- Top of page
Abstract The measures of prefrontal cognition have been used as endophenotype in molecular-genetic studies. Brain-derived neurotrophic factor (BDNF) has been implicated in cognitive functions and in the pathogenesis of major psychoses. This study investigates the relationship between Val66Met polymorphisms of the BDNF gene and prefrontal cognitive function in 129 patients with schizophrenia and 111 patients with bipolar mood disorder. Cognitive tests included the Wisconsin Card Sorting Test (WCST), with such domains as number of perseverative errors, non-perseverative errors, completed corrected categories, conceptual level responses, and set to the first category, and the N-back test, where mean reaction time and percent of correct reactions were measured. Genotyping for Val66Met BDNF polymorphism was done by polymerase chain reaction method. In schizophrenia, no relationship between Val66Met polymorphism of the BDNF gene and the results of the WCST was observed. Patients with Val/Val genotype had a higher percentage of correct reactions in the N-back test than those with the remaining genotypes. Bipolar patients with Val/Val genotype obtained significantly better results on three of five domains of the WCST. No relationship between BDNF polymorphism and the results of the N-back test was found in this group. A limitation to the results could be variable psychopathological state and medication during cognitive testing and lack of Hardy–Weinberg equilibrium in schizophrenia group. Val66Met polymorphism of the BDNF gene may be associated with cognitive performance on the WCST in bipolar mood disorder but not in schizophrenia. An association of this polymorphism with performance on the N-back test in schizophrenia and not in bipolar illness may suggest that in schizophrenia, the BDNF system may be connected with early phases of information processing.
- Top of page
The measures of cognition connected with prefrontal cortex have generated considerable interest as potential endophenotypes for molecular-genetic studies in major psychoses. The performance on the Wisconsin Card Sorting Test (WCST) has long been regarded as a neuropsychological marker of the efficiency of working memory and executive functions, depending on the activity of prefrontal cortex.1 In patients with schizophrenia, the deficits on the WCST performance have been recognized as enduring and a core feature of the illness.2,3 Such deficits have also been present in healthy first-degree relatives of patients with schizophrenia;4,5 what prompted their use as cognitive endophenotype for molecular-genetic studies in schizophrenia. An association of this phenotype with the Val158Met polymorphism of the gene for catechol-O-methyltransferase, the enzyme responsible for dopamine breakdown in prefrontal cortex was found by some researchers.6 In recent years, the association of this polymorphism was also investigated with another measure of working memory function, the N-back test.7,8
Brain-derived neurotrophic factor (BDNF) is a potent modulator of synaptic transmission and plasticity in the central nervous system and has been implicated in such cognitive processes as memory and learning.9,10 Substantial evidence has been accumulated pointing to the role of BDNF in the pathophysiology of mood disorders and in the mechanism of action of therapeutic agents.11 The possible role of BDNF in schizophrenia has also been postulated, especially in the context of the neurodevelopmental theory of this illness.12
Several polymorphisms of the BDNF gene have been reported, the most intensively investigated being Val66Met (196G/A) polymorphis.13 Two studies of association between Val66Met BDNF gene polymorphism and schizophrenia have yielded negative results.14,15 In contrast, studies in bipolar mood disorder brought about several positive reports, all showing an association of Val allele with a predisposition to bipolar illness.16–19 However, in a number of studies such association has not been found.15,20–23
So far, the studies of cognitive performance on neuropsychological tests measuring prefrontal cortex activity in relation to BDNF gene polymorphism in patients with schizophrenia have not been performed. However, in Egan et al.’s24 study it was found that Val66Met polymorphism of the BDNF gene was connected with the quality of performance on episodic memory test in patients with schizophrenia and control subjects; the presence of Met allele was associated with poorer performance. Episodic memory is mainly dependent on hippocampal function and in that study, the efficiency of such memory was also related to hippocampal activation.
In the authors’ preliminary study, an association between Val66Met BDNF polymorphism and cognitive performance on prefrontal cortex test in bipolar patients was found. Significantly better results on all domains of the WCST were obtained in patients with Val/Val than in those having Val/Met genotype.25
The aim of this study was to extend the preliminary investigation to larger groups of schizophrenic and bipolar patients and also to correlate Val66Met BDNF polymorphism with the results of another test measuring prefrontal function – the N-back test.
