S2.1.1
PART-WHOLE ANALYSIS & STRATEGY FORMATION IN AUTISM. N.J. Minshew,* E.S. Kuschner and G. Goldstein. Univ. of Pittsburgh School of Medicine, Pittsburgh, PA 15213.
This study examined the relationship between strategy formation (planning) and a local processing approach (fragmentation) to problem solving in individuals with high-functioning autism. Participants were 30 affected adolescents and adults aged 15-55 years + 42 group-matched normal controls and 30 affected children aged 8-14 years + 32 controls, all with average visuospatial skills. The participants completed copy, immediate memory, and delayed memory conditions for the Rey Osterrieth Complex Figure Test, scored with the Boston Qualitative Scoring System. The older autism group had significantly decreased planning and increased fragmentation as compared to the control group. Fewer differences occurred in the performance of the younger autism group relative to the controls. Conclusions: 1) These higher order skills emerge in the second decade, a second wave of abilities resulting from frontal maturation; 2) Individuals with autism use a piecemeal strategy and reduced global strategies demonstrating the reciprocal relationship between local processing and deficient concept formation and showing that these are two facets of the same cognitive phenomenon; and, 3) The absence of maturation of higher order circuitry and cognitive abilities results in residual reliance on lower order abilities and circuitry. Some refer to this cognitive constellation as lack of central coherence. However, we characterize it as a deficit in information processing that disproportionately impacts on complex information processing and on higher order integrative circuitry.
S2.1.2
GLOBAL AND CONFIGURAL PROCESSING IN CHILDREN WITH AUTISM AND ASPERGER SYNDROME. C.Rondan, J.Fagot, B.Wicker, B.Gepner, & C.Deruelle. CNRS-INPC, 31 chemin Joseph Aiguier, 13402 Marseille, France
Autistic people process visual information differently from normal populations. In facial perception tasks they seem to rely on a local strategy, while the normal population tend to use a configural strategy. On the other hand, results from studies using global figures tend to show contradictory results. In these tasks autistic people use both local and global strategies, while the normal population show a global bias. Although the terms “global” and “configural” are often taken as equivalent, Kimchi (1992) proposed that global and configural processes correspond to two distinct modes of object’s encoding.Global processing is defined as the processing of the highest hierarchical level of the stimuli, while configural processing is defined as the processing of the inter-element relations. Two experiments tested if the dissociation between global and configural processing would apply to autistic children. Eleven children with autistic spectrum disorders were individually matched with children of two control groups and tested in two tasks, one involving global versus local processing, the other involving configural versus local processing. Most of the children with autism exhibited a perceptual bias for processing the global stimulus level in the first task, but exhibited a local bias in the second experience. We conclude from these findings that autistic children are particularly deficient in configural processing, but most of them have a preserved ability for global processing.
S2.1.3
Network Model of Context Utilization in Autism Spectrum Disorder David Q. Beversdorf, Ashleigh J. Hillier, Ralph Brooker, John D. Hughes
Parallel distributed processing (PDP) modeling has been suggested to help with the understanding of autism spectrum disorder (ASD). We report an application of this process to context utilization in ASD. Individuals with ASD demonstrate impaired utilization of context in understanding their environment, referred to by some researchers as “weak central coherence”. This impaired utilization of context allows for superior performance among high-functioning individuals with ASD on recognition memory testing on the ‘false memory test’. We propose a simplified PDP-based model of how this might occur, which may have an anatomical basis. In individuals without ASD, semantic priming experiments support a model wherein presentation of one word, such as ‘apple’, strongly activates the neighboring nodes of closely related words such as ‘fruit’, ‘tree’, ‘red’, and ‘juice’, with the greatest activation at the node for ‘apple’. Other less related nodes are more weakly activated, such as ‘roots’, and ‘pear’. We propose that in ASD, the ‘apple’ node remains most strongly activated, but the neighboring nodes are relatively less activated. We demonstrate this model to be consistent with the superior performance on recognition testing on the ‘false memory test’, but not on free recall. This may have an anatomic basis in the diminished degree of hippocampal neuronal arborization reported in some studies, or in the abnormal minicolumnar pathology in ASD.
S2.1.4
VISUAL DISORGANIZATION IN HIGH-FUNCTIONING AUTISM AND ASPERGER’S DISORDER. L. Kenworthy*, T. Ahluvalia, G. Wallace, A.Wagner, L. Gilotty, L. Sirian, K. Towbin. Children’s National Medical Center, Washington, DC 20016.
Executive dysfunction, specifically deficits in flexibility and planning, is well documented in high functioning children with autism spectrum disorders. Theories of autism emphasizing “weak central coherence,” a piecemeal processing style, or deficits in “complex information processing” raise questions about the role of organizational abilities in the disorder. We hypothesized organization deficits in both HFA and AD. Two experienced clinicians reviewed charts of 33 HFA and 29 AD patients who had been consecutively evaluated in a neuropsychology service to confirm diagnoses (DSM-IV criteria used). Groups matched on age (range =6-17 years; HFA M=10.9; AD M=10.0), PIQ (HFA M=95.2; AD M=99.8) and gender (30 HFA males, 26 AD males). All patients had FSIQ> 70. Patients completed the Rey-Osterrieth Complex Figure and the Beery VMI. Rey figures were assigned Bernstein-Waber organization scores. There were no significant differences between the HFA and AD groups on either of these measures. Parent report obtained via the Behavior Rating Inventory of Executive Function showed no difference between groups on the Plan/Organize scale, though scores for both groups were impaired relative to the normative sample (p S2.1.5
COGNITIVE PROFILES IN AUTISM AND ADHD: DISTINGUISHING WEAK COHERENCE AND EXECUTIVE DYSFUNCTION. R. Booth, R. Charlton, C. Hughes, and F. HappÈ. Social, Genetic and Developmental Psychiatry Research Centre, Institute of Psychiatry, Kings College London, London, SE5 8AF.
Although it is clear that autism and ADHD are distinct disorders, interesting areas of overlap have been shown. Most noticeably, specific impairments in executive functioning (EF) are thought to characterise each group but may differ in type, degree and severity. Weak central coherence (CC) has been suggested to explain the peculiarities in perceptual processing observed in autism, although it is not established if and how this processing style interacts with EF. This study was undertaken to clarify the relationship between EF and CC by contrasting the cognitive profiles of 30 boys with ADHD, 32 boys with high-functioning autism and 32 typically developing control boys. Groups were matched for chronological age (mean = 11 years, range 8 to 16 years) and full scale IQ (mean = 102, range 70 to 140). A battery of tasks were administered to assess various components of EF (set shifting, inhibition, planning, working memory, self-monitoring and generativity) and to tap CC processing style (e.g., embedded figures, block design, homograph reading, un/connected recall, and sentence completion). Novel tasks were also administered that were designed to pit demands for EF against demands for CC. The two clinical groups were found to show distinctive cognitive profiles. Furthermore, executive dysfunction and weak coherence appear to be dissociable; within each group, executive performance did not predict performance on CC tests. The implications of these findings for theories of perception in autism will be discussed.
