984; 19: associations 703–719 Verbal pairs subscale Total number of correct word 20. Cuesta MJ, Peralta V, de Leon J: Neurological frontal signs associations and neuropsychological deficits in schizophrenic patients. Visual reproduction subscale Total number of accuracy crite- Schizophr Res 1996; 20:15–20 ria met in drawing designs 21. Mohr F, Hubmann W, Cohen R, Bender W, Haslacher C, Hon- Continuous Performance Test d′ (overall performance metric) icke S, Schlenker R, Wahlheim C, Werther P: Neurological Span of Apprehension Percent correct per group size soft signs in schizophrenia: assessment and correlates. Eur (6- to 12-item displays) Arch Psychiatry Clin Neurosci 1996; 246:240–248 Mini-Mental State examination Total score 22. Flashman LA, Flaum M, Gupta S, Andreasen NC: Soft signs WAIS-R verbal scale Verbal IQ and neuropsychological performance in schizophrenia. Am J WAIS-R performance scale Performance IQ Psychiatry 1996; 153:526–532 a See Method section of the text for a more detailed description of 23. Spitzer RL, Williams JBW, Gibbon M, First MB: Structured scoring procedures. Clinical Interview for DSM-III-R—Patient Version (SCID-P). 1356 Am J Psychiatry 156:9, September 1999
  9. 9. ARANGO, BARTKO, GOLD, ET AL. New York, New York State Psychiatric Institute, Biometrics 36. Russell EW: A multiple scoring method for assessment of Research, 1989 complex memory functions. J Consult Clin Psychol 1975; 43: 24. Overall JE, Gorham DR: The Brief Psychiatric Rating Scale. 800–809 Psychol Rep 1962; 10:799–812 37. Lezak MD: Neuropsychological Assessment, 2nd ed. New 25. Cassady SL, Thaker GK, Summerfelt A, Tamminga CA: The York, Oxford University Press, 1983 Maryland Psychiatric Research Center Scale and the charac- 38. Folstein MF, Folstein SE, McHugh PR: “Mini-Mental State”: a terization of involuntary movements. Psychiatry Res 1997; 70: practical method for grading the cognitive state of patients for 21–37 the clinician. J Psychiatr Res 1975; 12:189–198 26. Schooler N: Antipsychotic medications in schizophrenia: ef- 39. Buchanan RW, Strauss ME, Kirkpatrick B, Holstein C, Breier fects in acute and maintenance treatment of the illness, in A, Carpenter WT Jr: Neuropsychological impairments in defi- Schizophrenia: Origins, Processes, Treatment, and Outcome. cit vs nondeficit forms of schizophrenia. Arch Gen Psychiatry Edited by Cromwell RL, Snyder CR. New York, Oxford Univer- 1994; 51:804–811 sity Press, 1993, pp 284–295 40. King DJ, Wilson A, Cooper SJ, Waddington JL: The clinical 27. Buchanan RW, Heinrichs DW: The Neurological Evaluation correlates of neurological soft signs in chronic schizophrenia. Scale (NES): a structured instrument for the assessment of Br J Psychiatry 1991; 158:770–775 neurological signs in schizophrenia. Psychiatry Res 1989; 27: 41. Gupta S, Andreasen NC, Arndt S, Flaum M, Schultz SK, Hub- 335–350 bard WC, Smith M: Neurological soft signs in neuroleptic-na- 28. Taylor LB: Psychological assessment of neurosurgical pa- ive and neuroleptic-treated schizophrenic patients and in nor- tients, in Functional Neurosurgery. Edited by Rasmussen T, mal comparison subjects. Am J Psychiatry 1995; 152:191– Marino R. New York, Raven Press, 1979, pp 165–180 196 29. Milner B: Some effects of frontal lobectomy in man, in The Frontal Granular Cortex and Behavior. Edited by Warren JM, 42. Barnes TR, Crichton P, Nelson HE, Halstead S: Primitive (de- Akert K. New York, McGraw-Hill, 1964, pp 313–334 velopmental) reflexes, tardive dyskinesia and intellectual im- 30. Golden CJ: Stroop Color and Word Test: A Manual for Clinical pairment in schizophrenia. Schizophr Res 1995; 16:47–52 and Experimental Uses. Chicago, Stoelting, 1978 43. Efron B: The Jackknife, the Bootstrap, and Other Resampling 31. Reitan RM, Wolfson D: The Halstead-Reitan Neuropsycho- Plans: CBMS Monograph 38. Philadelphia, Society for Indus- logical Test Battery: Theory and Clinical Interpretation. Tuc- trial and Applied Mathematics, 1981 son, Ariz, Neuropsychology Press, 1975 44. Brown JW: Frontal lobe syndromes, in Handbook of Clinical 32. Mooney CM: Age in the development of closure ability in chil- Neurology, vol 45 (revised series 1): Clinical Neuropsychol- dren. Can J Psychol 1957; 11:219–226 ogy. Edited by Fredericks JAM. New York, Elsevier, 1985, pp 33. Benton AL, Hannay HJ, Varney NR: Visual perception of line 23–41 direction in patients with unilateral disease. Neurology 1975; 45. Pearlson GD, Petty RG, Ross CA, Tien AY: Schizophrenia: a 25:907–910 disease of heteromodal association cortex? Neuropsychop- 34. Wechsler D: Wechsler Adult Intelligence Scale—Revised, harmacology 1996; 14:1–17 Manual. Cleveland, Psychological Corp, 1981 46. Ismail B, Cantor-Graae E, McNeil TF: Neurological abnormal- 35. Wechsler D: A standardized memory scale for clinical use. J ities in schizophrenic patients and their siblings. Am J Psychi- Psychol 1945; 19:87–95 atry 1998; 155:84–89 Am J Psychiatry 156:9, September 1999 1357