International Child Development Resource Center

POSTER SESSION 1 TOPIC 1: PARENT-CHILD RELATIONS P1.1.1
MATERNAL ADAPTATION: EFFECTS OF AD CHILD TEMPERAMENT AND COMPETENCE. M.M. Konstantareas and V.Papageorgiou.

Temperament has been shown to impact development. Little is available on the temperamental characteristics of children with Autistic Disorder (AD)or how they may affect maternal adaptation. These issues were examined in a study employing Greek children with AD and their mothers. The mothers of 36 male and 7 female children, meeting criteria for AD on the CARS (Mean=44), were given the Dimensions of Temperament Scale (DOTS-R) and the Clarke QRS, both translated into Greek and back-translated. As well, the CARS and the Psychoeducational Profile (PEP) were used. The children were assessed for verbal nonverbal status and for degree of independence (low-average-high) by a Child Psychiatrist. Most displayed low to moderate independence and half were verbal. Results revealed that the PEP scores were positively related to clinican-rated independence(r=.312, p P1.1.2
ATTACHMENT SECURITY IN AUTISM: CORRELATES AND SEQUELAE. A. Rozga, M. Sigman* and L. Beckwith. Department of Psychiatry, University of California Los Angeles, Los Angeles, CA 90024.

The aim of this short-term longitudinal study was to investigate cognitive and social correlates and sequelae of attachment security in a group of children with autism. Twenty-nine children with autism were observed in Ainsworth’s Strange Situation, as well as interacting with their mothers in the home. Almost half (45%) of the children were classified as securely attached, and this group demonstrated greater receptive and expressive language ability and mental level than the insecurely attached group, concurrently as well as during a one-year follow-up. Securely attached children also demonstrated greater gains in empathic ability over the course of one year than children who received insecure classifications. Mothers of children classified as securely attached displayed greater sensitivity than mothers of children who were insecurely attached. Children classified as securely attached more frequently initiated social interactions with their mothers, and were more responsive to Mothers’ bids for social interaction than children classified as insecurely attached. These results indicate that although child functional level may impact attachment security and maternal sensitivity in this population, the interactional antecedents and consequences of attachment security in autism are similar to those reported for typically developing children.

P1.1.3
MATERNAL ATTITUDES ABOUT PARENTING A YOUNG CHILD WITH AUTISM. M. Siller, A. Bundy and M. Sigman*. Department of Psychology, University of California at Los Angeles, Los Angeles, CA 90024.

Parenting a child with autism is a very complex task. Many of the developmental milestones that come natural to typically developing children require the parent’s special attention in the case of autism. This study was concerned with how mothers prioritize the different tasks involved in parenting a child with autism. Mothers of 23 young children with autism (CA = 71 months; LA = 40 months) were presented with 14 self-descriptive statements, each emphasizing a different domain of parenting (i.e. promoting independence, assuring services, teaching academic skills, promoting communication skills, promoting peer/ sibling interactions, providing emotional support). Mothers were asked to rank order these statements according to how descriptive they are of their role as being a mother of a child with autism. The results showed that priority was given to assuring services and interventions, providing emotional support and promoting communication skills. On the other hand, promoting academic skills, promoting peer interactions, and sharing the child’s interest were considered least important. In addition we found evidence that the maternal attitudes were dependent on the child’s developmental level. Mothers of children with relatively advanced language skills emphasized the need to promote socially acceptable behavior and peer/sibling interactions while mothers of children with relatively little language skills emphasize the need to follow the child’s lead and to promote the child’s requesting skills. The implications of these findings for parent training interventions will be discussed.

P1.1.4
FACILITATIVE FACTORS OF THE HOME ENVIRONMENTS OF CHILDREN WITH AUTISM. C. Zierhut

The present study was designed to identify the predictive characteristics of the home environment that were associated with gains in nonverbal communication, language skills, play, and empathy skills over the course of one year in very young children with autism. The central organizing hypothesis is that, although autism is caused by biological determinants, environmental factors contribute to those social communication skills, which crucially determine the later development of children with autism. The specific hypothesis examined was that the optimal social and communicative development of children with autism requires both responsive caregiving and the provision of structured experiences. The study found that caregivers of children who were more sensitive and responsive in their interactions, had children who developed superior language skills over the course of a year, than did children of caregivers who showed less sensitivity and responsivity initially. The identification of these facilitative factors of social-emotional and communication skills is crucial for parent training interventions.

POSTER SESSION 1 TOPIC 1: PARENT-CHILD RELATIONS

P1.1.1
MATERNAL ADAPTATION: EFFECTS OF AD CHILD TEMPERAMENT AND COMPETENCE. M.M. Konstantareas and V.Papageorgiou.

Temperament has been shown to impact development. Little is available on the temperamental characteristics of children with Autistic Disorder (AD)or how they may affect maternal adaptation. These issues were examined in a study employing Greek children with AD and their mothers. The mothers of 36 male and 7 female children, meeting criteria for AD on the CARS (Mean=44), were given the Dimensions of Temperament Scale (DOTS-R) and the Clarke QRS, both translated into Greek and back-translated. As well, the CARS and the Psychoeducational Profile (PEP) were used. The children were assessed for verbal nonverbal status and for degree of independence (low-average-high) by a Child Psychiatrist. Most displayed low to moderate independence and half were verbal. Results revealed that the PEP scores were positively related to clinican-rated independence(r=.312, p P1.1.2
ATTACHMENT SECURITY IN AUTISM: CORRELATES AND SEQUELAE. A. Rozga, M. Sigman* and L. Beckwith. Department of Psychiatry, University of California Los Angeles, Los Angeles, CA 90024.

The aim of this short-term longitudinal study was to investigate cognitive and social correlates and sequelae of attachment security in a group of children with autism. Twenty-nine children with autism were observed in Ainsworth’s Strange Situation, as well as interacting with their mothers in the home. Almost half (45%) of the children were classified as securely attached, and this group demonstrated greater receptive and expressive language ability and mental level than the insecurely attached group, concurrently as well as during a one-year follow-up. Securely attached children also demonstrated greater gains in empathic ability over the course of one year than children who received insecure classifications. Mothers of children classified as securely attached displayed greater sensitivity than mothers of children who were insecurely attached. Children classified as securely attached more frequently initiated social interactions with their mothers, and were more responsive to Mothers’ bids for social interaction than children classified as insecurely attached. These results indicate that although child functional level may impact attachment security and maternal sensitivity in this population, the interactional antecedents and consequences of attachment security in autism are similar to those reported for typically developing children.

P1.1.3
MATERNAL ATTITUDES ABOUT PARENTING A YOUNG CHILD WITH AUTISM. M. Siller, A. Bundy and M. Sigman*. Department of Psychology, University of California at Los Angeles, Los Angeles, CA 90024.

Parenting a child with autism is a very complex task. Many of the developmental milestones that come natural to typically developing children require the parent’s special attention in the case of autism. This study was concerned with how mothers prioritize the different tasks involved in parenting a child with autism. Mothers of 23 young children with autism (CA = 71 months; LA = 40 months) were presented with 14 self-descriptive statements, each emphasizing a different domain of parenting (i.e. promoting independence, assuring services, teaching academic skills, promoting communication skills, promoting peer/ sibling interactions, providing emotional support). Mothers were asked to rank order these statements according to how descriptive they are of their role as being a mother of a child with autism. The results showed that priority was given to assuring services and interventions, providing emotional support and promoting communication skills. On the other hand, promoting academic skills, promoting peer interactions, and sharing the child’s interest were considered least important. In addition we found evidence that the maternal attitudes were dependent on the child’s developmental level. Mothers of children with relatively advanced language skills emphasized the need to promote socially acceptable behavior and peer/sibling interactions while mothers of children with relatively little language skills emphasize the need to follow the child’s lead and to promote the child’s requesting skills. The implications of these findings for parent training interventions will be discussed.

P1.1.4
FACILITATIVE FACTORS OF THE HOME ENVIRONMENTS OF CHILDREN WITH AUTISM. C. Zierhut

The present study was designed to identify the predictive characteristics of the home environment that were associated with gains in nonverbal communication, language skills, play, and empathy skills over the course of one year in very young children with autism. The central organizing hypothesis is that, although autism is caused by biological determinants, environmental factors contribute to those social communication skills, which crucially determine the later development of children with autism. The specific hypothesis examined was that the optimal social and communicative development of children with autism requires both responsive caregiving and the provision of structured experiences. The study found that caregivers of children who were more sensitive and responsive in their interactions, had children who developed superior language skills over the course of a year, than did children of caregivers who showed less sensitivity and responsivity initially. The identification of these facilitative factors of social-emotional and communication skills is crucial for parent training interventions.

POSTER SESSION 1 TOPIC 2: ANIMAL MODELS

P1.2.1
NEURAL AND BEHAVIORAL CONSEQUENCES FOLLOWING EMBROYONIC EXPOSURE TO SODIUM VALPROATE IN MICE. M.Cheh, A.K. Halladay, P.McRae and G.C. Wagner. Rutgers University, Piscataway, NJ 08854.