S2.1.6
THE DEVELOPMENT OF THEORY-OF-MIND IN CHILDREN WITH LESSER VARIANTS OF AUTISM: A LONGITUDINAL STUDY. E.M.A. Blijd-Hoogewys, P.L.C. van Geert*, M. Serra and R.B. Minderaa. Department of Clinical & Developmental Psychology, Univ. of Groningen, the Netherlands, 9712 TS.
Children with a pervasive developmental disorder are known to have Theory-of-Mind problems. Theory-of-Mind (ToM) is the social cognitive ability to attribute mental states -such as thoughts, intentions and emotions- to oneself and others and to use these in predicting and explaining behaviour of others. Children with lesser variants of autism (PDD-NOS) also often have ToM problems. Thirtyfive children with PDD-NOS were tested in a period that is supposed to be critical for the development of Theory of Mind skills, i.e. from three to seven years old. They were followed over a period of two years; being tested every four months. A new test, The Theory of Mind Story Books, based on the work of Wellman and colleagues, was used. It consists of four alternative versions of a set of storybooks, illustrated with full-colour pictures. The test includes a broad range of tasks tapping recognition of emotions, the ability to distinguish between mental and physical entities and the ability to predict both actions and emotions on the basis of beliefs and desires. We will discuss the psychometric qualities of the test, the pattern of performances on the various aspects of Theory of Mind, and the developmental profile of children with PDD-NOS. Our research showed that children with PDD-NOS have a lower ToM quotient and a greater variability between measurements than normally developing children (N=220). The development of ToM skills in children with PDD-NOS is not only delayed but the growth pattern is also qualitatively different from normally developing children. S2.2.1IMPAIRMENTS IN BOTH CORTICAL AND SUBCORTICAL MOTOR CONTROL IN AUTISM Jansiewicz, E.M.*, Goldberg, M.C., Mostofsky, S.H., Cutting, L.E., & Denckla, M.B. Kennedy Krieger Institute, Baltimore, MD 21205.
Brain volume and function studies suggest that in autism there are abnormalities of the cerebellum and frontal lobes, areas that play a key role in motor processes, in autism. In this study, we hypothesized that children/adolescents with High-Functioning Autism (HFA) would show impaired motor coordination on the Physical and Neurological Examination for Subtle Signs (PANESS). We examined 8 to 17 year-olds (total 86), all with FSIQ > 80. Thirty-four subjects met ADI-R and ADOS-G criteria for autism (mean age = 10.97), and 52 had no medical diagnoses (mean age = 11.56). From the PANESS, tasks selected were: stressed gaits (e.g. on heels), one-footed balance and hopping, and timed (repetitive and patterned) movements (feet and hands). Mann-Whitney and Chi Square analyses were used to examine differences between groups. Results indicated that the HFA group showed more overall motor dysfunction than the control group, including a greater number of postural and balance errors on axial tasks (e.g. gaits, standing, hopping), increased overflow elicited by gaits and timed motor tasks, and more dysrhythmia during timed motor tasks. The HFA group was more impaired relative to controls on timed motor tasks involving the feet (e.g. foot tap, heel/toe tap) than the hands (e.g. hand pat, hand pronate-supinate, finger tap, and finger succession). Our results suggest that there are impairments in both the subcortical and cortical control of motor coordination in individuals with HFA.
S2.2.2
SENSORY OVERREACTIVITY, SEEKING, AND OVERFOCUSED ATTENTION IN PDD M. Liss*, C. Saulnier & D. Fein, Dept. of Psychology, Univ. of Connecticut, Stoors, CT 06269
Sensory abnormalities, especially over-reaction to sensory stimuli, are often observed and described in first person accounts of individuals with pervasive developmental disorders. However, systematic research on this phenomenon is sorely lacking. A recent theoretical model has hypothesized that sensory disturbances are related to overarousal and overfocused attention. This investigation examined parent report of sensory overreactivity, underreactivity and seeking behaviors in 252 individuals with PDD (range 1-23, mean age 8). Additionally, parents were given a measure of diagnostic severity, overselctive attention and an assessment of exceptional memory. Furthermore, 191 of the parents were administered the Vineland Adaptive Behavior Scales (VABS). A cluster analysis was performed, the results of which lend support to the overfocusing hypothesis. Approximately 10% of the sample had a strongly overfocused pattern of symptoms including overreactivity, perseverative behavior, overfocusing, and having an exceptional memory. An additional 33% of the sample had a mildly overfocused pattern. Hypotheses about the origins of overfocusing behavior are discussed.
S2.2.3
INFANT AND TODDLER ORAL AND MANUAL MOTOR DEVELOPMENT PREDICTS SPEECH OUTCOME. M. A. Gernsbacher, H. H. Goldsmith, E. A. Sauer, M. C. O’Reilly, and J. L. DeRuyter, Waisman Center, Univ. of Wisconsin, Madison, WI 53706.
The oral-motor and manual-motor development of a sample of 160 children with autism and 40 typically developing children were studied both retrospectively and concurrently. The retrospective data were collected via care-provider interview and analysis of home video. A total of 31 critical oral-motor and manual-motor markers were assayed at 6, 12, 18, 24, and 36 months of age and included failure to “blow raspberries” And grab dangling earrings at 6 months; clap at 12 months; produce animal sounds and draw horizontal lines at 18 months; suck liquids through a straw and turn round door knobs at 24 months; and blow out candles and catch and kick a medium-sized ball at 36 months. These critical oral-motor and manual-motor markers distinguished children with autism from typically developing children, clustered in a subgroup of children with autism, and predicted minimal current speech. Children from the subgroup who were characterized by the most oral-motor and manual-motor markers were examined in their homes (their ages ranged from 4:3 to 13:1), along with an age- and gender-matched comparison group of children from the autism sample who were characterized by the least oral-motor and manual-motor markers. Children from the subgroup who were characterized by the most oral-motor and manual-motor markers were severely impaired in their current-day performance on the Praxis Test of the Boston Diagnostic Aphasia Exam and the Kaufman Speech Praxis Text for Children. They all had minimal current-day speech.
