Oral 2

[O-2-01] Evaluating the Factor Structure of the 25-item Autism-Spectrum Quotient Japanese Version as a Potential Screening Tool in Perinatal Women

*Ekachaeryanti Zain^1,2, Naoki Fukui¹, Yuichiro Watanabe^1,3, Koyo Hashijiri¹, Takaharu Motegi¹, Maki Ogawa¹, Jun Egawa¹, Koji Nishijima⁴, Toshiyuki Someya¹ (1.Department of Psychiatry, Niigata University Graduate School of Medical and Dental Sciences(Japan), 2.Department of Psychiatry, Faculty of Medicine, Mulawarman University(Indonesia), 3.Department of Psychiatry, Uonuma Kikan Hospital(Japan), 4.Department of Obstetrics and Gynecology, Niigata University Graduate School of Medical and Dental Sciences(Japan))

Keywords:

Autism-Spectrum Quotient (AQ),autism spectrum conditions (ASC),autism spectrum disorder (ASD),perinatal women

Background: Autism screening in pregnant women is increasingly recognized as vital in Japan for promoting perinatal health and supporting healthy motherhood. Objective: This study aimed to identify the optimal factor structure of the Japanese version of the Autism-Spectrum Quotient (AQ-J) as a prospective screening tool for autistic traits in perinatal women. Methods: The AQ-J was administered to 4,287 Japanese perinatal women (currently pregnant or within one month postpartum). Participants were divided into two groups for analysis. Exploratory factor analysis (EFA) was conducted on the first group (n = 2,154) to identify factor structures and refine final item selection. Confirmatory factor analysis (CFA) was performed on the second group (n = 2,133) to validate the model and assess fit indices. The resulting model was compared with 13 previously proposed models to determine the best fit. Results: The EFA identified a 25-item model comprising three factors: “Social Interaction” (9 items), “Non-Verbal Communication” (11 items), and “Restricted Interest” (5 items). Cronbach’s alpha coefficients were 0.829 for the total AQ-J, 0.829 for “Social Interaction,” 0.755 for “Non-Verbal Communication,” and 0.576 for “Restricted Interest.” McDonald’s omega coefficients (95% confidence intervals) were 0.826 (0.821–0.836) for the total scale, 0.835 (0.821–0.837) for “Social Interaction,” 0.755 (0.744–0.766) for “Non-Verbal Communication,” and 0.603 (0.556–0.596) for “Restricted Interest.” The CFA confirmed acceptable fit indices for the three-factor structure (goodness-of-fit index: 0.900, comparative fit index: 0.860, root mean square error of approximation: 0.066), indicating it provided a superior fit compared to all 13 previously proposed models. Conclusion: The 25-item AQ-J model is suitable for assessing autistic traits in perinatal women and outperforms previous models. Further research involving clinical populations is needed to establish a cutoff score, enhancing the utility of the 25-item AQ-J for screening autistic traits in non-clinical perinatal populations.