It has been reported that gestational exposure to sodium valproate (VPA) produces changes in cerebellar morphology similar to that seen in autistic children. The goal of the current experiments was to determine the effects of prenatal VPA on neurobehavioral development of mice. Both male and female offspring of BALB/c mice treated with 600 mg/kg VPA on gestational day 13 were tested daily on PND 4-26 for either surface or mid-air righting, negative geotaxis, grip-strength, habituation of motor activity, and acquisition of a water-maze task. Prenatal VPA exposure resulted in a significant delay in the appearance of both the surface and mid-air righting response, both of which have been suggested to be mediated by cerebellar function. Both male and female VPA- exposed offspring did not show significant improvement in the development of a hippocampal-dependent water maze task. However, other skills such as negative geotaxis and habituation of motor activity were unaffected by prenatal VPA exposure, suggesting that the deficts were specific and not due to global impairments in functioning. The delay in surface righting ability was accompanied by a concomitant decrease in the levels of cerebellar synaptophysin and GFAP on PND5. These results suggest that prenatal exposure to VPA results in behavioral teratogenicity, which may be mediated by changes in the expression of proteins important for synaptogenesis, axonal sprouting, and structural support during brain development.

P1.2.2
THIMEROSAL TOXICITY IN INFANT MICE: APPROXIMATION OF HUMAN INFANT VACCINATION CONDITIONS B. Done, J.B. Adams, and E. Castaneda, Arizona State University, Tempe, AZ 85287-6006

Thimerosal is a mercury-based preservative widely used in childhood vaccines, and there are suspicions that it may contribute to the etiology of some cases of autism since it is a known neurodevelopmental toxin. A research study was undertaken to more fully assess the toxicity of thimerosal, for the conditions experienced by human infants. This study involved the use of a mouse model, Mouse Crl: CD-1 (ICR) BR from Charles River Labs. This mouse is commonly used for toxicological studies because it is an outbred strain (wide genetic variability, which is more typical of human populations than inbred strains). This study first involved the determination of the LD50 for single acute doses in adults compared to 4-day-old neonates. Next, it compared single acute doses vs chronic doses. The acute phase involved injections at age four days and the chronic injections occur at days 4, 8, 12 and 16. This approximates the human vaccination schedule (months 2, 4, 6, and 12). The effect of oral antibiotics was investigated because it has been shown that their use increases the half-life for the excretion of mercury from 10 days to over 100 days in rats. The effect of aluminum was investigated because it is included in some vaccines, and Haley has demonstrated that it increases the toxicity of thimerosal.

P1.2.3
EFFECTS OF POSTNATAL EXPOSURE TO SODIUM VALPROATE IN MICE: A NEW ANIMAL MODEL OF AUTISM. A.K. Halladay, P. McRae, M. Cheh and G.C. Wagner*. Department of Psychology, Rutgers University, Piscataway, NJ, 08854.

Early exposure to environmental agents may have a role in the etiology of pervasive developmental disorders (PDD) including autism. Such exposure may cause neurobehavioral retardation, regression, or intrusions. Teratogenic effects of the antiepileptic drug sodium valproate (VPA) have been well documented in the form of decreases in Purkinje cell number which, in turn, have been associated with autism. In addition, gestational exposure to VPA has been linked to autistic-like behavior in children. In the present study, PND 14 male BALB/c mice were tested for skills corresponding to cerebellar development. Following initial testing, pups were treated with 200 or 400 mg/kg sc VPA and retested on these same tasks 1 to 5 days later, and also examined for emergence of habituation of motor activity and spatial learning ability on PND 23-26. Treatment with VPA on day 14 produced a loss of the mid-air righting reflex as well as the negative geotropic response; that is, while pups had developed these abilities before treatment, these behaviors were lost for up to 72 hours after VPA administration (defined here as regression). These behavioral deficits were accompanied by an increase in serotonin turnover in the striatum and hippocampus. In addition, treated pups showed a delay in acquisition of the water maze task (retardation) with no alterations in the habituation of motor activity. These observations are discussed in the context of a new behavioral model for PDD which can be used to assess the role of known teratogens on behavioral development.

P1.2.4
QUANTIFICATION OF SELF-INJURIOUS BEHAVIOR IN AN ANIMAL MODEL OF PEMOLINE-INDUCED SELF-INJURY S. D. Kies and D. P. Devine Dept of Psychology, University of Florida, Gainesville, FL 32611

Self-injurious behavior (SIB) is a devastating behavior disorder in which an individual inflicts tissue damage by various methods, such as self-biting, skin-picking, and head-banging. This maladaptive behavior disorder is often chronic and stereotypic, and SIB often co-occurs with other stereotypic behaviors. A sub-population of autistic and intellectually handicapped individuals exhibits this maladaptive behavior disorder, wherein the severity of the self-injury ranges from mild to life-threatening. Several animal models have been developed to investigate the neurobiology of this behavior disorder, and these animal models have consistently been evaluated by examining self-inflicted tissue trauma using qualitative scales. Pemoline, an indirect dopamine agonist, produces self-inflicted tissue damage in rats. However, we have previously found that some, but not all, rats self-injure when treated with moderately high doses of pemoline. Furthermore, the rats that self-injured exhibited great variability in self-induced tissue damage. We have investigated the potential that we can more accurately describe the severity of SIB in individual rats by actually quantifying the behavior rather than the resulting tissue trauma. Accordingly, we measured the frequency and duration of self-injurious contacts between the mouth and specific body regions, and evaluated the relationship between this behavior and the extent of tissue damage. Our method of quantifying self-injurious behavior will increase the utility of animal models of SIB, as we define their actual behaviors more precisely.

P1.2.5
THE CEREBELLUM AND SOCIAL BEHAVIOR: IS THERE A CONNECTION? L.A. Martin(1)*;D. Goldowitz(1); and G. Mittleman(2) 1. Dept. Anat & Neurobiology, Univ Tenn. Health Sci. Ctr., Memphis, TN, 38163 2. Dept. of Psych., Univ. Memphis, Memphis, TN, 38152

Cerebellar abnormality has been linked to autism through both the loss of Purkinje cells demonstrated in post-mortem studies of autistic brains and the hypoplasia of cerebellar lobules shown in MRI studies of autistic patients. Traditionally considered solely a motoric structure, recent research has demonstrated new functions for the cerebellum that extend beyond motor control. One of the defining characteristics of autism is an impairment of social interaction. In order to study the potential link between cerebellar neuropathology and abnormal social behavior, we created a mouse model with varying loss of cerebellar Purkinje cells by making aggregation chimeras between heterozygous lurcher (Lc/+) mutant and wildtype (+/+) embryos. Lc/+ mice undergo a complete loss of Purkinje cells during postnatal development as a result of a mutation in the 2 glutamate receptor gene. Chimeric mice have varying numbers of Purkinje cells, depending on the degree of incorporation of the +/+ lineage. In contrast to the ataxia shown by Lc/+ mice, chimeras have no obvious motor dysfunction. Using this chimeric model of cerebellar neuropathology, we tested mice in a series of investigations into their social behavior. Specifically, we tested Lc/+, chimeric, and +/+ mice in social intruder, social dominance, and social memory tests to detect differences in these various social behaviors that can be attributed to Purkinje cell function. Previously, we reported on our investigation of the behavioral effects of Purkinje cell loss on motivation, working memory, and spatial navigation abilities. Here we present our analysis of the relationship between Purkinje cell loss and social behavior.

P1.2.6
IMPROVEMENT BY REPEATED ADMINISTRATION OF 6R-TETRAHYDROBIOPTERIN OF THE MONOAMINE DEPLETION-INDUCED ABNORMAL BEHAVIORS IN IMMATURE RATS. H. Mizuma1, K. Fukuoka1, S. Nozaki2, H. Iizuka1, A. Kabasawa1, H. Tohyama1, N. Nishimura1, Y. Watanabe2 and R. Kohashi11Dept. Clini. Pathol., Kyorin Univ. Sch. Health Sci., Hachioji, 192-8508, Japan 2Dept. Physiol., Osaka City Univ. Gradu. Sch. Med., Osaka, 545-8585, Japan

Background. 6R-L-erythro-5,6,7,8-Tetrahydrobiopterin (6R-BH4) is a natural cofactor for rate-limiting enzymes of dopamine (DA) and serotonin (5-HT) biosynthesis. 6R-BH4 is also a cofactor for NO synthase. Administration of 6R-BH4 has been reported to improve the clinical symptoms in certain cases of several neurological disorders, such as Parkinson’s disease, depression and infantile autism. This study was attempted to clarify the therapeutic effects of 6R-BH4 on the abnormal behaviors induced by neonatal dopaminergic or serotonergic denervation in immature rats. Methods. 6-Hydroxydopamine (6-OHDA) or 5,7-dihydroxytryptamine (5,7-DHT) were injected into the lateral ventricle on neonatal period. 6R-BH4 (20mg/kg) was administered intraperitoneally for 1week (once a day) in immature rats. In the 4 week-old rats, the locomotion activities wereassessed during 20 min in open-field apparatus. Results and Interpretation. Locomotion activities increased in 6-OHDA-treated rats and decreased in 5,7-DHT-treated rats (p P1.2.7
NEUROTRANSMITTER AND GENE EXPRESSION CHANGES IN THE CENTRAL AMYGDALA AFTER SECRETIN TREATMENT IN RATS. M. Goulet*, C. Ware, R.Strong, J. Tay, P. Shiromani1, R. Boismenu and J. Rusche. Repligen Corp., Waltham, MA 02453; 1West Roxbury VA-Harvard, MA 01451.