S2.2.4
EVIDENCE FROM POSTURAL STABILITY TESTING FOR A VESTIBULAR DEFICIT IN CHILDREN WITH ASD. C. A. Molloy, K. Dietrich, A Bhattacharya. Depts. of Pediatrics and Environmental Health. Cincinnati Children’s Hospital Medical Center, University of Cincinnati, Ohio 45229
Problems with balance and postural stability are well described in children with Autism Spectrum Disorder (ASD). Maintaining upright posture is a complex process involving multiple afferent systems. The aim of this study was to measure the postural stability of children with ASD (cases) compared to children with typical neurodevelopment (controls), and to measure the relative contributions of the visual, proprioceptive and vestibular afferent systems in cases and controls. METHODS: Eight boys with ASD and eight age, race and gender matched controls participated in this study using force platform technology with customized software to measure vertical and horizontal forces and moments around the x,y and z axes. These measurements determine sway area, a measure of the movement of the body’s center of pressure in the horizontal plane during quiet standing. For all participants, sway area was measured under baseline conditions of normal afferent input and test conditions using blindfolds and foam padding to eliminate or modify visual and proprioceptive input. RESULTS: Children with ASD had a significantly larger mean sway area under all afferent conditions (p S2.2.5
EVALUATION OF MOTOR PROFICIENCY AND APRAXIA IN AUTISM: EFFECTS ON SIGN LANGUAGE ACQUISITION. L.V. Soorya and R.G. Romanczyk Institute for Child Development (ICD), SUNY-Binghamton, Binghamton, NY 13902
The present study further explores questions about the relationship between motor skills, apraxia, and sign language acquisition initially posed by Seal and Bonvillian (1997). The present study extends the previous research in a number of ways including the use of control groups and standardized instruction in sign language acquisition. Twelve children with autism (ages 3-10 yrs.) were matched to typically developing children in two control groups, matched on chronological age (CA) and mental age (MA). The study was conducted in two phases. Phase 1 included tests of motor proficiency and apraxia. Phase 2 included a six-session instructional program in sign language comprehension and production using ABA methods. Assessments of motor proficiency included the Bruininks-Oseretsky Test of Motor Proficiency and the Bayley Infant Development, Motor Scale. Measures of apraxia included the Kimura Movement Copying Test, Kimura Manual Sequencing Box, and the Apraxia subtest of the Boston Diagnostic Aphasia Exam. Analyses indicate significant differences between children with autism and CA controls on motor skills and measures of apraxia (p=.000). Children with autism did not differ significantly from MA controls on motor tests (p=.495) while differences on apraxia measures approached significance (p=.06). Analysis of sign language instruction data suggests a moderate, inverse correlation between apraxia and sign language production. Implications of the results will be discussed in terms of motor functioning in autism, the apraxia hypotheses for imitation problems, and individual and instructional variables affecting sign language instruction.
S2.2.6
DIFFERENCES IN POSTERIOR CORTICAL FMRI ACTIVATION DURING FINGER SEQUENCING IN CHILDREN WITH AUTISM. S. H. Mostofsky*, M.C. Goldberg, J. G. B. Schafer, A. C. Boyce, A. A. Flower, M. B. Denckla, J. J. Pekar. Kennedy Krieger Institute, Johns Hopkins Sch. Of Med., Baltimore, MD 21205.
Abnormalities in performance of skilled motor tasks (often referred to as “dyspraxia”) are well documented in autism. To investigate the neural basis of these deficits, fMRI was used to examine activation associated with finger sequencing in 12 children with autism (mean age 10.7 yrs) and 16 controls (mean age 10.3 yrs). Thirty-second blocks of right-handed finger sequencing (RHFS) were alternated with blocks of left-handed finger sequencing (LHFS) and blocks of rest. FMRI group data were analyzed using separate random effects models designed to determine signal change patterns associated with RHFS and LHFS in children with autism and controls. Random effects analyses were also used to examine between-group differences in activation during RHFS and LHFS. In children with autism and controls, activation (uncorrected p < 0.001) associated with RHFS and with LHFS was seen in the primary motor/sensory cortices (predominantly contralateral), supplementary motor area (bilateral), basal ganglia (bilateral), and cerebellum (predominantly ipsilateral). Statistical contrast between the two groups revealed that during both right and left finger sequencing controls have significantly greater activation near the border of the contralateral inferior parietal/superior temporal cortices, regions that provide perceptual representations of movement important for accurate performance of skilled motor actions. Our findings suggest that abnormalities in perceptual representations of movement may contribute to motor skill deficits (dyspraxia) observed in autism. S2.3.1ANALYSIS OF CANDIDATE GENES IN AutD LINKAGE REGION OF CHROMOSOME 2q Rabionet R (1), Menold MM (1), Raiford KL (1), Shao Y (1), Bost PL (1), Chua ET (1), Wolpert CM (1), Donnelly SL (1), Abramson RK (2), Wright HH (2), Cuccaro ML (1), Gilbert JR (1), Pericak-Vance MA* (1). (1) Dept of Medicine and Center for Human Genetics, Duke University Medical Center, Durham, NC 27710 (2) WS Hall Psychiatric Institute, University of South Carolina, Columbia, SC 29208
Genomic screens for autistic disorder (AutD) have shown linkage to chromosome 2q32-33. Support for an AutD risk gene in this region was found by stratifying the data sets based on delay in the acquisition of phrase speech (PSD). With this stratification, the HLOD score for D2S116 increased to 2.45 (recessive model) or 2.12 (dominant model). Several candidate genes including CTLA4 and CD28, both belonging to the Ig superfamily, DRAK2, a serine/threonine protein kinase involved in regulating apoptosis, BZAP45, a leucine-zipper protein, and ABI-2, an Abl interacting protein potentially involved in CNS development lie in the candidate region. The Collaborative Autism Team (CAT) is looking for association between AutD and these candidate genes. SNP analysis has been performed on CTLA4, CD28 and DRAK2 by oligonucleotide ligation assay (OLA) on a data set that includes 99 CAT multiplex families and 86 multiplex families from the Autism Genetic Resource Exchange (AGRE). Analysis of these data using the Pedigree Disequilibrium Test (PDT) showed no evidence of association at any of the loci (P>0.4 and P>0.5 respectively). Haplotype analysis using TRANSMIT, including PSD stratification, also demonstrated no association between the AutD and any particular haplotype. Expansion of this analysis, including application of TaqMan and pooling protocols, will be applied to other candidate genes in the region.
S2.3.2
WNT2 SEQUENCE DIFFERENCES BETWEEN AUTISM AND CONTROL SUBJECTS. W.M. McMahon, H. Coon, R. Weiss, D. Dunn, A. von Niederhausern, J. Lainhart, M. Leppert. NICHD* and Utah Autism Foundation,* Departments of Psychiatry and Human Genetics, University of Utah School of Medicine, Salt Lake City, UT 84108.
The WNT2 gene is located on chromosome 7q31-33 in a candidate region detected by several whole-genome scans of autism. The WNT family of genes encode secreted signaling proteins involved in development, and WNT genes are also implicated as candidates for autism by a mouse knockout (Dvl1) which displays some autistic-like behaviors. The coding sequence of WNT2 has been previously sequenced (Wassink et al., 2001), revealing two interesting coding sequence variants and positive evidence for linkage disequilibrium. In a sample of 70 autism subjects and 21 controls from the same population, we have re-sequenced a larger region after noting that a comparison of human and mouse WNT2 indicates high conservation of genomic sequence for 13 kb starting 3 kb upstream of exon 1. Five pairs of forward and reverse PCR primer sequences were chosen from the publicly available genomic sequence. A total of 66 internal primers were used to cover each product. Sixteen polymorphic positions were identified and verified by manual evaluation of the individual sequence traces. The missense mutations previously reported did not occur in our autism or control subjects. One new missense mutation was found in exon 5 in three of our autism subjects and none of our controls. Significant frequency differences were found between the autistic individuals vs. the control subjects for two polymorphisms in the 5′ flanking region, a proximal promotor polymorphism, a polymorphism in intron 3, two polymorphisms in the 3′ untranslated region, and two polymorphisms in the 3′ flanking sequence.
s2.3.3
The Premutation and Autistic Spectrum Disorders Hagerman, R., Goodlin-Jones, B.L., Spence, S., Albrect, L., Bacalman, S., Tassone, F., Gane, L., and Hagerman, P.
