To Örebro University

Background: In clinical practice, the differential diagnosis of Asperger syndrome (AS) versus schizophrenia can be a challenge. Some self-report instruments—such as the Autism-spectrum Quotient (AQ)—have been portrayed as proxies for the diagnosis of AS. However, it has not been demonstrated to what extent autistic traits—as measured by the AQ—separate AS from schizophrenia.

Aim: To examine the AS–schizophrenia discriminating ability of the AQ.

Method: The AQ is a 50-item self-administered questionnaire (with score range 0–50) for measuring “autistic traits” in adults. Here, it was completed by 136 individuals: 36 with schizophrenic psychosis, 51 with AS and 49 non-clinical comparison cases. A receiver operating characteristic (ROC) analysis for the total AQ score was performed to examine the discriminating power of the instrument.

Result: Both individuals with schizophrenia and individuals with AS scored significantly higher on AQ than the non-clinical group. The mean total AQ score (± standard deviation) of the AS group (26.7 ± 8.9; range 9–44) was significantly higher than that of the schizophrenia group (22.7 ± 6.2; range 10–35) (P = 0.041). However, when using the full Likert scale for scoring, the difference did not reach significance. In the ROC analysis of total AQ scores for AS versus schizophrenia, the area under the curve (AUC) was 0.65 (P = 0.02).

Conclusion: Although mean AQ scores separated AS and schizophrenia at a group comparison level, significant overlap of AQ scores across the two diagnostic groups clearly reduces the discriminating power of the AQ in the separation of schizophrenia from AS.