Recent observations in a comprehensive clinical study showed improvements in autism symptoms (ADOS social interaction scores) specifically in 3-4 year old children receiving secretin. We explored the effect of secretin on the rat brain by examining gene activation in neurons, defining brain regions that expressed secretin receptor, and neurotransmitter changes in specific brain regions. We showed previously that secretin induces Fos protein in the central amygdala (CeA) and the area postrema in rats. The peak blood levels of secretin that induce Fos protein within the CeA are equivalent to the peak blood levels observed during pharmacologic, human use of secretin. Secretin infusion did not result in Fos expression in the NTS or PVN, brain regions normally activated in a response to stress. Secretin receptor (secR) is only present in the AP and cerebellum of adult rats but expressed in many other brain regions in juvenile rats. QPCR has been used to document secretin, secR and other gene expression changes post secretin dosing. Noradrenaline levels were decreased in the amygdala by secretin treatment. There was no effect on 5-HT. Neurotransmitters from brain regions containing CeA neural projections have also been measured. The presence of secR mRNA in multiple regions of juvenile rat brain suggests an important role during development. Neurotransmitter and gene expression changes in the amygdala by secretin may indicate a site of physiologic and clinical significance in treatment of autistic children.

POSTER SESSION 1 TOPIC 3: TREATMENT & OUTCOMES 1

P1.3.1
PREDICTORS OF EDUCATIONAL PLACEMENT FOR SCHOOL AGE CHILDREN WITH AUTISM SPECTRUM DISORDERS (ASD). J. Charles*, J. Nicholas, D. Conway, L. Horowitz. Dept. Pediatrics, Med. Univ. S. Carolina, Charleston, SC 29425.

The objective of this study was to determine predictors of educational placement of school age children with ASD and to assess demographics of a cohort of children with ASD in a tertiary developmental clinic. 109 children born 1991-93 with ICD-9 codes 299.00 or 299.80 were identified from clinic records. Demographics, early intervention history, cognitive testing, and current class placement were abstracted from the active clinic charts (86). Univariate logistic regression was used to obtain crude odds ratios between placement in regular education and participation in the following services: Part C (70, 30 with IQ70 (odds ratio 7.159, p=0.004). Effects of different levels of intervention did not reach significance in the small sample available. Only 9% of children with IQ>70 received Part C services, compared to 20% with IQ70 and 77% with IQ70 was the best univariate predictor of regular education placement. The low percentage of children (especially higher functioning) receiving Part C services emphasizes the importance of early diagnosis of ASD.

P1.3.2
COMPARING TWO EARLY INTERVENTION METHODS FOR CHILDREN WITH AUTISM: FULL-DAY PRESCHOOL AND HOME-BASED APPLIED BEHAVIOR THERAPY. R. E. Daniels*, D. Breiger, D. Rubovits, C. Oliver and A. Maxwell. Children’s Clinic, LLC, Chicago, IL 60614, University of Washington School of Medicine, Seattle, WA, 98105, Jewish Children’s Bureau, Chicago, IL 60606, Behavior Consultation Services, Inc., Chicago, IL 60622.

This prospective treatment outcome study is one of the first investigations comparing the efficacy of two intervention programs for young children with autism. Specific aims: 1. To develop a methodology to assess the efficacy of early intervention programs for children with autism; 2. To determine the degree of improvement observed in children who attend a full-day therapeutic preschool (TEACCH model); and 3. To compare the efficacy of the preschool program to a home based applied behavior analysis therapy (ABA) program. Fifteen children (35 to 50 months of age) across both programs were evaluated every 3 months by an independent, clinical psychologist using parent observation checklists and direct, standardized assessments measuring cognitive functioning, adaptive behavior, communication skills, visual perceptual skills, fine motor coordination, and atypical behavior. Results from the first 12 months of the study will be presented. Strengths and weaknesses of the research methodology and implications for future outcome research will be discussed.

P1.3.3
SOCIAL INTERACTION SKILLS OF CHILDREN WITH AUTISM: TEACHING SELF-INITIATIONS WITH A NATURAL ENVIRONMENT. A.L. Donaldson & L.B. Olswang, Child Language Laboratory, Department of Speech & Hearing Sciences, University of Washington, Seattle, WA 98105

A defining feature of autism is impairment in social interaction skills. One expression of this deficit is a marked difficulty initiating communication with others. Employing a hierarchical treatment model, young children with autism were taught to self-initiate with classroom peers using three types of initiations (attention-getting/greeting, commenting and requesting nformation). Three Kindergarten-aged children with a diagnosis of autism participated in the study. Classroom teachers nominated the children based on their demonstrated social skill deficits and difficulty initiating with peers. The investigation utilized a multiple baseline across behaviors (initiation types) time series design to determine the effects of direct self-initiation training on the participants’ peer interactions and overall appearance of social competency. Following treatment all participants demonstrated an overall increase in self-initiations with peers across two classroom activities (snack and free-choice). They also demonstrated increased interest in peer activities and classroom teachers noted an overall increase in the participants’ social competency when interacting with peers and developing peer relationships.

P1.3.4
PHARMAUTISME (RV): A QUALITY ORIENTED PHARMACOTHERAPY SYSTEM FOR PROGRAMS SERVING VULNERABLE PSYCHIATRIC PATIENTS Joaquin Fuentes, M.D., GAUTENA and Policlínica Gipuzkoa, San Sebastian, Spain & Andrés Martin, M.D., M.P.H., Yale Child Study Center, New Haven, Connecticut

Persons with autism and other developmental disorders can be seen as potentially vulnerable patients for pharmacotherapy, as medical systems are rarely prepared to adapt their practice to their specific needs. Thus, in many countries, these patients receive unchallenged long-term and/or unnecessarily high dosages of psychotropic drugs, without the appropriate regard to professional accountability, informed consent (by the person him or herself and/or his or her legally authorized representative), established indications, systematic consideration to side effects and possible interactions, and dissemination of essential data about the medication used, among involved non-medical professionals. The Pharmautisme system addresses all these educational and ethical issues, and has been instrumental as part of an ISO/IQ NET quality registered program in Spain, to rationalize the practice of pharmacotherapy in this population. In the five years it has been in use, the program has contributed to a decrease in psychotropic drug use to 28% of the clients of all ages served in a community program, a figure much lower than usually reported. The present refinement of this instrument involves a transnational collaboration to ensure international applicability of the system, and to grant its dissemination among the medical community around the Spanish-speaking world.

P1.3.5
THE EFFICACY OF MELATONIN ON SLEEP PROBLEMS IN CHILDREN WITH AUTISTIC SPECTRUM DISORDERS AND FRAGILE X S. Jacquemont, B. Goodlin-Jones, S. Bacalman, M. Sasan, RJ. Hagerman, T. Anders.

A significant number of children who carry a diagnosis of Autistic Spectrum Disorder or fragile X are reported to have sleep disturbances that contribute to parental stress and may impact the children’s overall functioning. The objective of this study was to determine whether melatonin would improve sleep problems in children with ASD or fragile X syndrome. Parents completed a Sleep Habits Questionnaire that asked about their child’s usual sleep patterns. Those children who had a sleep onset of greater than 30 minutes and/or frequent middle of the night waking were included in the study. The study design involved a double blind crossover and was conducted over a 5 -week period. Baseline data was collected during the first week. The remaining four weeks were divided into two phases-during the first two weeks children took a capsule containing either active medication or a placebo and then received a 2-week crossover period. Data regarding the children’s sleep patterns were gathered using a sleep diary completed by the parents and an actiwatch. The actiwatch monitors activity and the children wore the watch in a specially designed ankle bracelet 24 hours per day for the entire duration of the study. These instruments enhance the description of children’s sleep patterns. Our preliminary data on 10 patients shows significant sleep improvement on 2 outcome measures: sleep latency and time of sleep onset. Sleep duration was not significantly improved on pooled (10 patients) data. However, improvement of the sleep duration was significant in 2 patients when the data was analyzed individually, as a repetitive measure. There were no significant side effects reported. This study provides preliminary data on the effectiveness of melatonin in treating sleep disturbances in children with autistic spectrum disorder and/or Fragile X. Further research results will describe what type of sleep problem is most responsive to this medication in children with ASD or Fragile X.

P1.3.6
THE EFFECT OF RECIPROCAL IMITATION TRAINING ON IMITATIVE AND SPONTANEOUS PLAY IN CHILDREN WITH AUTISM. B. Ingersoll* and L. Schreibman. Autism Research Lab., Univ. of Calif., San Diego, CA 92093.

Children with autism exhibit deficits in their use of both imitative and spontaneous play. Play deficits are most pronounced in the use of pretend play actions. A multiple-baseline design across five participants was used to investigate the effect of reciprocal imitation training, a naturalistic, behavioral intervention, on the development of imitative and spontaneous pretend play in young children with autism. Results indicate that reciprocal imitation training increased the total number of pretend play actions used by the children during treatment. In addition, several of the participants also exhibited an increase in their use of pretend actions in their spontaneous play which suggests that actions acquired during imitative play were incorporated into the children’s spontaneous play. These results indicate that pretend play can be taught effectively using in a naturalistic, play-based intervention.

P1.3.7
NEUROFEEDBACK AS PART OF TOTAL THERAPY IN AUTISM SCHOOL: PILOT PROJECT. B.Jarusiewicz. Atlantic Research Institute, Inc., Atlantic Highlands, NJ 07716.

This pilot project was established to determine if Neurofeedback increased therapy progress of autistic students. The principal selected one class of six students, ages 8 and 9 for participation. The parents, after demonstration, were given the opportunity to have their child participate or not. Five participated. The school compared progress before and after NFB by reviewing on-going charts before and after NFB. NFB is being provided approximately twice a week for an expected total of 20 sessions. The NFB providers used the Autism Treatment Evaluation Checklist (ATEC at www.autism.com), a major issue checklist and other data forms developed by the principal investigator for before and after measurements. The school and principal investigator will meet sometime before school begins in the fall to determine if sufficient data was developed to determine if NFB will be used on a trial basis at the school.