This paper reports on nine subjects, ages 3 years to 33 years, with the fragile X premutation and a diagnosis of Autistic Spectrum Disorder. In these 9 subjects the CGG repeats ranged form 56 to 180 while the FMRP levels were in the normal or mildly deficient range. All subjects exhibited features of fragile X syndrome, including cognitive and language problems, avoidance of eye contact, and some hand stereotypies. All subjects were assessed with standardized measures for autism (e.g. ADOS, ADI-R) as part of the clinical protocol. Five subjects made criterion for autism both clinically and with the standardized measures while three made criterion for autistic spectrum disorder (including Asperger and PDD-NOS). The subjects had autistic features including impairment in social reciprocity, repetitive behaviors, and impoverished social overtures which were observed in addition to the features of fragile X syndrome. Information will be presented about a range of mRNA elevations in these patients. One hypothesis may be that this association of autism and the fragile X premutation is related to the FMRP gene dysfunction, specifically a translation defect that begins in the premutation range. In addition, a second gene affect may be additive to the premutation to predispose patients to an autism spectrum disorder.
S2.3.4
ALTERATIONS IN IMPRINTED GENE EXPRESSION FOLLOWING MECP2 MUTATION MAY ACCOUNT FOR THE PHENOTYPIC OVERLAP BETWEEN AUTISM, RETT SYNDROME, & ANGELMAN SYNDROME. L.B.K. Herzing1 S-J. Kim2 E.H. Cook2 & D.H. Ledbetter3,. Depts. of Pediatrics1, Psychiatry2 & Human Genetics3. Univ. of Chicago, Chicago, IL 60637.
Phenotypic overlap between autism, Rett syndrome (RS) and Angelman syndrome (AS) has been spotlighted by the recent identification of MeCP2 (the RS gene) mutations in several patients with AS or autism, and evidence for linkage to UBE3A (the AS gene) in autism. UBE3A and ATP10C, maternally expressed imprinted genes regulated by the AS-imprinting center (IC), map to 15q11-q13, a strong candidate region for containing an autism gene. Semi-quantitative RT-PCR suggests that the ratio of maternal:paternal ATP10C expression differs between RS cell lines expressing wild-type or mutant MECP2 protein. Maternal ATP10C expression is considerably reduced in both RS lymphoblast (LB) clones expressing mutant MeCP2 and from a LB line carrying an AS-IC deletion. Imprinted expression of UBE3A is mediated by an overlapping antisense transcript, and we have identified a similar, paternally expressed antisense transcript overlapping ATP10C. Loss of maternal ATP10C expression in LBs with AS-IC and MeCP2 mutations is coordinate with a gain of expression of additional, maternal ATP10C antisense transcript in these cells. Furthermore, biallelic antisense transcription is also observed in cerebral tissue from male patients with MeCP2 mutations; analyses of gene expression levels in these tissues are ongoing. Together, these results suggest that MeCP2 mediates control of the maternally expressed genes on 15q11-13 through the AS-IC, and that loss of expression of these genes may contribute to chromosome 15-associated autism and the AS and RS phenotypes.
s2.3.5
Mapping of a Candidate Region for Autism on Chromosome 2q32. Sean Ennis1, 2, Louise Gallagher3, Judith Conroy3, Geraldine Kearney3, Michael Fitzgerald3, Raymond Stallings1, 2, David E Barton1, 2, Andrew J Green1, 2, and Michael Gill3 1Department of Medical Genetics, University College Dublin, Crumlin, Dublin 12, Ireland 2National Centre for Medical Genetics, Our Lady’s Hospital for Sick Children, Crumlin, Dublin 12, Ireland 3Department of Psychiatry, Trinity Centre for Health Sciences, St. James’s Hospital, Dublin 8.
Autism is a relatively common primary developmental disorder, with a significant genetic component. Routine investigations for the cause of autism, including chromosome and fragile X analysis, almost always are normal. There is a polygenic component to autism, and large sib-pair studies have been carried out by a number of groups worldwide. Several groups have identified associations between autism and a wide area of chromosome 2q, flanked by markers D2S364 and D2S2188 (Buxbaum et al & the IMGSAC respectively) which shows the strongest evidence for linkage. These results strongly suggest a predisposing gene(s) to autism within the 2q region. We have recently identified a patient with high-functioning autism, who has a small but cytogenetically visible de novo deletion of chromosome 2q32 which falls within these linkage findings. This would appear to be the smallest known deletion of this part of chromosome 2, suggesting that the deletion includes a predisposing gene(s) for autism. Fine mapping of the deleted region was carried out using markers from 2q32, to map the exact size of the deleted region. This reduces the very large linkage region by >75% to approximately 8.7 megabases. The region contains approximately 16 known genes and 16 ESTs, a number of which are potential candidates for autism. An initial fine mapping experiment of the region in a sample of Irish autistic trios involved 16 polymorphic microsattellite markers has been undertaken. Association analysis is currently being undertaken and the results of these analyses will also be presented.
S2.3.6
GENETIC ANALYSIS OF THE PARAOXONASE GENE CLUSTER IN AUTISTIC DISORDER. M. Conciatori1, L. D’Agruma2, L. Zelante2, R. Militerni3, C. Bravaccio3, C. Schneider4, R. Melmed4, S. Trillo5, F. Montecchi5, M. Elia6, M. Palermo7, T. Pascucci8,9, S. Puglisi-Allegra8,9, K-L. Reichelt10, L. Muscarella2, V. Guarnieri2, A.M. Persico1*. 1) Mol. Psychiatry & Neurogenet., UCBM, Rome, Italy; 2) Med. Genetics, IRCCS “CSS”, S.Giovanni Rotondo (FG), Italy; 3) Child Neuropsychiatry, II Univ. of Naples, Naples, Italy; 4) S.A.R.C., Phoenix, AZ, USA; 5) Child Neuropsychiatry, IRCCS “Ospedale Bambino-Ges˘”, Rome, Italy; 6) Neurology, IRCCS “Oasi Maria S.S.”, Troina (EN), Italy; 7) Assoc. “Anni Verdi”, Rome, Italy; 8) Dept. of Psychol., Univ. “La Sapienza”, Rome, Italy; 9) IRCCS “Fond. S. Lucia”, Rome, Italy; 10) Pediatric Res., Univ. of Oslo, Norway.
