The purpose of this study was to investigate whether the individuals in the general population with high scores on the Autism Spectrum Quotient constituted a single homogeneous group or not. A cohort of university students (n = 4901) was investigated by cluster analysis based on the original five subscales of the Autism Spectrum Quotient. Based on the results of the analysis, the students could be divided into six clusters: the first with low scores on all the five subscales, the second with high scores on only the 'attention to detail' subscale, the third and fourth with intermediate scores on all the subscales, the fifth with high scores on four of the five subscales but low scores on the 'attention to detail' subscale and the sixth with high scores on all the five subscales. The students with high total Autism Spectrum Quotient scores (n = 166) were divided into two groups: one with high scores on four subscales but low scores on the 'attention to detail' subscale and the other with high scores on all the five subscales. The results of this study suggested that individuals from the general population with high Autism Spectrum Quotient scores may consist of two qualitatively different groups.

The AQ measured at admission was correlated with the UPI score, regardless of the sex or department; in the second survey, the scores decreased significantly in those with high AQ values at admission, suggesting that an unstable mental state can produce a temporary increase of the AQ scores.

Background: Symptom overlap between autism spectrum disorders (ASD) and social anxiety disorder has been reported. However, few reports exist on the associations between autistic traits and social anxiety in young adults. Methods: Autistic traits and social anxiety were assessed using the Autism-Spectrum Quotient (AQ) and the Liebowitz Social Anxiety Scale (LSAS), respectively. The survey was performed as part of the student health check-up at admission to our school, Kochi University. The relationship between the AO and LSAS scores were analyzed. Results: In all, 1479 students (77.2%) completed both the questionnaires. The overall AQ score was significantly related to the score on the Japanese version of the LSAS (LSAS-J) (Spearman’s correlation coefficient; 0.46, p<0.001). The group of students with AQ≥33 had significantly higher total and subscale scores on the LSAS-J as compared to the group with AQ<33, while the group with LSAS-J≥44 showed significantly higher scores on all the subscales of the AQ, except “Attention to detail,” as compared to the group with LSAS-J<44. Conclusions: The results suggest that both students diagnosed as having ASD and those with autistic traits may experience anxiety symptoms. The score for “Attention to detail” on the AQ, which was not related to the presence of social anxiety, may be useful in discriminating between autistic traits and social anxiety.