P1.3.8
USE OF COMPLEMENTARY AND ALTERNATIVE MEDICINE AMONG FAMILIES OF CHILDREN DIAGNOSED WITH AUTISM. D Mandell, J Pinto-Martin, M Souders, E Giarelli, J Bloch, S Levy.

Several descriptive reviews have been written about Complementary and Alternative Medicine(CAM) for children with autism and how clinicians should respond to their use, but there is a paucity of research on how frequently these treatments are used and child and family characteristics associated with their use. The purpose of this study was to estimate and describe the use of CAM with children with autism. Chart review was conducted for the first 500 consecutive children seen for assessment and diagnosed with autism at The Children’s Hospital of Philadelphia, between May 1, 2001 and April 30, 2002. Caregivers responses to the question, “what other types of treatment have you tried for your child?” were abstracted from medical records and encoded into mutually exclusive categories, including different forms of vitamin supplements, over-the-counter medications, specialized diets, chelation and others. The prevalence of different CAM strategies and associated child and family and characteristics were estimated. Review of the first 24 charts suggests that 21% of families report some use of CAM. Eight percent reported use of gluten/casein-free diets and 13% report use of vitamins or other supplements (i.e., melatonin and DMG). Complete results on all 500 subjects will be presented, along with variables associated with the use of CAM. The high proportion of those reporting use of CAM has important implications for treatment, education and research. Chart review most likely under-represents use of CAM, since families may be unwilling to report treatment strategies of which they think clinicians will disapprove. Clinicians should ask families about their use of alternative therapies and create a nonjudgmental atmosphere in which families feel comfortable disclosing. Families should receive education concerning what is known about different treatments. Finally, research agendas should include studies of the efficacy and iatrogenic effects of different widely used treatments, as well as information on their interactions with other therapies.

P1.3.9
A RETROSPECTIVE ASSESSMENT OF CITALOPRAM IN CHILDREN AND ADOLESCENTS WITH PERVASIVE DEVELOPMENTAL DISORDERS R. Nicolson*, J. Couturier. Department of Psychiatry, University of Western Ontario, London, Ontario, Canada. N6H-3W7.

Introduction: Although selective serotonin reuptake inhibitors (SSRI’s) have been used to treat symptoms of aggression and anxiety in children and adolescents with pervasive developmental disorders (PDD’s), there are no published reports of the use of citalopram in this population. The purpose of this study was to examine the benefits and safety of citalopram in a group of children and adolescents with PDD’s. Methods: Twenty-nine patients with PDD’s (25 with autistic disorder, four with Asperger’s disorder) (mean age: 9.4±2.9 years; range: 4 to 15 years) were treated with citalopram for at least two months (mean duration of treatment: 7.4±5.3 months; range: 1 to 15 months). Treatment was initiated at a low dose (5mg daily) and was increased by 5mg weekly as tolerated and as necessary. The mean final dose was 19.7±7.8mg (range: 5mg to 40mg). Target behaviours included aggression, anxiety, stereotypies, and preoccupations. Outcome was based upon a consensus between clinician and parents, using the Improvement Item of the Clinical Global Impressions Scale (CGI) as a guide. Results: Nineteen (66%) children were judged to be much improved or very much improved regarding target behaviours. The symptoms which showed the greatest amelioration with citalopram were anxiety and aggression. Core symptoms of PDD’s (socialinteractions, communication) did not show clinically significant improvement. Citalopram was generally well tolerated, although five patients developed treatment limiting adverse effects: three with increased agitation, one with insomnia, and one with possible tics. Conclusions: The results of this case series suggest that citalopram has beneficial effects on some interfering behaviours associated with PDD’s with few adverse effects. Controlled trials are warranted.

P1.3.10
INSIGHT INTO THE NEURONAL MECHANISM OF SECRETIN THERAPY IN AUTISTIC CHILDREN S Nozaki1*, H Mizuma2 , N Nishimura2, R Kohashi2, Y Watanabe1 1 Dept Physiol, Osaka City Univ Grad Sch Med, 1-4-3 Ashahi-machi, Abeno-ku, Osaka 545-8585; 2 Dept Clin Pathol, Kyorin Univ Sch Health Sci, Tokyo 192-8508, JAPAN

[Background.] In 1998, in the course of the treatment for digestive dysfunction of the patients with autism, the dramatic effect of secretin was reported in the social behavior of autistic children (Horvath, K. et al., 1998.). However, the therapeutic mechanism of secretin treatment has not been clarified yet. Furthermore, the existence of the receptor for secretin in the brain has been suggested, the localization of secretin receptor and the neuronal function of secretin have not been elucidated yet. We have already published the localization of secretin receptor by in vitro autoradiography. We here report more detailed informations of the localization of secretin receptor and interaction between secretin and monoaminergic system. [Results.] In vitro autoradiography with 125I-secretin showed intense binding in the nucleus of solitary tract, laterodorsal thalamic nucleus, and caudate/putamen (Nozaki, S. et al., 2002.). The depletion of dopaminergic neurons during neonatal period resulted in the increase in secretin binding in the caudate/putamen (p P1.3.11
PILOT STUDY OF A DYNAMIC LEARNING ASSESSMENT FOR CHILDREN WITH AUTISM WHO ARE INVOLVED IN STRUCTURED BEHAVIORAL TEACHING SESSIONS. J. Reitzel and J. Summers*. Chedoke Child and Family Centre, Hamilton Health Sciences, Hamilton, Ontario L8N 3Z5.

Recognizing that children with autism are heterogeneous in regard to their pretreatment characteristics and response to behavioral treatment, Schreibman (2000) has called upon researchers to develop strategies for tailoring specific treatment components to their individualized learning needs. One approach that shows promise but that has not been extensively researched is to evaluate children’s rate of progress during the initial phases of behavioral treatment. There is some evidence that a dynamic learning assessment may be a better predictor of eventual outcome than standardized assessments (Smith et al 2000). In order to study this issue, we selected children with autism who had very similar pretreatment characteristics. Our aim was to determine whether differences in their rate and pattern of learning would emerge after even a brief period of behavioral intervention. Four non-verbal male children with an independent diagnosis of autism participated in the study. Their pretreatment characteristics were as follows: mean CA of 74 months, mean CARS score of 44.3; mean MA on the Bayley Scales of Infant Development of 13.2 months; mean composite A-E score on the Vineland Adaptive Behavior Scales of 15.8 months. All 4 children participated in several structured teaching sessions per week that were based on the principles of applied behavior analysis. Children’s acquisition of skills according to a redetermined criterion for mastery was assessed after approximately 100 hours of therapy. Despite the similarity of children’s pretreatment characteristics, differences emerged in their rate and pattern of skill acquisition. While no child made any progress in imitating sounds and words, 3 of 4 children were able to acquire motor imitation skills; and 2 of 4 responded successfully on visual matching tasks. Preliminary data suggest that a dynamic learning assessment may help to develop and refine the most effective individualized teaching strategies for children with autism.

P1.3.12
TIME DELAY AND EXTRASTIMULUS PROMPTS TO TEACH BASIC COMPUTER SKILLS TO CHILDREN WITH AUTISM. D.E. Ross and C.A. Rody. Florida Atlantic University, Department of Exceptional Student Education, Boca Raton, FL 33431.

Increasingly, children with autism need basic computer skills such as operating a mouse or utilizing a software program to participate in many leisure and instructional activities. However, acquiring such skills can be challenging for some children with cognitive, motor, or language delays. Consequently, a lack of basic computer skills can inhibit the beneficial application of computer technology in various settings. The objective of the present study was to evaluate a treatment package to teach basic computer skills to two children with moderate and severe autism who were experiencing significant cognitive, motor, and language delays. In a multiple baseline across subjects design, we tested the acquisition of basic computer skills for one nonverbal preschool male with Fragile X syndrome and one nonverbal school-age female with severe autism. Operating a mouse, tracking a cursor, and sequencing computer screens in a child’s software program were the target behaviors. Each behavior was trained separately using a treatment package consisting of extrastimulus prompts, system of most-to-least prompts, and progressive time delay. Generalization probes were conducted on an untrained software program after training. The number of trials to criterion and the percentage of correct responses indicate the effectiveness of the treatment package. Limitations of the intervention package are also presented with suggestions for future research.

P1.3.13
ENGAGEMENT OF CHILDREN WITH AUTISM AND CHILDREN WITH DOWN SYNDROME IN PUBLIC SCHOOL PROGRAMS L. RUBLE, University of Louisville

Engagement is a key ingredient in effective programs for children with autism (Dunlap, 1999; National Academy of Sciences, 2001), and 25 to 30 hours a week of engaged time is recommended. The importance placed on engagement begs the question “What does it mean for a child to be engaged?” Researchers recognize that engagement is a multidimensional construct comprised of both quantitative and qualitativecomponents. For children with autism, it may be more important to evaluate not only the quantifiable aspects, such as how much time did a child spend engaged, but also the qualitative and interactive features of engagement. For example, assessment of teacher and child behaviors may indicate that although a child is engaged during instructional time, the child requires physical prompting to complete tasks. Another child may writing his ABCs repeatedly during class time, while the teacher is providing an overview on plants to the rest of the class. Both of the children are engaged; but on closer examination, the quality of their engagement is discrepant from peers’ engagement. To examine these aspects of engagement, an ecological psychology methodological approach was used (Barker & Wright, 1950; Ruble, 2001;Ruble and Scott, in press). Two-hour observations of 4 children with autism and 4 children with Down syndrome in public school settings were collected and transcribed into written transcripts called chronologs. Chronologs are currently being analyzed in order to answer the following questions: 1. How much time do children spend in activities that are congruent with classroom expectations? 2. How much time do children spend in activities that appear compliant to teacher expectation, but inconsistent with setting expectations? 3.What instructional strategies and types of activities elicit high congruency with setting expectations and / or teacher compliance?

