Cognitive dysfunction and associated factors in patients with chronic schizophrenia.

BACKGROUND: Deficits in neurocognitive function are a hallmark of schizophrenia. They are associated with clinical manifestations and the course of the illness. A study of cognitive dysfunction in Indian patients with schizophrenia is of significance in view of a more benign course and outcome of the illness in this region. AIM: To study cognitive deficits and associated factors in patients with chronic schizophrenia and compare them with those in the normal population.
METHODS: We compared 100 patients with chronic schizophrenia with 100 matched normal controls on multiple measures of attention, executive function and memory.
RESULTS: Compared to normal individuals, patients with schizophrenia performed poorly in all cognitive tests. Cognitive deficits in patients were related to gender, education, age, duration of illness, and presence of positive and negative symptoms.
CONCLUSION: The neurocognitive profile of Indian patients with chronic schizophrenia resembles those of patients in developed countries.

INTRODUCTION

Schizophrenia is accompanied by impairments in several domains of cognitive function.1 Patients with schizophrenia have been found to perform more poorly than normal controls on tasks of attention, memory, executive function, language, learning and motor control.2–4 In recent times, cognitive impairment has gained importance in terms of emerging theories on the aetiology and treatment of schizophrenia.5

Cognitive impairment in schizophrenia has been found to be related to measures of psychopathology67 and outcome.89 Much research on cognition in schizophrenia has been done in developed countries where the outcome was found to be poorer than that in developing countries such as India. It is of interest to know the degree and nature of cognitive dysfunction in Indian patients with schizophrenia. Studies in India have described cognitive deficits in schizophrenia.1011 However, a comprehensive evaluation of deficits in all major cognitive domains, and their relation with demographic and clinical variables, has not been done. We compared cognitive deficits and associated factors in patients with chronic schizophrenia with those of a matched normal population.

METHODS

The case group was a consecutive sample selected from outpatients attending the treatment and rehabilitation centre of the Schizophrenia Research Foundation (India) in Chennai and comprised 100 subjects (men: 60; women: 40) fulfilling the DSM-IV criteria for chronic schizophrenia. A clinical interview and chart review established the diagnosis. All of them were on antipsychotic drug treatment at the time of evaluation. Subjects between the ages of 18 and 45 years, with at least 10 years of school education, were selected. The control group comprised 100 healthy subjects (men: 60; women: 40) with no current, past or family history of any psychiatric disorder. They were selected from among volunteers by the stratified sampling method and matched with subjects from the study group for age, sex and education. All participants gave a written informed consent after being explained the nature of the study. The cases and controls did not differ significantly in their mean age (33.6 years, SD±8.2 vs 33.9 years, SD±8.1; t=0.251) and years of formal education (14.3 years, SD±3.1 vs 13.9 years, SD±2.8; t=0.893). The patients were ill for a mean duration of 10.4 years (SD±6.8). The neuropsychological tests done are listed in Table 1.12–16

Table 1

Neuropsychological tests done on the study sample

Test done (subtests)	Function measured
Digit Span Test12 (forward and backward)	Span of attention (verbal task)
Visual Memory Span12 (forward and backward)	Span of attention (non-verbal task)
Digit Symbol Substitution Test13	Sustained attention and speed
Visual Number Scanning Ability Test*(time taken, number/minute)	Visual scanning and attention
Ideational Fluency Test*	Executive function—verbal fluency
Ruff Figural Fluency Test14 (unique and perseverative responses)	Executive function—non-verbal fluency
Wisconsin Card Sorting Test15	Executive functions and cognitive flexibility
Letter–Number Span test16 (correct and longest)	Working memory
Delayed Response Learning Test*	Working memory
Verbal Learning and Memory* (immediate, delayed recall)	Logical memory and learning
Verbal Paired Associate Learning Test12 (immediate, delayed recall)	Associate learning (verbal)
Visual Paired Associate Learning Test12 (immediate, delayed recall)	Associate learning (visual)
Visual Reproduction Test12 (immediate, delayed recall)	Immediate and delayed visual memory

NIMHANS Battery (unpublished)

Data analysis

The Statistical Package for Social Sciences (SPSS)17 was used for data analysis. The chi-square and t tests were applied for univariate analysis. Simple correlation and partial correlation analyses were done to measure the relationship between continuous variables. The variables significant at univariate analysis were entered into classification analysis using the Mahalanobi distant statistic method to identify neuro-psychological tests that differentiated normals from patients.

RESULTS

The mean scores on the Positive and Negative Syndrome Scale (PANSS)18 were 10.2 (SD±3.9) for the positive subscale (PS), 9.6 (SD±3.2) for the negative subscale (NS) and 23.6 (SD±5.7) for the general psychopathology subscale (GS).