We have previously shown genetic association/linkage between autism and Reelin gene alleles resulting in reduced protein synthesis. Reelin’s serine protease activity, critical to neuronal migration, is inhibited by organophosphates. These compounds, routinely used as pesticides, could thus act upon vulnerable individuals during critical periods in prenatal neurodevelopment, yielding an autistic syndrome. Organophosphates are inactivated by serum paraoxonase (PON1), an LDL-bound enzyme physiologically acting as a lipid oxidizer and against bacterial endotoxins. The PON1 gene is located on chromosome 7q21.3, in a gene cluster also encompassing PON2 and PON3, whose function is less clear. We are currently genotyping the functional PON1 ñ108 C/T, PON1 192 Q/R and PON2 311 C/S polymorphisms for TDT and haplotype association analyses. Initial results from the PON1 ñ108 C/T polymorphism yield no evidence of linkage/association S2.4.1JOINT ATTENTION, SOCIAL REFERENCING AND REQUESTING ABILITIES IN YOUNG CHILDREN WITH AN AUTISM SPECTRUM DISORDER. P. Warreyn and H. Roeyers Research Group Developmental Disorders, Ghent University. Ghent, Belgium, B-9000.
Introduction: During the last decade, there has been growing scientific interest in the early social-communicative abilities of young children with an autism spectrum disorder (ASD). Especially the ability to share attention with others is seen as crucial for the later development of language and perspective taking. Aim: The aim of this study was to investigate the different aspects of sharing attention: joint attention, requesting and social referencing and their interrelation in young children (2y1m – 5y6m) with an ASD. Method: The children were invited with their mothers to the lab. The early social-communicative abilities were elicited in a 40-minute session. During the trials, mothers were instructed through a wireless earphone. All sessions were videotaped, and coded afterwards with the Observer Video Pro, a computer program enabling very detailed frame-by-frame coding. The children with ASD were compared with a group of age-matched children, who had been referred to an infant psychiatric unit, but who did not show any signs of ASD. Results: The children with ASD looked, pointed and vocalized almost as much as did the control children. Especially the older, more verbal children with an ASD initiated many instances of social interaction with their mothers, while some of the (clinical) control children communicated very little. However, qualitative differences in the coordination of looking behavior were found between both groups.
S2.4.2
Concurrent and longitudinal associations between joint attention, imitation and play in infancy and language in the pre-school years in children with autism spectrum disorder Tony Charman, Simon Baron-Cohen, John Swettenham, Gillian Baird, Auriol Drew, Antony Cox
Primary objective: To examine concurrent and longitudinal associations between joint attention, play and imitation abilities and language from infancy to the pre-school years in children with autism spectrum disorder. Research design: A within-group cross-sectional and longitudinal correlational design was adopted. Methods and procedures: Experimental measures of joint attention, play and imitation were conducted with a sample of infants with autism spectrum disorder at age 20 months. Language competence was assessed at age 20 months and again at age 42 months. Main outcomes and results: Only a measure of joint attention ability was concurrently associated with language ability at age 20 months. Longitudinally, measures of both joint attention and imitation were associated with later expressive and receptive language ability, even when the effects of language and IQ at the initial timepoint had been partialled out. Conclusions: The results confirmed findings from previous studies with older pre-school children with autism spectrum disorder. Current and longitudinal associations between infant social-communication abilities and language development suggest that they may be important prognostic factors and targets for intervention.
S2.4.3
A LONGITUDINAL STUDY OF PRETEND PLAY IN AUTISM M.D. Rutherford and S.J. Rogers, JFK Partners University of Colorado Health Sciences Center, Denver CO 80262
In this study theories that consider either theory of mind (ToM) or executive function (EF) important deficits in autism and important factors in the development of pretend play are considered and compared. Each of these two theories posits a cognitive precursor to pretense which should be present in typical development, and whose absence could explain pretend play deficits in children with autism. We tested which of these theories better predicts a child’s production of pretend play in a longitudinal study. At time 1 (approx. 2 yrs MA), the subjects were 28 children with idiopathic autism (IA), 24 children with other developmental disorders (DD), and 26 typically developing children. At time 2 (approx. 4 yrs MA) there were 21 in the IA group, 13 in the DD group, and 9 in the typical group. The Fewell play scale was used to measure pretend play. Joint attention measures were our index of ToM development. Measures of EF were a generativity task and a spatial reversal task. With data from all three groups combined, joint attention at time 1 was correlated with pretend play at time 2 significantly and more strongly than any other time 1 measure. In a regression analysis, joint attention at time 1 was the strongest predictor of pretend play at time 2; time 1 generativity scores also significantly predicted time 2 pretend play scores. Considering data from the autism group alone, joint attention was the only time 1 variable to correlate significantly with time 2 pretend play. A joint attention deficit, but no EF deficit is measurable early enough to explain the pretend play deficit in autism. One of the EF measures, the generativity task, had a strong relationship with pretend play scores at both time 1 and time 2, so one cannot rule out the idea that EF contributes to pretend play performance. At time 1, joint attention scores, our index of ToM development, best predicted pretend play at time 2, favoring the ToM theory of the development of pretend play in autism.
S2.4.4
RESPONSE TO JOINT ATTENTION MODERATES THE RELATION BETWEEN AMOUNT OF INTERVENTION AND GAIN IN LANGUAGE AGE IN YOUNG CHILDREN WITH AUTISM. M. A. Bono, T. Daley, L. Travis, & M. Sigman UCLA/NPI, Los Angeles, CA 90024.
Joint attention skills were hypothesized to predict language gain and moderate the effect of amount of intervention on language gain in 29 young children with autism. All children met DSM-IV criteria for a clinical diagnosis of autism and had above-cutoff scores on either the ADI-R or the ADOS. Children were assessed at the beginning and the end of a one-year interval. Children were, on average, 46.68 months of age (SD = 9.64) at the first evaluation and 59.06 months of age (SD = 9.98) at the second. Initiation and response to joint attention were assessed at the first visit with the Early Social Communication Scales. Intervention data were collected from parents’ reports of the amount of time per week children participated in structured group and individual settings. Language abilities were assessed twice with either the Reynell or the CELF-R and total language age was employed as the dependent variable. Difference scores were calculated as a measure of change. More initiation of joint attention and a more consistent response to joint attention were associated with greater gain in language age, r = .37, p < .05 and r = 64, p < .001, respectively. The relation between amount of intervention and gain in language age was weak, r = .18, p = .33. Linear regression analyses suggested that the relation between intervention and language gain depended upon children’s ability to respond to bids for joint attention, F (3, 25) = 11.9, p < .001, R2 = .59. The findings suggest that autistic children with more consistent responses to bids for joint attention gain more language as a result of interventions than others with less consistent responses.
S2.4.5
COMPARISON OF INDIVIDUALS WITH ASPERGER SYNDROME AND AUTISM ON MEASURES SENSITIVE TO THE ABILITY TO SHIFT ATTENTION E. Hanson,* A. Lincoln Children’s Hospital, Harvard Medical School, 300 Longwood Avenue, Boston, MA 02115
There have been a number of studies showing the difficulty that individuals with Autism (AD) have in shifting attention rapidly, especially when required to do so in a very short (less than 2.5 second) time frame. (Akshoomoff & Courchesne, 1994; Casey, Gordon, Mannheim & Rumsey, 1993; Townsend & Courchesne, 1994). This method has been shown to be sensitive to cerebellar abnormalities. The primary goal of this study was to replicate these earlier findings and to evaluate whether individuals with Asperger Syndrome (AS) also showed impairment in their ability to rapidly shift attention. Findings from the present study supported the replication hypothesis that AD individuals would have a lower percent accuracy in the shifting conditions at the fastest time interval than normal controls (NC). AD individuals performed similarly to the other groups in longer time bins. Their sustained attention appeared intact as did their ability to shift attention given more time. The study did not support the hypothesis that AS participants would perform differently from AD and NC individuals in the fastest time bins and shifting attention tasks did not appear to be sensitive to differences between these groups. However, there were a number of interesting trends in the data. In particular, AS individuals tended to perform better on auditory shifting tasks in the fastest time bin but not in the visual shifting modality. Other trends were for AS individuals to have more difficulty with sustained attention vs. AD individuals who had difficulty with shifting attention.