P1.3.14
THE SCOTTISH CENTRE FOR AUTISM: AN EFFECTIVE DEVELOPMENTAL PRESCHOOL INTERVENTION. J. Salt. Scottish Centre for Autism, Glasgow, Scotland.Currently University of Chicago,Illinois.

Objectives: All previously published outcome studies of pre-school treatment programmes are American; and are predominantly behaviourial inorientation. This paper evaluates the effectiveness of a Scottish, developmentally based, early intervention programme for children with Autism. The treatment protocol includes 1:1 intervention with a therapist and parent-child sessions. Parents are trained in the treatment approaches through workshops, and in vivo with therapists. The aims of treatment are to improve the child’s early social communication and social interaction skills, leading to the potential development of play, flexibility of behaviour, and pre-school social skills. Design: A Group (treatment, waiting list) X Time (pre-post treatment) design was utilised. Methods: Children were allocated to the treatment group (n=14) and control group (n=6) in the usual clinical manner. Standardised assessments were administered by an independent clinician. Pre-treatment comparisons revealed that the control group had a significantly higher pre-treatment IQ; but the two groups were comparable for age, mental age, socio-economic status and number of hours non-experimental therapy. Results: Scores were entered into a repeated measures multivariate analysis of variance. Results demonstrated that children in the treatment group improved significantly more than the control group on measures of joint attention, social interaction, imitation, daily living skills, motor skills and an adaptive behaviour composite. A measure of requesting behaviour fell shy of statistical significance. The total stress index reduced for treatment group parents and increased for the control group parents (but not significantly). Conclusions: The results of the study are considered to support the efficacy of this treatment approach.

P1.3.15
PARENT REPORT OF EFFECTIVE SLEEP INTERVENTIONS FOR CHILDREN WITH AUTISM. P.G. Williams, J. Hersh, L.L. Sears. Department of Pediatrics, University of Louisville, Louisville, KY 40202.

Autism is a developmental disability characterized by pervasive deficits in communication, social interaction, and range of interests and activities. Sleep problems appear to constitute one of the stressful aspects of autism for many families. Prevalence estimates of sleep problems, based on parent report, range from 44 to 83 percent. Among the more common sleep disorders reported in children with autism are those associated with the sleep/wake cycle including long sleep latencies, night waking, early morning waking, shortened night sleep and daytime napping. Little is known regarding the effectiveness of interventions for these types of sleep problems in autism. In order to determine interventions that parents use, we surveyed 700 families with children having an autism spectrum disorder. Based on the 210 respondents (30 % return rate), parents utilized a variety of behavioral and medical interventions to enhance sleep. Over 90 % of the respondents had attempted some intervention to enhance sleep. Establishing a regular bedtime routine, endorsed as helpful by 78 % of parents attempting this strategy, was reported to be the most effective sleep intervention strategy. Medications for sleep were used less frequently than behavioral interventions. Of the medications used for sleep problems, parents reported the most frequent perceived benefit with Benadryl (49 % of parents who tried Benadryl for their child’s sleep reported a benefit). Beneficial sleep interventions were also related to child characteristics since some interventions were less effective for children having significant cognitive delay. Results of this sleep survey provide information that may help parents and professionals develop and monitor sleep interventions likely to enhance behavioral functioning of the child with autism and reduce family stress.

P1.3.16
SPEECH AND LANGUAGE OUTCOMES IN CHILDREN WITH AUTISM AFTER PARENT TRAINING Seung, H.K., Elder, J. H., & Ashwell, S. University of Florida, Gainesville, FL 32608

This study examined language outcomes of 10 children with autism after father received training on social interaction behaviors. Father training consisted of imitating, waiting expectantly, and following child’s lead. Language transcriptions of mother-child and father-child interactions during baseline, intervention, and maintenance periods, were examined. We examined the following questions: 1) Do children with autism demonstrate gains in verbal communication after their parents receive interactive play training? 2) Are there differences in language production depending on a play partner (mother vs father)? Measures of Mean Length of Utterance (MLU), total words and different words produced, and number of vocalizations will be analyzed.

P1.3.17
USING COMPUTER-ENHANCED ACTIVITY SCHEDULES TO TEACH A CHILD WITH AUTISM TO PLAY CREATIVELY. M. Dauphin, E.M. Kinney, BEACON Services, Milford, MA 01757, and R. Stromer* Univ. of Mas. Med. Sch.,Shriver Center, Waltham, MA 02452

The present study evaluated the effectiveness of a four-component teaching package (activity schedules, video modeling, matrix training, and computerized instruction) to teach creative play and play related comments to a three-year-old boy with an autism spectrum disorder. After teaching him to follow notebook activity schedules, a multiple baseline probe design across matrices was used to assess the effects of video modeling on play and commenting during play routines depicted by video in computer-mediated activity schedules. Generalization from trained to the untrained Say and Do components in the play routines, within matrices was analyzed. In addition, generalization of the same components to a notebook activity schedule was assessed. The boy acquired all nine play routines and comments taught via computer presented video models, in the context of schedule following. As well, the boy was on-schedule and highly accurate on the Say and Do components for the 18 routines not directly trained. Results replicate prior matrix-training studies showing recombinative generalization, and recommend video models embedded in computer activity schedules to teach children with autism to independently creatively play and comment during play.

POSTER SESSION 1 TOPIC 4: BIOCHEMISTRY

P1.4.1
REDUCED SERUM LEVELS OF CERULOPLASMIN AND TRANSFERRIN – ANTIOXIDANT PROTEINS IN AUTISM. A. Chauhan, V.P.S. Chauhan and I. Cohen. N.Y.S. Institute for Basic Research in Developmental Disabilities, 1050 Forest Hill Road, Staten Island, New York 10314, U.S.A.

We have analyzed lipid peroxidation status in the plasma and erythrocytes of children with autism, and their non-autistic siblings. Lipid-peroxidation was found to be increased in autism indicating an oxidative stress in this disease. Ceruloplasmin, the copper-transport protein, and transferrin, the iron-transport protein, are the major plasma anti-oxidant proteins that are synthesized in several tissues, including brain. These proteins serve in antioxidant defense by limiting concentration of free ferrous ion that is known to catalyze the conversion of hydrogen peroxide to highly toxic hydroxyl radicals by Fenton reaction. In this study, we quantified concentrations of ceruloplasmin and transferring in the serum samples of children with autism and their non-autistic siblings (n=19) using the MininephTM nephalometer system. The autistic children were diagnosed based on ADI-R and ADOS-G criteria. Both transferrin and ceruloplasmin levels were reduced significantly in the serum of children with autism as compared to non-autistic siblings. These results suggest abnormal iron and copper metabolism in autism that may lead to oxidative stress in autism.

P1.4.2
REDUCED NAA IN CORPUS CALLOSUM IN AUTISTIC DISORDER T.J. DeVito, D.J. Drost, P.C. Williamson, N. Rajakumar, D. Ellison, R. Nicolson*. University of Western Ontario, London, ON, Canada, N6A 4V2

Objective: Magnetic resonance imaging (MRI) studies have reported reduced volume of the posterior regions of the corpus callosum (CC) in patients with autistic disorder (AD). However, the neuronal and metabolic abnormalities underlying this volume reduction have not been examined using magnetic resonance spectroscopy (MRS). The purpose of this pilot study was to assess potential neuronal abnormalities associated with the reduction of the CC in children with AD. Methods: Six boys with AD (ages 6-11) and seven healthy controls (ages7-16) completed an MRS scan using a 3.0-tesla MRI system. All patients had non-verbal IQ > 70, and were diagnosed using the Autism Diagnostic Interview and Autism Diagnostic Observation Schedule. A multislice proton MR Spectroscopic Imaging sequence was developed to collect spectra from two 10-mm thick oblique-axial slices (TI/TE/TR = 235/135/1800 ms, voxel size ~1 ml). Spectra from the posterior region of the CC were analysed, and the groups were compared with respect to their levels of N-acetylaspartate (NAA), a neuronal marker. Results: Subjects with AD showed significantly reduced levels of NAA relative to controls in the posterior region of the CC (p=0.03). Conclusions: Preliminary results from this pilot study indicate that patients with autism have reductions of NAA in the posterior CC. This may indicate that the volumetric reduction previously seen in the CC may be related to a reduction in the number or density of neurons in this region.

P1.4.3
RELATIONS BETWEEN PRENATAL TESTOSTERONE AND CHILDHOOD SOCIAL COGNITION: IMPLICATIONS FOR UNDERSTANDING THE SEX RATIO IN AUTISM. R. C. Knickmeyer, S. Baron-Cohen and P. Raggatt. Autism Research Centre,Departments of Experimental Psychology and Psychiatry, University of Cambridge, Cambridge CB2 3EB and Department of Clinical Biochemistry, Addenbrooke’s Hospital, Cambridge CB2 2QQ.

