INDIVIDUALS WITH NON-CLINICAL schizotypy, a healthy population with schizotypal features but without any diagnosis of mental illness, have been found to have poorer Wisconsin Card-Sorting Test (WCST) performance in both adolescents and university students.1–3 Jahshan and Sergi, however, debated these aforementioned studies and found no executive functioning impairment in schizotypic undergraduates.4 Noguchi et al. also stated that no correlation was found between adults with schizotypy and WCST performance.5 Because both Jahshan and Sergi, and Noguchi et al. recruited more female participants, their results raised the possibility that gender might affect scale measurement of schizotypy.6 The aim of the present study was therefore to confirm the hypothesis that non-clinical gender-equivalent adults with higher positive and negative schizotypal traits will also exhibit poorer performance in the WCST.
Non-clinical schizotypy was found to be related to poorer Wisconsin Card-Sorting Test (WCST) performance, but the results were inconsistent. Two subgroups, the higher negative–higher positive and the lower negative–lower positive (15 vs 16), were selected from the top and the bottom quartiles of negative and positive scale scores of the Schizotypal Personality Questionnaire (SPQ) completed by 177 healthy volunteers, respectively. The higher negative–higher positive SPQ score subgroup had significantly poorer performance regarding the completed categories of WCST than the lower negative–lower positive SPQ score subgroup. Subjects with higher non-clinical schizotypy trait showed relatively mild cognitive dysfunction.
One hundred and seventy-seven participants (83 male, 94 female) were recruited in various community studies as healthy controls through research advertisements. The local ethics committee approved the study protocol. The Chinese version of the Mini International Neuropsychiatry Interview7 was used by a senior psychiatrist, who has been performing this interview for more than 10 years, to exclude individuals with mental illness. None of the participants had a history of mental illness, substance abuse or dependence (including cigarette smoking).
Wisconsin Card-Sorting Test
Schizotypal Personality Questionnaire
The Schizotypal Personality Questionnaire (SPQ) is a 74-item self-report scale based on the DSM-III-R criteria for schizotypal personality disorders, which has been used in dimensional research to identify the correlates of schizotypal features in the normal population.11 The internal consistency of the total SPQ score (0.90) is high in adults.12 Sarkin's model for the SPQ subtyping of the positive and negative domains was applied.13 We modified the grouping and chose two subgroups for further investigation, namely the higher negative–higher positive and the lower negative–lower positive. Fifteen participants with the highest 25th percentile of both positive and negative scale scores were categorized as the higher negative–higher positive SPQ score (SPQ-H) group; and 16 participants with the lowest 25th percentile of both positive and negative scale scores were categorized as the lower negative–lower positive SPQ score (SPQ-L) group.13 The positive scale consisted of perceptual aberration, magical ideation, and ideas of reference subscales; the negative scale consisted of constricted affect, no close friends, and social anxiety subscales.
The Mann–Whitney U-test and partial correlation analysis controlling for age between SPQ and WCST were performed. The age-controlled correlation analysis between SPQ and WCST in different genders was also performed. The threshold for statistical significance was set at P < 0.05.
The demographic data and WCST comparison of the subgroups are given in Table 1. The SPQ-H subgroup had a poorer performance in the completed categories of WCST (Mann–Whitney U = 70, Z = −2.02; P = 0.04). No significant correlation was found between the total SPQ scores and WCST performance in the whole group after controlling for age (n = 177, perseverative errors r = 0.02, P = 0.82; completed categories, r = −0.11, P = 0.30).
|Higher negative–higher positive SPQ score (n = 15)||Lower negative–lower positive SPQ score (n = 16)||Mann–Whitney U (Z)/χ2||P|
|Age (years) (mean ± SD)||36.73 ± 14.27||33.63 ± 8.55||127 (0.28)||0.78|
|Below senior high school degree||8||6|
|Above bachelor degree||7||10|
|SPQ subscale scores (mean ± SD)|
|Positive||9.20 ± 2.34||1.06 ± 0.85||240 (4.8)||<0.01**|
|Negative||12.33 ± 2.02||0.88 ± 0.81||240 (4.8)||<0.01**|
|WCST (mean ± SD)|
|Perseverative errors||14.13 ± 12.62||7.19 ± 4.65||159.5 (1.57)||0.12|
|Completed categories||2.13 ± 1.81||3.38 ± 1.20||70 (−2.02)||0.04*|
No significant age or educational level differences were found between the SPQ-H and the SPQ-L subgroups. The correlation between the positive and negative scales within the total participants was statistically significant (Spearman's rho, 0.31; P < 0.001).
The SPQ-H subgroup performed significantly poorer than the SPQ-L subgroup in the completed categories, but not in the perseverative errors of WCST. This is consistent with the findings of previous studies that non-clinical schizotypy individuals with higher SPQ scores could have poorer WCST performance.1,2
The perseverative errors is the most characteristic WCST feature of schizophrenia.14,15 Both perseverative errors and completed categories of WCST, however, are the most commonly used indices in neuroimaging studies, and the change of the completed categories was demonstrated to be more apparent than that of the perseverative errors.16 The present findings suggest that the difference in cognitive performance between the SPQ-H subgroup and the SPQ-L subgroup, might be presented by another index besides the perseverative errors of WCST. The completed categories of WCST might be more sensitive in reflecting the mild cognitive deterioration in certain subjects such as non-clinical schizotypy individuals in the present study.
Furthermore, the correlation between the positive and negative scales within the present study participants was statistically significant, and thus we assumed that a dimension, such as the model of Sarkin et al.,13 might be a suitable approach.
Schizotypic adolescents and university students with higher SPQ scores might have much more cognitive and biological vulnerability,17 and there is a high possibility that they would move further along the schizophrenia spectrum disorders before reaching adulthood.18 Such individuals who do not develop schizophrenia and other related disorders tend to have lower SPQ scores and less cognitive dysfunction compared to individuals who do develop schizophrenia. This might explain why adults with non-clinical schizotypy, as in the present sample, showed weak evidence of poor performance on the WCST. Because the present SPQ-L subgroup could be thought of as ultra-psychologically stable participants, it should be noted that the present result could be interpreted as a comparison between the ultra-psychologically stable participants and those with schizotypal traits.
Several limitations in the present study should be noted. First, WCST scores should be interpreted within the context of education or IQ,19 although years of education might be used in the clinical neuropsychological practice as an approximate estimate of IQ.20 Second, the present sample size for both subgroups was small. Third, regarding the subgrouping method, positive and negative dimensions of SPQ scores might possibly have independent influence on cognitive performance.
This research was supported by the Changhua Christian Hospital and the National Science Council (NSC 91–2314-B-006–074, NSC 93–2314-B-006–107, NSC 95–2314-B-006–115 MY2). The authors would like to thank Yee Ping Chen, Tsai-Hua Chang, Ching Ling Chu and Chien Ting Lin for their administrative help.