S2.4.6
TOWARDS A MODEL OF SHARED ATTENTION IN INFANCY: TYPICAL DEVELOPMENT AND AUTISM G. O. De·k, J. Triesch, J. Movellan, I. Fasel, Dept. of Cognitive Science & INC, UCSD, La Jolla, CA 92093-0515
A deficit in attention sharing is among the first behavioral signs of infantile autism. To understand this we need a plausible model of the emergence of shared attention during normal development. We propose that gaze following, an early form of attention sharing, emerges from natural social interactions. Through generic learning mechanisms, and perceptual and motivational biases, infants learn that the parent’s gaze direction predicts the location of rewarding stimuli. We are combining behavioral evidence and computational models to test the plausibility of this proposal. In an observational study of healthy 3- to 12-month-old infants, we are deriving patterns of parents’ and infants’ attention shifting during social interactions. The observed patterns of parental behavior are used as input to train a computational model. This model is in the form of a computer-simulated virtual “infant” endowed with a typical 3-month-old’s perceptual, motivational, and learning capacities. Using this model we can determine under what conditions gaze following is learned, given a pattern of realistic social input. By perturbing parameters of learning and perceptual bias in the model, we can describe the conditions under which gaze following fails to emerge. This will offer clues to possible learning and attentional underpinnings of autism, which can be tested in behavioral experiments.
SLIDE SESSION 2.2: SENSORY & MOTOR 1
S2.2.1
IMPAIRMENTS IN BOTH CORTICAL AND SUBCORTICAL MOTOR CONTROL IN AUTISM Jansiewicz, E.M.*, Goldberg, M.C., Mostofsky, S.H., Cutting, L.E., & Denckla, M.B. Kennedy Krieger Institute, Baltimore, MD 21205.
Brain volume and function studies suggest that in autism there are abnormalities of the cerebellum and frontal lobes, areas that play a key role in motor processes, in autism. In this study, we hypothesized that children/adolescents with High-Functioning Autism (HFA) would show impaired motor coordination on the Physical and Neurological Examination for Subtle Signs (PANESS). We examined 8 to 17 year-olds (total 86), all with FSIQ > 80. Thirty-four subjects met ADI-R and ADOS-G criteria for autism (mean age = 10.97), and 52 had no medical diagnoses (mean age = 11.56). From the PANESS, tasks selected were: stressed gaits (e.g. on heels), one-footed balance and hopping, and timed (repetitive and patterned) movements (feet and hands). Mann-Whitney and Chi Square analyses were used to examine differences between groups. Results indicated that the HFA group showed more overall motor dysfunction than the control group, including a greater number of postural and balance errors on axial tasks (e.g. gaits, standing, hopping), increased overflow elicited by gaits and timed motor tasks, and more dysrhythmia during timed motor tasks. The HFA group was more impaired relative to controls on timed motor tasks involving the feet (e.g. foot tap, heel/toe tap) than the hands (e.g. hand pat, hand pronate-supinate, finger tap, and finger succession). Our results suggest that there are impairments in both the subcortical and cortical control of motor coordination in individuals with HFA.
S2.2.2
SENSORY OVERREACTIVITY, SEEKING, AND OVERFOCUSED ATTENTION IN PDD M. Liss*, C. Saulnier & D. Fein, Dept. of Psychology, Univ. of Connecticut, Stoors, CT 06269
Sensory abnormalities, especially over-reaction to sensory stimuli, are often observed and described in first person accounts of individuals with pervasive developmental disorders. However, systematic research on this phenomenon is sorely lacking. A recent theoretical model has hypothesized that sensory disturbances are related to overarousal and overfocused attention. This investigation examined parent report of sensory overreactivity, underreactivity and seeking behaviors in 252 individuals with PDD (range 1-23, mean age 8). Additionally, parents were given a measure of diagnostic severity, overselctive attention and an assessment of exceptional memory. Furthermore, 191 of the parents were administered the Vineland Adaptive Behavior Scales (VABS). A cluster analysis was performed, the results of which lend support to the overfocusing hypothesis. Approximately 10% of the sample had a strongly overfocused pattern of symptoms including overreactivity, perseverative behavior, overfocusing, and having an exceptional memory. An additional 33% of the sample had a mildly overfocused pattern. Hypotheses about the origins of overfocusing behavior are discussed.
S2.2.3
INFANT AND TODDLER ORAL AND MANUAL MOTOR DEVELOPMENT PREDICTS SPEECH OUTCOME. M. A. Gernsbacher, H. H. Goldsmith, E. A. Sauer, M. C. O’Reilly, and J. L. DeRuyter, Waisman Center, Univ. of Wisconsin, Madison, WI 53706.
The oral-motor and manual-motor development of a sample of 160 children with autism and 40 typically developing children were studied both retrospectively and concurrently. The retrospective data were collected via care-provider interview and analysis of home video. A total of 31 critical oral-motor and manual-motor markers were assayed at 6, 12, 18, 24, and 36 months of age and included failure to “blow raspberries” And grab dangling earrings at 6 months; clap at 12 months; produce animal sounds and draw horizontal lines at 18 months; suck liquids through a straw and turn round door knobs at 24 months; and blow out candles and catch and kick a medium-sized ball at 36 months. These critical oral-motor and manual-motor markers distinguished children with autism from typically developing children, clustered in a subgroup of children with autism, and predicted minimal current speech. Children from the subgroup who were characterized by the most oral-motor and manual-motor markers were examined in their homes (their ages ranged from 4:3 to 13:1), along with an age- and gender-matched comparison group of children from the autism sample who were characterized by the least oral-motor and manual-motor markers. Children from the subgroup who were characterized by the most oral-motor and manual-motor markers were severely impaired in their current-day performance on the Praxis Test of the Boston Diagnostic Aphasia Exam and the Kaufman Speech Praxis Text for Children. They all had minimal current-day speech.