Prenatal testosterone plays a central role in organizing the brain for later social behaviour in animals, and has been implicated in a range of cognitive skills in human beings. In this study 58 children (male and female), whose prenatal testosterone level was previously analysed in amniotic fluid, were followed up at age 4. Their parents completed the Children’s Communication Checklist, a questionnaire assessing pragmatic language ability. Results showed an inverse relationship between foetal testosterone and (1) the ability to use conversational context appropriately (2) conversational rapport and (3) quality of social relationships. An opposite relationship was seen for restricted interests. In addition, the children were shown cartoons of moving shapes and asked to describe what was happening. Results show a sex difference in the tendency to interpret the shapes socially, with females using more mental state terms than males. These findings are compatible with the Extreme Male Brain theory of autism and reaffirm the role of testosterone in structuring the social brain. They may also have important implications for understanding the sex ratio in autism.

P1.4.4
LONGITUDINAL BIOPROFILING OF AN AUTISTIC POPULATION. A.Marolewski*, R. Strong, G. Rivers, A. Springer, E. Smith, J. Rusche. Repligen Corporation, Waltham, MA 02453.

Previous bioprofiling studies have tended to focus on small populations and limited study duration. We have monitored 89 autistic children (age 3-6) for over 1 year, focusing on markers of immune activation, metabolism, and inflammation. Immune activation was assessed via cytokine profiling (IL-2, -4, -5, -10, TNFa, and IFNg). Elevated levels of multiple cytokines were detected in 16% of subjects while significant and consistent IFNg elevations were seen in 5%. Chymotrypsin was studied as a marker of pancreatic function. Stool chymotrypsin levels were low in approximately 30% of samples at any timepoint. Analysis of 2 other populations showed low chymotrypsin correlated with autism and not GI symptoms. Exocrine pancreatic insufficiency was considered but not supported by additional experiments. Two urinary metabolic makers studied were uric acid and 7-methyl xanthine (7-MX). This autistic population showed increased frequency of hyperuricosuria when compared to neurotypical control group. Hyperuricosuria remained consistent at both the population and patient level. 7-MX was decreased versus control group over the entire study. The source of urinary 7-MX is unknown although it did not correlate with xanthine or urate. Elevations of stool calprotectin (neutrophile protein marker of inflammation) suggested gut inflammation, however, individual patient values were inconsistent and frequencies were not statistically different from an age-matched neurotypical control group. Systemic measures of inflammation such as c-reactive protein (CRP) also were not elevated. This long-term large-scale evaluation suggests a heterogenous autistic population comprising several biological phenotypes which warrants further study.

P1.4.5
ALTERED SEROTONINERGIC NEUROTRANSMISSION AND AUTISM-LIKE DEFICITS: INSIGHTS FROM THE BDV MODEL. M.Pletnikov*1,2; D.Dietz1; S.Rubin2; T.H.Moran1; K.Carbone1,2. 1Dept. of Psychiatry, Johns Hopkins Univ. Sch. of Med., Baltimore, MD 21205; 2CBER/FDA, Bethesda, MD 20892.

Neonatal Borna disease virus (BDV) infection produces selective autism-like neurodevelopmental damage and behavioral deficits in rats. The BDV rat model was used to explore the mechanisms of serotonin (5-HT) alterations that may mediate autism-like behavioral deficits and might be responsible for effects of 5-HT compounds in autistic patients. Neonatal BDV infection produced increased tissue levels of 5-HT in brain regions damaged by the virus and at the level of cell bodies in developing rats. Simultaneous BDV-associated up-regulation of postsynaptic 5-HT receptors was also noted. Compared to control animals, young and adult BDV-infected rats also demonstrated greater responsiveness to compounds that stimulate 5-HT neurotransmission. The character of the neurochemical outcomes and pharmacological effects appear to indicate that in spite of elevated tissue content, 5-HT neurotransmission may be decreased in BDV-infected rats. The present findings provide a plausible model of autism-like 5-HT disturbances (e.g., hyperserotoninemia) and explain effects of 5-HT compounds observed in some autistic patients.

P1.4.6
PROTEOLYSIS OF NEUROTROPHIN 4, A BIOMARKER OF AUTISM, BY CATHEPSIN D. A. M. Sheikh, V. P. S. Chauhan, A. Chauhan, W. D. Spivack and M. N. Malik. N.Y.S. Institute for Basic Research in Developmental Disabilities, 1050 Forest Hill Road, Staten Island, New York 10314, USA

Recently, it has been reported that Neurotrophin 4 (NT4) levels are increased in the serum of children with autism [Nelson et al., Ann Neurol. (2001) 49, 597-606]. The mechanism of increase in serum NT 4 in autism is not known. We report here that NT4 is proteolytically cleaved by cathepsin D. Cathepsin D cleaved NT 4 in a time- dependent manner that resulted in the appearance of two peptides of approximately 9 kDa and 11 kDa molecular weight. The digestion of NT4 by capthepsin D was also dependent on the concentration of the enzyme. The proteolysis by cathepsin D as a function of pH showed that acidic pH of 4 resulted in the maximum proteolysis of NT 4. The peptide was not digested at pHs 5.5, 6.2 and 7.5. These results indicate that NT4 may be processed in the lysosomes. It is suggested that levels of cathepsin D may be decreased in the sera of children with autism thereby increasing the levels of serum neurotrophin 4 in autism. Further studies are in progress to compare the levels of cathepsin D in the sera of children with autism and their non-autistic siblings.

POSTER SESSION 1 TOPIC 6: GASTROINTESTINAL & TOXICOLOGY

P1.6.1
ABNORMAL HEAVY METAL EXCRETION IN CHILDREN WITH AUTISM J.B. Adams, C. Holloway, F. George, B. Done, Arizona State University, Tempe, AZ 85286-6006

Our previous pilot study found that most children with autism were exposed to above average levels of mercury in utero, and that their ability to excrete mercury was diminished in most cases by the excessive use of oral antibiotics. Also, recent analysis of baby hair by A. Holmes has found that children with autism have unusally low levels of mercury in their hair, suggestive of a diminished excretion, and the level of mercury had an inverse correlation with the severity of autism. Finally, there have been many anecdotal reports of the benefit of chelation treatments on children with autism. Therefore, we carried out a DMSA challenge study involving 15 children with autism and 15 typical children. The children received a single dose of meso-2,3-dimercaptosuccinic acid (DMSA), at a dose of 10 mg/kg, followed by a 10-hour urine collection. The DMSA resulted in a much greater increase in heavy metal excretion in the children with autism compared to the controls. Many of the children with autism excreted high levels of one or more heavy metals, although there was wide varation in the amount and type of metal excreted. Therefore, our data suggests that many children with autism have a greatly diminished ability to excrete heavy metals, and thus would be unusally vulnerable to exposures to those metals.

P1.6.2
AN INVESTIGATION INTO THE PATTERN OF PRESENTATION OF AUTISTIC CHILDREN RECEIVING MORE THAN ONE COMBINED MEASLES CONTAINING VACCINE, K. Aitken

A significant question which has been raised in both the UK and North America over the possible association between MMR vaccine administration and the development of ASDs concerns the issue of typical time of first parental worry over development. As this, on most accounts, coincides with the period of MMR administration, it could easily lead to inappropriate parental attribution of the basis to the child’s difficulties to vaccine administration through temporal coincidence. In an attempt to critically address this issue, a cohort of autistic children whose history exhibited both parental and clinical concern only after initial MMR exposure, and who recieved a second, ‘booster’ MMR vaccination were ubjected to a systematic casenote review, using a standardised clinical protocol to identify any evidence of a second regression after ‘booster’ vaccination. The presence of any such second effect would significantly increase the plausability of a vaccine-triggered phenomenon. A group of diagnosed children who had had a single vaccine exposure were also reviewed to ensure that any apparent second exposure effect not also seen in the typical development of autistic children. At the time of abstract submission, the data analysis for this project has not been finalised.

P1.6.3
UPPER AND LOWER GASTROINTESTINAL PATHOLOGY IN 182 AUTISTIC CHILDREN A. Anthony, S. H. Murch, A. J. Wakefield. Depts of Histopathology & Paediatric Gastroenterology, Royal Free Hosp, London NW3 2PF, UK.

An inflammatory bowel disorder characterised by lymphoid nodular hyperplasia (LNH) and mild acute and chronic inflammatory changes of the colorectum have been reported previously by our group in 60 autistic children with GI symptoms. An underlying autoimmune GI lesion has also been proposed for autism. In this report we characterise the gastrointestinal pathology in a larger group of 182 children that comprised 142 children with either autism or autistic spectrum disorder, 13 with Asperger’s syndrome and a further 27 children with other behavioural problems including some autistic traits. All 182 children underwent ileo-colonoscopy. 78/182 (43%) children were on non-restricted diets. Of 75 of the 182 cases who also underwent upper GI endoscopy, 53 (71%) showed some degree of either erythema or inflammation in the oesophagus and stomach. Ileal LNH grades 0 (normal), 1, 2, 3 were seen in 14, 21, 30, and 35% of the children, respectively. Colonoscopy showed no abnormalities, erythematous/inflamed mucosa, prominent lymphoid follicles or mucosal granularity/vascular changes in 13, 24, 53 and 36% of cases, respectively. Either acute or chronic histological inflammatory changes were seen in the esophagus, stomach and duodenum of 15, 24 and 25% of cases. In the colorectum, acute inflammation, chronic inflammation and raised mucosal eosinophils were seen in 47, 71 and 41% of cases, respectively. In 20% of cases the acute inflammatory changes were mildly active in nature. Histologic melanosis coli was also seen in 12% of cases. The pathology of this appears consistent. However, links with cognitive function remain to be explored.