- Top of page
Demographic features and the results of cognitive prefrontal tests in whole groups of patients with schizophrenia and patients with bipolar who were studied compared with a group of healthy control persons are shown in Table 1.
Table 1. Demographic features and the results of cognitive prefrontal tests in whole groups of patients with schizophrenia and patients with bipolar who were studied compared with a group of healthy control persons (values are given as means ± standard deviation)
| ||Schizophrenia (n = 129)||Bipolar disorder (n = 111)||Healthy controls (n = 160)|
|Age (years)||27.1 ± 9.6*†||43.4 ± 13.7‡||32.9 ± 11.5|
|Education (years)||11.8 ± 2.1*†||12.6 ± 2.0‡||14.9 ± 2.6|
|WCST-P||15.9 ± 9.9*†||11.5 ± 7.0‡|| 7.9 ± 3.0|
|WCST-NP||13.4 ± 9.7*†||10.4 ± 6.3‡|| 7.7 ± 3.4|
|WCST-CC|| 4.8 ± 1.8*†|| 5.7 ± 1.0‡|| 6.0 ± 0.2|
|WCST % CONC||64.1 ± 20.7*†||74.1 ± 14.0‡||81.4 ± 7.4|
|WCST-1st CAT||26.1 ± 27.1*†||18.5 ± 15.7‡||13.4 ± 7.2|
|N-back time|| 938 ± 449†|| 983 ± 301‡|| 516 ± 238|
|N-back %CORR||33.1 ± 21.4*†||39.4 ± 20.9‡||93.6 ± 12.1|
As seen in the table, on the WCST and N-back test, both patients with schizophrenia and patients with bipolar obtained significantly worse results compared to control persons. In the WCST, schizophrenic subjects performed significantly poorer on all parameters compared with bipolar patients. In the N-back test, the results of both clinical groups were similar on N-back time and slightly better in patients with bipolar on N-back %CORR. Concerning demographic differences it may be noticed that patients with bipolar were older than the remaining groups, while the healthy controls had more years of education.
Distributions of genotypes were in Hardy–Weinberg equilibrium in patients with bipolar (P = 0.289) but not in schizophrenia (P = 0.01).
The results of the WCST and the N-back test in relation to Val66Met polymorphism of the BDNF gene in patients with schizophrenia are shown in Table 2.
Table 2. The results of the Wisconsin Card Sorting Test and the N-back test in relation to Val66Met polymorphism of the BDNF gene in patients with schizophrenia (values are given as means ± standard deviation)
|Cognitive domain||Val/Val (n = 84)||Val/Met (n = 34)||Met/Met (n = 11)|
|WCST-P||15.8 ± 9.6|| 16.4 ± 11.2|| 15.6 ± 9.5|
|WCST-NP||12.7 ± 9.9|| 14.6 ± 9.3|| 15.0 ± 9.1|
|WCST-CC|| 4.9 ± 1.8|| 4.7 ± 1.9|| 4.8 ± 1.6|
|WCST %CONC||65.5 ± 20.4|| 60.6 ± 22.4|| 64.2 ± 17.5|
|WCST-1st CAT||25.7 ± 29.3|| 27.4 ± 23.8|| 25.2 ± 19.7|
|N-back time|| 906 ± 400||1076 ± 601||1023 ± 646|
|N-back %CORR||34.7 ± 22.7*|| 26.1 ± 13.8|| 29.4 ± 15.9|
As seen in the table, no difference in any domain of the WCST was found between patients with different genotypes. In contrast, in the N-back test, patients with Val/Val genotype performed significantly better than those with the remaining genotypes. The Val/Val patients obtained a significantly higher percentage of correct reactions and were also numerically better in the mean reaction time although that did not reach statistical significance.
The results of the WCST and the N-back test in relation to Val66Met polymorphism of the BDNF gene in patients with bipolar are shown in Table 3. Since only three patients with bipolar (2.7%) had Met/Met genotype, the patients with Val/Met and Met/Met genotype were combined in one group and compared with the Val/Val genotype group.