S2.2.4
EVIDENCE FROM POSTURAL STABILITY TESTING FOR A VESTIBULAR DEFICIT IN CHILDREN WITH ASD. C. A. Molloy, K. Dietrich, A Bhattacharya. Depts. of Pediatrics and Environmental Health. Cincinnati Children’s Hospital Medical Center, University of Cincinnati, Ohio 45229
Problems with balance and postural stability are well described in children with Autism Spectrum Disorder (ASD). Maintaining upright posture is a complex process involving multiple afferent systems. The aim of this study was to measure the postural stability of children with ASD (cases) compared to children with typical neurodevelopment (controls), and to measure the relative contributions of the visual, proprioceptive and vestibular afferent systems in cases and controls. METHODS: Eight boys with ASD and eight age, race and gender matched controls participated in this study using force platform technology with customized software to measure vertical and horizontal forces and moments around the x,y and z axes. These measurements determine sway area, a measure of the movement of the body’s center of pressure in the horizontal plane during quiet standing. For all participants, sway area was measured under baseline conditions of normal afferent input and test conditions using blindfolds and foam padding to eliminate or modify visual and proprioceptive input. RESULTS: Children with ASD had a significantly larger mean sway area under all afferent conditions (p S2.2.5
EVALUATION OF MOTOR PROFICIENCY AND APRAXIA IN AUTISM: EFFECTS ON SIGN LANGUAGE ACQUISITION. L.V. Soorya and R.G. Romanczyk Institute for Child Development (ICD), SUNY-Binghamton, Binghamton, NY 13902
The present study further explores questions about the relationship between motor skills, apraxia, and sign language acquisition initially posed by Seal and Bonvillian (1997). The present study extends the previous research in a number of ways including the use of control groups and standardized instruction in sign language acquisition. Twelve children with autism (ages 3-10 yrs.) were matched to typically developing children in two control groups, matched on chronological age (CA) and mental age (MA). The study was conducted in two phases. Phase 1 included tests of motor proficiency and apraxia. Phase 2 included a six-session instructional program in sign language comprehension and production using ABA methods. Assessments of motor proficiency included the Bruininks-Oseretsky Test of Motor Proficiency and the Bayley Infant Development, Motor Scale. Measures of apraxia included the Kimura Movement Copying Test, Kimura Manual Sequencing Box, and the Apraxia subtest of the Boston Diagnostic Aphasia Exam. Analyses indicate significant differences between children with autism and CA controls on motor skills and measures of apraxia (p=.000). Children with autism did not differ significantly from MA controls on motor tests (p=.495) while differences on apraxia measures approached significance (p=.06). Analysis of sign language instruction data suggests a moderate, inverse correlation between apraxia and sign language production. Implications of the results will be discussed in terms of motor functioning in autism, the apraxia hypotheses for imitation problems, and individual and instructional variables affecting sign language instruction.
S2.2.6
DIFFERENCES IN POSTERIOR CORTICAL FMRI ACTIVATION DURING FINGER SEQUENCING IN CHILDREN WITH AUTISM. S. H. Mostofsky*, M.C. Goldberg, J. G. B. Schafer, A. C. Boyce, A. A. Flower, M. B. Denckla, J. J. Pekar. Kennedy Krieger Institute, Johns Hopkins Sch. Of Med., Baltimore, MD 21205.
Abnormalities in performance of skilled motor tasks (often referred to as “dyspraxia”) are well documented in autism. To investigate the neural basis of these deficits, fMRI was used to examine activation associated with finger sequencing in 12 children with autism (mean age 10.7 yrs) and 16 controls (mean age 10.3 yrs). Thirty-second blocks of right-handed finger sequencing (RHFS) were alternated with blocks of left-handed finger sequencing (LHFS) and blocks of rest. FMRI group data were analyzed using separate random effects models designed to determine signal change patterns associated with RHFS and LHFS in children with autism and controls. Random effects analyses were also used to examine between-group differences in activation during RHFS and LHFS. In children with autism and controls, activation (uncorrected p < 0.001) associated with RHFS and with LHFS was seen in the primary motor/sensory cortices (predominantly contralateral), supplementary motor area (bilateral), basal ganglia (bilateral), and cerebellum (predominantly ipsilateral). Statistical contrast between the two groups revealed that during both right and left finger sequencing controls have significantly greater activation near the border of the contralateral inferior parietal/superior temporal cortices, regions that provide perceptual representations of movement important for accurate performance of skilled motor actions. Our findings suggest that abnormalities in perceptual representations of movement may contribute to motor skill deficits (dyspraxia) observed in autism.
SLIDE SESSION 2.3: GENETICS 2
S2.3.1
ANALYSIS OF CANDIDATE GENES IN AutD LINKAGE REGION OF CHROMOSOME 2q Rabionet R (1), Menold MM (1), Raiford KL (1), Shao Y (1), Bost PL (1), Chua ET (1), Wolpert CM (1), Donnelly SL (1), Abramson RK (2), Wright HH (2), Cuccaro ML (1), Gilbert JR (1), Pericak-Vance MA* (1). (1) Dept of Medicine and Center for Human Genetics, Duke University Medical Center, Durham, NC 27710 (2) WS Hall Psychiatric Institute, University of South Carolina, Columbia, SC 29208
Genomic screens for autistic disorder (AutD) have shown linkage to chromosome 2q32-33. Support for an AutD risk gene in this region was found by stratifying the data sets based on delay in the acquisition of phrase speech (PSD). With this stratification, the HLOD score for D2S116 increased to 2.45 (recessive model) or 2.12 (dominant model). Several candidate genes including CTLA4 and CD28, both belonging to the Ig superfamily, DRAK2, a serine/threonine protein kinase involved in regulating apoptosis, BZAP45, a leucine-zipper protein, and ABI-2, an Abl interacting protein potentially involved in CNS development lie in the candidate region. The Collaborative Autism Team (CAT) is looking for association between AutD and these candidate genes. SNP analysis has been performed on CTLA4, CD28 and DRAK2 by oligonucleotide ligation assay (OLA) on a data set that includes 99 CAT multiplex families and 86 multiplex families from the Autism Genetic Resource Exchange (AGRE). Analysis of these data using the Pedigree Disequilibrium Test (PDT) showed no evidence of association at any of the loci (P>0.4 and P>0.5 respectively). Haplotype analysis using TRANSMIT, including PSD stratification, also demonstrated no association between the AutD and any particular haplotype. Expansion of this analysis, including application of TaqMan and pooling protocols, will be applied to other candidate genes in the region.
S2.3.2
WNT2 SEQUENCE DIFFERENCES BETWEEN AUTISM AND CONTROL SUBJECTS. W.M. McMahon, H. Coon, R. Weiss, D. Dunn, A. von Niederhausern, J. Lainhart, M. Leppert. NICHD* and Utah Autism Foundation,* Departments of Psychiatry and Human Genetics, University of Utah School of Medicine, Salt Lake City, UT 84108.
The WNT2 gene is located on chromosome 7q31-33 in a candidate region detected by several whole-genome scans of autism. The WNT family of genes encode secreted signaling proteins involved in development, and WNT genes are also implicated as candidates for autism by a mouse knockout (Dvl1) which displays some autistic-like behaviors. The coding sequence of WNT2 has been previously sequenced (Wassink et al., 2001), revealing two interesting coding sequence variants and positive evidence for linkage disequilibrium. In a sample of 70 autism subjects and 21 controls from the same population, we have re-sequenced a larger region after noting that a comparison of human and mouse WNT2 indicates high conservation of genomic sequence for 13 kb starting 3 kb upstream of exon 1. Five pairs of forward and reverse PCR primer sequences were chosen from the publicly available genomic sequence. A total of 66 internal primers were used to cover each product. Sixteen polymorphic positions were identified and verified by manual evaluation of the individual sequence traces. The missense mutations previously reported did not occur in our autism or control subjects. One new missense mutation was found in exon 5 in three of our autism subjects and none of our controls. Significant frequency differences were found between the autistic individuals vs. the control subjects for two polymorphisms in the 5′ flanking region, a proximal promotor polymorphism, a polymorphism in intron 3, two polymorphisms in the 3′ untranslated region, and two polymorphisms in the 3′ flanking sequence.