P1.6.4
SPONTANEOUS PRO-INFLAMMATORY CYTOKINE PRODUCTION BY DUODENAL AND COLONIC LYMPHOCYTES IN CHILDREN WITH AUTISTIC SPECTRUM DISORDER. P. Ashwood, S.H. Murch, A.J. Wakefield. Paed. Gastro, Royal Free Hospital, London, UK

There have been previous reports of a novel lymphocytic entero-colitis in children with autistic spectrum disorder (ASD). Systemic immune responses consistent with a dysregulated innate immune response, with raised PBMC-TNFalpha production and TH2 skewing have been reported in similarly affected children. The purpose of this study was to characterize spontaneous pro-inflammatory intracellular cytokine production in the duodenal and colonic lymphocytes of ASD children. Duodenal and colonic biopsies from 14 ASD children, 24 histologically normal controls and 33 histologically inflamed controls were compared. Single cell suspensions were isolated from the epithelial compartment following washes in EDTA and lamina propria, by incubation in collagenase. Detection of CD3+ lymphocytes spontaneously producing intracellular TNFalpha, IL-2, IL-4, IFNgamma, IL-10 was performed by multicolour flow cytometry. Controls included isotype-matched antibodies and stimulated PBMNC. In ASD children, CD3+ lymphocytes had increased intracellular IL-2 and IFN in both lamina propria and epithelial compartments compared with controls. Furthermore, TNF was substantially increased in ASD children compared with controls (p P1.6.5
PRENATAL MODULATION OF THE BETA-2 ADRENERGIC RECEPTOR IN AUTISM: ANALYSIS OF DIZYGOTIC TWINS. S. L. Connors; D. E. Crowell; S. J. Spence; C. J. Newschaffer; A. W. Zimmerman. Kennedy Krieger Institute and Johns Hopkins University, Baltimore, MD 21205; UCLA, Los Angeles, CA 90095.

Epigenetic prenatal factors may be important for the development of autism. These might include stressors during pregnancy and twinning. Multiple pregnancy increases the chance for exposure to tocolytic agents such as terbutaline, a specific agonist for the beta-2 adrenergic receptor. Experimental maternal treatment with terbutaline during the second trimester in rats interferes with development of several neurotransmitter systems, synaptogenesis in the CNS, and peripheral receptor function (Slotkin TA, 1989). We analyzed dizygotic (DZ) twin pairs in the AGRE database and in 3 autism clinics for whom prenatal histories were available. Of 18 sets of DZ twins (10 concordant, 8 discordant for autism spectrum disorders), 5 of the concordant pairs had continuous prenatal exposure to terbutaline for 2 weeks or more during the 2nd or 3rd trimester. Observed concordance for the 18 DZ twin pairs (56%) was greater than predicted from twin studies (10%). In an additional concordant case the mother underwent surgery, a major stress, at 3 months gestation. Of the 8 discordant twin pairs, 7 had no substantive terbutaline exposure (p = 0.067; 2 tailed Fischer’s Exact test). Overstimulation of the beta-2 adrenergic receptor by terbutaline orstress, or interference by prenatal viral infections or environmental factors such as pesticides, may alter normal brain development by delaying maturation of the sympathetic nervous system and contribute to autism. We hypothesize that alterations in beta-2 adrenergic receptor signaling in the fetus may be an important factor in the development of autism. An association between terbutaline and DZ concordance, if real and not attributable to chance variation or ascertainment bias, would support this.

P1.6.6
MERCURY EXPOSURE AND AUTISM: A CASE-CONTROL STUDY C. Holloway, J.B. Adams, M. Margolis, X. Liu, Arizona State University, Tempe, AZ 85287-6006

The cause of autism is unknown, but there are suggestions that mercury and other heavy metals play a role in its pathogenesis. A case-control study of 50 children with autism and 50 age- and gender-matched controls was carried out to investigate the possible link between mercury and autism. Some of the highlights of the study include: 1) Maternal consumption of seafood during pregnancy was an important risk factor for autism, presumably due to mercury in the fish. This was consistent with higher levels of mercury in the hair of mothers of children with autism compared to the control mothers. 2) Children with autism had 5x as many ear infections as the controls during their first three years of life, so they also had 5x the exposure to oral antibiotics. Oral antibiotics have been shown in rats to increase the half-life for excretion of mercury from 10 days to over 100 days. 3) A detailed hair analysis of toxic metals will be discussed. Overall, the data suggests that children with autism were exposed to higher levels of mercury, and that they had a decreased ability to excrete mercury due to oral antibiotic use. Since mercury is a known neurodevelopmental toxicant, leading to symptoms similar to autism, our data suggests that mercury poisoning could be an important factor in exacerbating and/or causing many of the symptoms of autism. However, larger, more-controlled studies are needed to test this hypothesis.

P1.6.7
LIVER IMPAIRMENT IN CHILDREN WITH REGRESSIVE AUTISM.(1)Rosseneu SLM, (2) van Saene HKF,(1)Heuschkel R, (1) Murch S. (1) Centre of Paediatric Gastroenterology, Royal Free University Hospital, London,United Kingdom,(2)Department of Medical Microbiology, University of Liverpool, Liverpool, United Kingdom.

Background: Children with regressive autism have non-specific enterocolitis and complex immunopathology of the gut mucosa. Recently our research group reported that >50% of the children diagnosed with regressive autism and gastrointestinal symptoms, have overgrowth of abnormal flora including aerobic Gram-negative bacilli (AGNB). This abnormal condition is known to impair liver function. Outcome measure: whether children with regressive autism and gut disease, including bacterial overgrowth, have impaired liver function. Methods and results: liver function and abnormal gut bacteria were evaluated in 20 children diagnosed with regressive autism and gut symptoms and who were included in the ongoing flora study of our research group. The liver function tests included aminotransferases (AST,ALT), bilirubin,gamma-glutamyltransferase. Thirteen patients (65%) carried abnormal bacteria including Enterobacter, Citrobacter in high concentrations of = or > 10.5 AGNB per ml of saliva and/or g of faeces. A total of 10 patients (50%) had liver impairment reflected by an increased AST (median of 49 IU/L, range 41-56, normal: P1.6.8
ORAL GAMMAGLOBULIN TREATMENT OF GASTROINTESTINAL SYMPTOMS IN CHILDREN WITH AUTISM C.K. Schneider, R.D. Melmed, J.A. Ostrem, F.J. Enriquez, L.E. Barstow*, Southwest Autism Research Center, Phoenix, AZ 85006; Protein Therapeutics, Inc., Tucson, AZ 85747

Gastrointestinal pathology is common in children with autism and may be autoimmune in nature. An open label pilot trial was conducted with an oral formulation of human gammaglobulin in twelve children with chronic gastrointestinal (GI) disturbances and a diagnosis of Autistic Disorder. A Gastrointestinal Severity Index (GSI: scale 0 – 15) was the primary efficacy endpoint in the trial. Children ages 2 to 8 with a GSI score of at least 7 and onset of autistic symptoms after 15 months of age were enrolled in the study. Secondary endpoints included measures of behavior (Autism Behavior Checklist: ABC; Physician and Parent Clinical Global Impression), safety, and tolerability as monitored by physical examinations and laboratory tests. All children received a single daily dose of oral gammaglobulin (Oralgam(tm)) before bedtime for eight weeks. Nine subjects completed eight weeks of treatment. Seven of nine subjects (78%) met the definition of response and five of nine (55%) met the definition of clinical remission in GI symptoms. There were modest improvements in behavioral scores (ABC median score of 94, 90, and 75 at baseline, 4 weeks, and 8 weeks, respectively). Conclusions: A majority of children treated with oral gammaglobulin in this small open-label pilot study showed improvement in gastrointestinal disturbances. Larger randomized placebo-controlled trials are necessary to evaluate the therapeutic efficacy of oral gammaglobulin in the treatment of chronic gastrointestinal disturbances associated with Autistic Disorder and related developmental disorders.

POSTER SESSION 1 TOPIC 7: COGNITION Mottron & Mitchell Symposium

P1.7.1
ATYPICAL TEMPORO-OCCIPITAL FAST-FREQUENCY EEG ACTIVITY DURING REM SLEEP IN AUTISTIC SPECTRUM DISORDERS. R. Godbout,* É. Limoges, A.-M. Daoust, L. Mottron. Neurodevelopmental Disorders Program, Hôpital Rivière-des-Prairies & Dept. Psychiatry, Université de Montréal, Canada, H1E 1A4.

Physiopathological models of autistic spectrum disorders (ASD) involve atypical perceptivo-visual processing. Since REM sleep is a state of endogenous, spontaneous activation of the visual system, the present study explored whether REM sleep-selective (Beta) EEG activity could discriminate ASD participants from controls. Nine participants with ASD (8 M, 1 F; 22.2 ± 4.1 years old) and eight comparison participants (7M, 1F; 23.5 ± 4.9 years old) were included in this study. All participants were right-handed and had a full-scale IQ > 80. EEG activity during REM sleep was recorded with a 12-electrode montage, including the temporo-occipital (visual) areas. EEG samples were made of 24 four-seconds artefact-free segments taken in equal proportions from the first three REM sleep periods. EEG data was Fast Fourier transformed and spectral amplitude values (mV) were determined for the 13.0-19.75 Hz window. Groups were compared with Mann-Whitney U-tests. REM sleep EEG of ASD patients compared to control participants showed a significantly lower absolute beta spectral amplitude over the primary (O1: U = 15, p < .05; O2: U = 12.5, p < .03) and associative visual areas (T5: U = 11.5, p < .03; T6: U = 11, p < .02). Waking EEG recordings were also performed in the evening preceding, and the morning following sleep recordings: no group differences were found. These observations support the hypothesis according to which neurophysiological mechanisms responsible for visual perception may be atypical in ASD. We propose that REM sleep-active thalamo-cortical networks are more sensitive to such atypicalities than waking mechanisms.