Table 3. The results of the Wisconsin Card Sorting Test and the N-back test in relation to Val66Met polymorphism of the BDNF gene in patients with bipolar (values are given as means ± standard deviation)
|Cognitive domain||Val/Val (n = 81)||Val/Met (n = 27) + Met/Met (n = 3)|
|WCST-P||10.6 ± 4.5||15.0 ± 11.0**|
|WCST-NP|| 9.8 ± 4.9||11.3 ± 9.5|
|WCST-CC|| 5.8 ± 0.5|| 5.2 ± 1.6**|
|WCST %CONC||75.8 ± 10.7||67.7 ± 20.4*|
|WCST-1st CAT||16.5 ± 8.6||24.3 ± 26.2|
|N-back time|| 959 ± 336|| 967 ± 300|
|N-back %CORR||39.0 ± 20.2||41.1 ± 26.9|
Concerning Val66Met polymorphism of the BDNF gene, there was a significant difference between patients with Val/Val versus Val/Met + Met/Met genotype on WCST-P, WCST-CC and WCST %CONC. A numerical difference on WCST-1st CAT did not reach significance. No difference was found between these two groups in any of the N-back test measures.
- Top of page
The main finding of the present study shows that Val66Met polymorphism of the BDNF gene is connected with a quality of cognitive performance on Wisconsin test measuring prefrontal cortex activity in bipolar illness but not in schizophrenia. These results confirm the authors’ previous data obtained on a lower number of patients.25 Another finding is a possible association between this polymorphism and the results on the N-back test in patients with schizophrenia but not in patients with bipolar illness.
In the study of Egan et al.,24 it was shown that Val66Met polymorphism does not affect mature BDNF protein function but it significantly alters the intracellular trafficking and packaging of pro-BDNF. Val allele of BDNF polymorphism displayed increased activity in this respect. Therefore, Val/Val genotype can determine higher activity of the BDNF system with its possible consequences for neuronal function, for example, better cognitive performance. This would be compatible with the results of the current investigation. Patients with schizophrenia with Val/Val allele obtained better results on the N-back test, corrected responses. Also, patients with bipolar with Val/Val genotype performed significantly better on three of five domains of the WCST than those with Val/Met genotype. Corresponding findings have been recently achieved in a study of healthy Chinese women where those females with the Val/Val genotype had significantly higher performance intelligence quotient than Val/Met carriers.30 In light of evolutionary psychiatry, it may be speculated that Val allele, a connection of which with predisposition to bipolar illness has been suggested,16–19 may be at the same time associated with some aspect of cognitive advantage in such subjects.
The limitation of the current study may be deviation from the Hardy–Weinberg equilibrium for Val66Met polymorphism in patients with schizophrenia. In light of this, the difference between genotypes in respect to the N-back test should be regarded with caution. Also, variable psychopathological state and medication during cognitive testing could restrict the interpretation of cognitive results. However, all patients had symptoms of mild or mild–moderate intensity and none of them received high doses of psychotropic drugs during the period of study.
Bearing all these limitations in mind, it may be asked why the BDNF system is associated with a quality of performance on different prefrontal cognitive tests in schizophrenia and bipolar mood disorder? Although both the WCST and the N-back test measure aspects of prefrontal cortex activity there are some differences between their neuropsychological targets. The performance on the N-back test is to great extent dependent on a basic processing of visual information and on visuo-motor coordination. Cognitive deficits involving initial stages of information processing are well documented in schizophrenia but not in bipolar mood disorder.31 Therefore, in schizophrenia, the BDNF system could be more involved in these more primary cognitive processes. In contrast, the performance on the WCST is primarily dependent on the efficiency in planning and mental flexibility and not on the efficiency of processing of visual information. Therefore, in bipolar mood disorder, cognitive processes may be connected with the BDNF system on higher stages of information processing.
In conclusion, the current study suggests, on molecular-genetic ground, a possible role of the BDNF system for cognitive performance on the WCST in bipolar mood disorder but not in schizophrenia. An association of this polymorphism with performance on the N-back test in schizophrenia and not in bipolar illness may suggest that in schizophrenia, the BDNF system may be connected with early phases of information processing.