Cognitive deficits

The patients performed significantly poorer than normal subjects on all tests of cognitive functions evaluated— attention, executive function, memory—except the number of perseverative responses on the Ruff Figural Fluency test for executive function, and immediate recall on the Visual Reproduction task of memory (Table 2).

Table 2

Comparison of the cognitive functions in patients with schizophrenia and normal controls

Neuropsychological test	Normal subjects Mean±SD	Patients Mean±SD	t score
Attention
Visual Scanning—time taken	162.9±47.4	241.9±99.8	7.14*
Visual Scanning—number/minute	23.5±3.4	17.0±4.6	11.37*
Digit Span—forward	10.7±1.3	9.0±2.0	6.88*
Digit Span—backward	9.6±1.6	7.3±1.9	9.29*
Visual Memory Span—forward	10.9±1.5	9.0±2.0	7.08*
Visual Memory Span—backward	9.5±1.18	7.5±2.4	7.26*
Digit Symbol Substitution Test	57.0±10.5	41.0±11.6	10.18*
Executive function
Wisconsin Card Sorting Test
Trials administered	98.0±19.2	116.8±17.9	7.14*
Total correct	70.7±6.83	66.1±16.2	2.65*
Total errors	27.3±15.4	50.8±25.6	7.87*
Categories completed	5.8±0.6	3.8±2.2	8.71*
Trials to complete first category	13.2±5.1	35.7±40.7	5.49*
Perseverative response—total	18.1±12.9	41.1±32.8	6.53*
Perseverative errors—total	16.0±10.6	34.3±24.5	6.87*
Non-perseverative errors—total	11.5±6.4	16.6±10.0	4.31*
Conceptual level responses—total	63.9±6.6	51.6±22.0	5.36*
Failure to maintain set	0.21±0.43	0.84±1.2	4.99*
Other tests
Ideational Fluency	17.6±3.4	12.9±3.4	9.72*
Ruff Figural Fluency—perseveration	6.5±6.4	7.4±7.5	0.88 (NS)
Ruff Figural Fluency—unique responses	56.9±16.7	35.2±16.4	9.28*
Memory
Verbal Paired Association—immediate	21.6±1.7	19.0±3.6	6.43*
Verbal Paired Association—delayed	7.9±0.3	7.3±1.4	4.37*
Visual Paired Association—immediate	15.6±1.9	11.5±5.0	7.75*
Visual Paired Association—delayed	5.8±0.4	4.9±1.2	6.60*
Visual Reproduction—immediate	35.4±3.7	33.1±31.7	0.72 (NS)
Visual Reproduction—delayed	31.5±6.1	25.6±9.3	5.25*
Verbal Learning and Memory—delayed	22.3±1.3	20.1±3.3	6.30*
Visual Learning and Memory—delayed	16.6±3.2	13.5±4.9	5.20*
Delayed Response Learning	15.7±1.3	13.1±3.0	7.98*
Letter–Number Span—correct responses	18.0±2.7	14.9±3.5	6.92*
Letter–Number Span—longest item	5.9±0.7	5.3±1.0	5.10*

p≤0.01, which is significant

NS: not significant

The step-wise, discriminant function analysis identified 10 tests measuring tasks of attention, executive function and memory which differentiated most between patients and normal controls. The minimum D squared statistic and standardized canonical discriminant function coefficients (SCDFC) of the tests are listed in Table 3. A classification analysis based on the SCDFC of these 10 variables classified 92% of the study population appropriately into their original groups as patients and normal subjects.

Table 3

Function analysis of patients with schizophrenia and normal controls (standardized canonical discriminant function coefficient)

Neuropsychological test	Patients with schizophrenia	Normal subjects
Attention
Visual Scanning—time taken	6.593	0.319
Visual Scanning—number/minute	2.585	0.509
Digital Span—backward recall	5.214	0.220
Digit Symbol Substitution Test	6.333	0.341
Executive function
Wisconsin Card Sorting Test
Failure to maintain set	4.545	−0.359
Other tests
Ideational Fluency	3.766	0.391
Ruff Figural Fluency—unique respose	5.538	0.278
Memory
Verbal Paired Association—delayed	5.754	0.248
Verbal Learning and Memory—delayed	6.876	0.228
Visual Learning and Memory—delayed	6.007	−0.324

Social and clinical factors and cognition

Women performed better than men on only one task: the Visual Paired Associate learning test (mean scores: immediate recall=12.7, SD±3.8 vs 10.7, SD±5.5, t=2.06, p<0.05; delayed recall=5.4, SD±1.0 vs 4.7, SD±1.3, t=3.03, p<0.01). The years of education did not correlate with age or clinical factors. The age and duration of illness correlated with each other (r=0.723, p<0.001) but not with PANSS subscale scores. The three PANSS subscale scores correlated positively with each other at a significance level of 0.01 or less (correlation coefficients: PS with NS=0.280; PS with GS=0.499 and NS with GS=0.461).

