Quantitative Electroencephalographic Profiles for Children With Autistic Spectrum Disorder Agnes S. Chan and Sophia L. Sze Chinese University of Hong Kong Mei-chun Cheung Hong Kong Polytechnic University The present study examined quantitative electroencephalographic (QEEG) profile for children with autistic spectrum disorder (ASD). Five-minute QEEG data were obtained from 90 normal controls (NCs) and 66 children with ASD. Spectrum analyses revealed that ASD children showed significantly less relative alpha and more relative delta than NC. Specifically, 26% of ASD children and 2% of NCs showed 1.5 SDs of relative alpha below the normative mean. Children with this QEEG profile had 17 times the risk of having ASD than those without such a profile. Sensitivity and specificity of relative alpha were 91% and 73%, respectively. Split-half cross-validation yielded a sensitivity of 76%. Keywords: autistic spectrum disorder, QEEG, screening, classification Autistic spectrum disorder (ASD) is a neurodevelopmental dis- order that is clinically characterized by impaired social interaction, language impairments, behavioral stereotypes, and some cognitive deficits. Given that an ASD-identifying laboratory test is not available, diagnoses are primarily based on detailed clinical inter- view and behavioral observation, such as the Autism Diagnostic Interview (Le Couteur et al., 1989), the Autism Diagnostic Obser- vation Schedule (Lord et al., 1989), the Childhood Autism Rating Scale (CARS; Schopler, Reichler, & Renner, 1986), and the Checklist for Autism in Toddlers (Wong et al., 2004). Although behavioral observation remains the major diagnostic tool, it may be confounded by interrater bias (Waller, Armstrong, McGrath, & Sullivan, 1999). Thus, some efforts have been made to develop neurobiological measures including MRI measures (Akshoomoff et al., 2004; Courchesne, 2004; Courchesne et al., 2001) and genetic testing (Folstein & Rosen-Sheidley, 2001; Lord & Volk- mar, 2002) that may provide more objective and sensitive diag- noses for ASD. However, reliable and affordable neurobiological measures have not been established. Given that there is an increasing trend in the prevalence rate of ASD worldwide (Bertrand et al., 2001; Chakrabarti & Fombonne, 2001) and that early intervention is critical to remediate the symp- toms, it is clinically significant to develop a biological screening test that is relatively easy and possible even for young children to perform. As previous research has suggested that as little as 1-min noise-free electroencephalographic (EEG) data would yield reli- able and valid data (John, Prichep, Fridman, & Easton, 1988), it is possible to apply the EEG technique to children with special needs who may not be able to sit still for a long time or comply easily. Thus, the aim of the present study is to examine if quantitative EEG (QEEG) assessment is sensitive and yet simple enough to differentiate children with ASD from normal children. QEEG is a type of electrophysiological assessment that applies computerized mathematical analysis to convert the raw waveform data into different frequency ranges including delta, theta, alpha, and beta. Each frequency range is averaged across a sample of data and quantified into mean amplitude (i.e., voltage in microvolts). The absolute power and relative power (i.e., percentage of total power) in each frequency band can be calculated. Given that the QEEG technique has the advantages of being less expensive, easier to perform, and noninvasive compared with some other neuroim- aging techniques (e.g., positron emission tomography and com- puted tomography), it has been advocated as a potential clinical assessment for neurological and psychiatric disorders (Hughes & John, 1999). Several groups of researchers have attempted to develop the QEEG technique as a screening assessment for neurodevelopmen- tal disorders (Daoust, Limoges, Bolduc, Mottron, & Godbout, 2004; Dawson, Klinger, Panagiotides, Lewy, & Castelloe, 1995; Gasser, Rousson, & Gasser, 2003; Monastra, Lubar, & Linden, 2001; Monastra et al., 1999). Among the various disorders, atten- tion-deficit/hyperactivity disorder (ADHD) seems to have been studied relatively more extensively (Barry, Clarke, & Johnstone, 2003; Bresnahan, Anderson, & Barry, 1999; Bresnahan & Barry, 2002; Chabot & Serfontein, 1996; Kuperman, Johnson, Arndt, Lindgren, & Wolraich, 1996; Monastra et al., 1999). In a recent study by Monastra et al. (2001), the QEEG profile of 469 individ- uals with ADHD was examined to develop a single-channel QEEG screening test for differentiating individuals with ADHD from normal subjects. They reported that the theta– beta ratio measured from the vertex (Cz) yielded a specificity of 94% and a sensitivity of 90%, and that the test–retest reliability was 0.96. So far, the empirical evidence on ADHD seems to support the notion that QEEG, even single channel, may have the potential to be devel- oped into a sensitive measure for children with neurodevelopmen- tal disorder. Agnes S. Chan and Sophia L. Sze, Department of Psychology, Chinese University of Hong Kong, Shatin, Hong Kong, China; Mei-chun Cheung, Intitute of Textiles and Clothing, Hong Kong Polytechnic University, Hung Hom, Hong Kong, China. The research was supported by Hong Kong Research Grant Council Grant CUHK 4648/05H. We thank the Parents’ Association of Pre-School Handicapped Children in Hong Kong for their assistance in recruiting subjects. Correspondence concerning this article should be addressed to Agnes S. Chan, Department of Psychology, Chinese University of Hong Kong, Shatin, New Territories, Hong Kong, China. E-mail: aschan@psy.cuhk.edu.hk Neuropsychology Copyright 2007 by the American Psychological Association 2007, Vol. 21, No. 1, 74 – 81 0894-4105/07/$12.00 DOI: 10.1037/0894-4105.21.1.74 74