P1.7.2
CHANGE BLINDNESS FOR PEOPLE BUT NOT OBJECTS AMONG CHILDREN WITH AUTISM. S. Joseph, J. A. Burack*, J. T. Enns & D. I. Shore. Dept of Ed Psych, McGill U, 3700 McTavish, Montreal, QC H3A 1Y2.

Changes made to a scene during an interruption, such as an eye movement or brief flicker, often go undetected unless attention is directed toward the location of the change (Rensink, O’Regan, & Clark, 1997; Simons & Levin, 1998). The direction of attention to the change may be exogenous (e.g., a local transient in the image) or endogenous (e.g., voluntary inspection of a specific object). In the present study, change detection was examined with 2 paradigms among children with autism (MA of 4 -12 years) and typically developing children. In Experiment 1, one member of a flickering pair of pictures contained either (a) a color change in part of an object, (b) the removal of a part of an object, or (c) a picture plane rotation of the whole object. The observer’s task was to indicate as quickly as possible which picture had a change. Two rates of flicker were tested (50ms and 250ms) to compare exogenous and endogenous attention to the changed object. Children in both groups detected exogenous changes more rapidly, and were more likely to detect global (object rotation) as compared to local changes (part removal). The task in Experiment 2 was the same as in Exp 1 except that people were depicted interacting with objects. The people and objects could undergo either color changes or part removals. The children with autism showed more change blindness (i.e., they were slower at detecting change) than typical children, but only when the changes were made to the people. These results suggest that the change blindness task is sensitive to both the pervasive social deficits of children with autism as well as to their preoccupation with objects.

P1.7.3
SMOOTH PURSUIT AND SACCADIC EYE MOVEMENTS IN PERVASIVE DEVELOPMENTAL DISORDER C. Kemner, J.N. van der Geest, M.N. Verbaten and H. van Engeland. Department of Child and Adolescent Psychiatry, University Medical Center Utrecht, The Netherlands.

Several brain regions have been implicated in the etiology of Pervasive Developmental Disorders (PDD), such as the frontal lobes, the cerebellum, and the brainstem, but there is no general agreement on their functional involvement in PDD. Since these structures are all involved in the generation and control of smooth pursuit and saccadic eye movements, two oculomotor paradigms were used to study possible brainstem, cerebellar or frontal lobe dysfunctioning in a group of children with PDD. Using electro-oculography (EOG), smooth pursuit positional gains and number of saccadic intrusions, saccade metrics (amplitude and scatter of amplitude) and dynamics (duration and peak velocity), as well as the main sequence relationships between amplitude and duration or peak velocities were measured in a group of children with Autistic Disorder or PDD-NOS (about 10 years of age and average intelligence) and an age- and IQ matched control group. No differences between the groups were found on any of the aforementioned parameters. The results indicate that the saccadic and smooth pursuit system have matured to a normal level of functioning in children with PDD of this age. This suggests that brain abnormalities in PDD are unlikely to involve the oculomotor regions in the brainstem, cerebellum and frontal lobes.

P1.7.4
ATYPICAL VISUAL FIXATIONS IN INDIVIDUALS WITH AUTISM AND LOW LEVEL OF DEVELOPMENT. F. Marcil-Denault, S. L. Mottron*, Mineau, L. M. Sauvé, S. Palardy, M. Lemay. Lab of Information Processing in PDDs, Univ. of Montréal, Qc, Can., H1E1A4.

The Autism Diagnostic Observation Schedule-Generic (ADOS-G ; Lord, Rutter & DiLavore, 1998) emphasises the presence of abnormal sensory interests in various modalities. Among those behaviours, prolonged fixation of moving objects appears especially frequent, although this has never been empirically assessed. The current study aims at characterising atypical behaviours of visual fixation (ABVF) in natural settings. The study establishes an observation scale for the assessment of ABVF in autism, and validates the inter-judge reliability of this instrument. The ABVF scale will the be used for comparing the incidence of atypical visual behaviours in individuals with autism and with typically developing children matched in chronological age (CA) and verbal mental age (VMA). Only the first step of the study has been completed at the time of submission. Video recordings of 32 children with autism (mean CA: 43 months; mean VMA 25 ms) in natural settings were scored for visual behaviours that did not belong to the repertoire of typically developing children, and were evident in at least 20% of the children. These aberrant visual behaviours, were found in the form of 6 ABVF: exploration in the lateral visual field, close exploration, prolonged fixation in a mirror, fixation with one obstructed eye, fixation of a moving object and exploration of horizontal or vertical axes. This study, when completed, will allow us to relate early perceptually oriented behaviours observed in natural settings with recent laboratory findings of selective deficits within movement perception (Milne et al., 2002; Bertone, Mottron, Jelenic & Faubert, in press).

P1.7.5
A PREFERENCE FOR VISUAL FEAUTURES WHEN CATEGORISING IN AUTISM. P. Mitchell.*, & D. Ropar. Sch of Psy., Univ. of Notts. UK.

A categorisation task requiring individuals to sort books into two boxes was presented to children with and without autism. The books could be sorted into two equal sized groups according to colour (orange or green), size (large or small), or meaning (sports or games). Twenty children with autism, 22 with moderate learning difficulties and 23 typically developing 8 year olds took part in the study. Those without autism were significantly more likely to sort according to meaning than those with autism. In second experiment, children with autism and with MLD were asked to sort a different set of books which could only be sorted according to meaning. All visual features of the books were identical except for the pictures and labels which identified each book as a sport of game. Most of the individuals with autism were still unable to sort the books according to meaning. Some were able to categorise only when the experimenter explicitly stated that some were sports and some were games. This suggests that individuals with autism may have difficulty perceiving similarities on a thematic level, unless explicitly prompted, but are able to perceive similarities on a visual level.

P1.7.6
CONFIGURAL LEARNING IN AUTISM. KC Plaisted, LM Saksida. Dep’t of Experimental Psychology, University of Cambridge, Cambridge, CB2 3EB, UK.

The weak central coherence (WCC) hypothesis (Frith 1989) is a prominent theory concerning the abnormal performance of individuals with autism on tasks which involve local and global processing. However, studies have been unable to identify the locus of the mechanisms that may be responsible for weak central coherence effects. In the light of these studies, we propose that perception operates in autism to enhance the representation of individual perceptual features but that this does not impact adversely on representations that involve integration of features. These hypotheses were compared by presenting high-functioning children with autism and typically developing children with configural learning tasks. Models of configural learning state that when the significance of a stimulus is determined only by the combination of two or more features, those features are unified in a single representation. Thus, these models would identify abnormalities in configural representations as the locus of weak central coherence in autism. In contrast, the perceptual hypothesis, which states that features are more salient and acutely represented in autism, predicts no deficit in the unification of these features in a configural representation. The children were also assessed on a simple feature discrimination task. Children with autism showed no impairment on the configural learning tasks, but performed better than typically developing children on the feature discriminations. We argue that these results are not consistent with WCC, but rather that perceptual feature representations are particularly acute in autism.

P1.7.7
AUTISM AND REFLEXIVE ORIENTING TO EYE DIRECTION J. Ristic, L.Mottron*, G. Iarocci, J. Burack, Jelenic, P. and A. Kingstone Laboratoire pour le’étude du traitement de l’information dans les troubles envahissants du développement, Hôpital Rivière des Prairies Montréal, Québec, Canada H1E1A4

Where other people look can reveal where they are attending, and thus indicate sources of information in the environment. Normally, infants by 3 months will look preferentially at the eyes, and by 12 months they will attend to where someone else is looking. Work by ourselves and others has shown that healthy children and adults will attend to where someone else is looking, even if a shift in gaze does not predict target location. These reflexive shifts in attention produce faster responses for targets that appear at the gazed-at location. It has been suggested that this joint attention is crucial to the development of social cognition, and that it may be lacking in individuals with autism. We tested this by presenting adolescents with high-functioning autism of average intellectual functioning, as well as typically developing individuals matched in gender and FSIQ, with a simple schematic face that looked left or right. Then a stimulus demanding a speeded detection response appeared to the left or right of the face. Participants were told that gaze direction did not indicate where the response target would appear. Control observers responded more quickly when a gaze was towards the target. Importantly individuals with autism were unaffected by gaze direction. This study provides evidence that autistic individuals are unique in that they do not orient attention reflexively in response to gaze direction. Implications for theories of social attention are discussed and possible brain mechanisms are considered.

P1.7.8
THE ROLE OF PRIOR KNOWLEDGE IN AUTISTIC PERCEPTION. D. Ropar*, K. Ackroyd, & P. Mitchell. Sch of Psy., Univ. of Notts. UK.

The influence of conceptual knowledge on perception of shape was explored in individuals with and without autism. The Shepard illusion was presented to participants on a laptop computer. This illusion presents two identical rectangular tables, but one is standing up and the other islying on its side. The orientation of the stimuli gives rise to illusory distortion such that the table standing up tends to look long and the table lying on its side tends to look wide. The effect is more potent when the stimuli have legs and so look like solid objects. Individuals were required to use the computer keys to adjust the length and width of one table to make it look the same size as the other. The results showed that individuals without autism adjusted the shape more accurately when the stimuli did not have legs. Individuals with autism however performed equally well on both conditions suggesting their perception of shape is less influenced by conceptual knowledge. These findings support previous research which suggests individuals with autism are less influenced by conceptual knowledge on a shape constancy task (Ropar & Mitchell, 2002).

-TOP OF PAGE-