Table 4 presents the significant correlations (p<0.05) among scores on cognitive tests with education, age (controlling for duration of illness), duration of illness (controlling for age) and scores on each of the subscales of PANSS (controlling for scores on the other two subscales of PANSS). Increasing age correlated with scores on the Digit Span and Digit Symbol Substitution Tests of attention, Ruff Figural Fluency Test of executive function, and verbal working memory tested by the Letter–Number Span test. More years of education correlated with better performance on tasks of attention, executive function, verbal and visual memory. A longer duration of illness correlated with indicators of executive dysfunction on the Wisconsin Card Sorting Test (WCST) and verbal memory. The positive symptom score was related to deficit on a single test of verbal memory, and negative symptoms with performance on measures of attention, executive function and visual memory. The GS score did not correlate with any cognitive deficit.

Table 4

Cognitive deficits and social and clinical factors (correlation coefficients)

	Correlation coefficients (p<0.05)
				PANSS
Neuropsychological test	Age	Education	DOI	PS	NS	GS
Attention
Visual Scanning—time taken		−0.257			0.255
Visual Scanning—number/minute		0.203
Digit Span—forward	−0.245	0.344
Digit Span—backward	−0.269
Visual Memory Span—forward		0.344
Visual Memory Span—backward		0.311
Digit Symbol Substitution Test	−0.305	0.253
Executive function
Wisconsin Card Sorting Test
Trials administered		−0.380	0.223
Total errors		−0.400			0.280
Categories completed		0.373			−0.229
Trials to complete first category		−0.224			0.199
Perseverative response		−0.352			0.351
Perseverative errors		−0.356			0.338
Non-perseverative errors			0.255
Conceptual level responses					−0.329
Other tests
Ideational Fluency					−0.412
Ruff test—perseverative	0.212
Ruff test—unique responses	−0.213
Memory
Verbal Paired Association—immediate		0.208
Verbal Paired Association—delayed
Visual Paired Association—immediate		0.223			−0.245
Visual Paired Association—delayed		0.392			−0.213
Visual Reproduction—immediate
Visual Reproduction—delayed		0.340		−0.222	−0.333
Verbal Learning and Memory—delayed			−0.197
Visual Learning and Memory—delayed		0.286
Delayed Response Learning		0.439
Letter–Number Span—correct responses	−0.245	0.478
Letter–Number Span—longest item	−0.252	0.394

DOI: duration of illness, PANSS: Positive and Negative Syndrome Scale, PS: Positive subscale, NS: Negative subscale, GS: General psychopathology subscale

DISCUSSION

Cognitive deficits in chronic schizophrenia

We did not have any difficulty in using the neuropsychological tests developed in other cultures. The significant level of schooling of patients during which English was one of main languages taught seemed to facilitate their ability to understand and perform on tests that had numerate or verbal tasks. We feel cultural factors had little impact on performance in the neuropsychological tests.

Patients with schizophrenia performed poorly on all tests of cognitive function compared with the normal population matched with respect to gender, age and education. The classification analysis showed that patients with schizophrenia can often be clearly differentiated from the normal population based on their performance on some of the tests of attention, executive function and memory.

Factors associated with cognitive deficits

Gender differences in cognitive dysfunction have been reported. Males have been found to have more cognitive deficits than females, a trend attributed to the interplay of sex hormones, neurodevelopmental and psychosocial sex differences.19 We did not find any major gender difference except for a poorer performance of males on a memory task. Age-related decline across most neuropsychological functions has been demonstrated in schizophrenia.20 We found that increasing age was related to poorer performance on tasks of attention, executive function and memory, which has been pointed out to be the result of an ageing brain in patients.

More years of education positively influenced performance on tasks that tested attention, executive function, memory and constructional ability. The duration of formal academic training reflected good pre-morbid functioning, intellectual level and a higher level of information-processing skills in the past. Patients with good education thus did well on cognitive tasks because of this inherent capability. A parallel can be drawn with the influence of education on cognitive changes reported in other neurological disorders.21–23 Cognitive deficits have been found to remain relatively stable throughout the course of schizophrenia.24 We also found that all measures, except two measures of executive function on the WCST and one of verbal memory, were stable over a range of illness duration.

We observed that negative symptoms had a strong association with cognitive dysfunction in all the domains. This finding is in agreement with the results of studies which showed that both positive and negative symptoms were associated with distinct neuropsychological deficits.25 Heydebrand et al.26 observed that negative symptoms were related more frequently to cognitive dysfunction than positive symptoms.

CONCLUSION

In a group of patients with chronic schizophrenia in India, the nature and degree of cognitive deficits and their relationship to gender, age and clinical factors are comparable with observations made in developed countries. It would be of interest to explore the relationship between cognitive deficits in, and outcome of, schizophrenia among Indian patients, as they have a better outcome than patients with schizophrenia in developed countries.