S2.3.3
THE PREMUTATION AND AUTISTIC SPECTRUM DISORDERS Hagerman, R., Goodlin-Jones, B.L., Spence, S., Albrect, L., Bacalman, S., Tassone, F., Gane, L., and Hagerman, P.
This paper reports on nine subjects, ages 3 years to 33 years, with the fragile X premutation and a diagnosis of Autistic Spectrum Disorder. In these 9 subjects the CGG repeats ranged form 56 to 180 while the FMRP levels were in the normal or mildly deficient range. All subjects exhibited features of fragile X syndrome, including cognitive and language problems, avoidance of eye contact, and some hand stereotypies. All subjects were assessed with standardized measures for autism (e.g. ADOS, ADI-R) as part of the clinical protocol. Five subjects made criterion for autism both clinically and with the standardized measures while three made criterion for autistic spectrum disorder (including Asperger and PDD-NOS). The subjects had autistic features including impairment in social reciprocity, repetitive behaviors, and impoverished social overtures which were observed in addition to the features of fragile X syndrome. Information will be presented about a range of mRNA elevations in these patients. One hypothesis may be that this association of autism and the fragile X premutation is related to the FMRP gene dysfunction, specifically a translation defect that begins in the premutation range. In addition, a second gene affect may be additive to the premutation to predispose patients to an autism spectrum disorder.
S2.3.4
ALTERATIONS IN IMPRINTED GENE EXPRESSION FOLLOWING MECP2 MUTATION MAY ACCOUNT FOR THE PHENOTYPIC OVERLAP BETWEEN AUTISM, RETT SYNDROME, & ANGELMAN SYNDROME. L.B.K. Herzing1 S-J. Kim2 E.H. Cook2 & D.H. Ledbetter3,. Depts. of Pediatrics1, Psychiatry2 & Human Genetics3. Univ. of Chicago, Chicago, IL 60637.
Phenotypic overlap between autism, Rett syndrome (RS) and Angelman syndrome (AS) has been spotlighted by the recent identification of MeCP2 (the RS gene) mutations in several patients with AS or autism, and evidence for linkage to UBE3A (the AS gene) in autism. UBE3A and ATP10C, maternally expressed imprinted genes regulated by the AS-imprinting center (IC), map to 15q11-q13, a strong candidate region for containing an autism gene. Semi-quantitative RT-PCR suggests that the ratio of maternal:paternal ATP10C expression differs between RS cell lines expressing wild-type or mutant MECP2 protein. Maternal ATP10C expression is considerably reduced in both RS lymphoblast (LB) clones expressing mutant MeCP2 and from a LB line carrying an AS-IC deletion. Imprinted expression of UBE3A is mediated by an overlapping antisense transcript, and we have identified a similar, paternally expressed antisense transcript overlapping ATP10C. Loss of maternal ATP10C expression in LBs with AS-IC and MeCP2 mutations is coordinate with a gain of expression of additional, maternal ATP10C antisense transcript in these cells. Furthermore, biallelic antisense transcription is also observed in cerebral tissue from male patients with MeCP2 mutations; analyses of gene expression levels in these tissues are ongoing. Together, these results suggest that MeCP2 mediates control of the maternally expressed genes on 15q11-13 through the AS-IC, and that loss of expression of these genes may contribute to chromosome 15-associated autism and the AS and RS phenotypes.
S2.3.5
MAPPING OF A CANDIDATE REGION FOR AUTISM ON CHROMOSOME 2Q32. Sean Ennis1, 2, Louise Gallagher3, Judith Conroy3, Geraldine Kearney3, Michael Fitzgerald3, Raymond Stallings1, 2, David E Barton1, 2, Andrew J Green1, 2, and Michael Gill3 1Department of Medical Genetics, University College Dublin, Crumlin, Dublin 12, Ireland 2National Centre for Medical Genetics, Our Lady’s Hospital for Sick Children, Crumlin, Dublin 12, Ireland 3Department of Psychiatry, Trinity Centre for Health Sciences, St. James’s Hospital, Dublin 8.
Autism is a relatively common primary developmental disorder, with a significant genetic component. Routine investigations for the cause of autism, including chromosome and fragile X analysis, almost always are normal. There is a polygenic component to autism, and large sib-pair studies have been carried out by a number of groups worldwide. Several groups have identified associations between autism and a wide area of chromosome 2q, flanked by markers D2S364 and D2S2188 (Buxbaum et al & the IMGSAC respectively) which shows the strongest evidence for linkage. These results strongly suggest a predisposing gene(s) to autism within the 2q region. We have recently identified a patient with high-functioning autism, who has a small but cytogenetically visible de novo deletion of chromosome 2q32 which falls within these linkage findings. This would appear to be the smallest known deletion of this part of chromosome 2, suggesting that the deletion includes a predisposing gene(s) for autism. Fine mapping of the deleted region was carried out using markers from 2q32, to map the exact size of the deleted region. This reduces the very large linkage region by >75% to approximately 8.7 megabases. The region contains approximately 16 known genes and 16 ESTs, a number of which are potential candidates for autism. An initial fine mapping experiment of the region in a sample of Irish autistic trios involved 16 polymorphic microsattellite markers has been undertaken. Association analysis is currently being undertaken and the results of these analyses will also be presented.
S2.3.6
GENETIC ANALYSIS OF THE PARAOXONASE GENE CLUSTER IN AUTISTIC DISORDER. M. Conciatori1, L. D’Agruma2, L. Zelante2, R. Militerni3, C. Bravaccio3, C. Schneider4, R. Melmed4, S. Trillo5, F. Montecchi5, M. Elia6, M. Palermo7, T. Pascucci8,9, S. Puglisi-Allegra8,9, K-L. Reichelt10, L. Muscarella2, V. Guarnieri2, A.M. Persico1*. 1) Mol. Psychiatry & Neurogenet., UCBM, Rome, Italy; 2) Med. Genetics, IRCCS “CSS”, S.Giovanni Rotondo (FG), Italy; 3) Child Neuropsychiatry, II Univ. of Naples, Naples, Italy; 4) S.A.R.C., Phoenix, AZ, USA; 5) Child Neuropsychiatry, IRCCS “Ospedale Bambino-Gesù”, Rome, Italy; 6) Neurology, IRCCS “Oasi Maria S.S.”, Troina (EN), Italy; 7) Assoc. “Anni Verdi”, Rome, Italy; 8) Dept. of Psychol., Univ. “La Sapienza”, Rome, Italy; 9) IRCCS “Fond. S. Lucia”, Rome, Italy; 10) Pediatric Res., Univ. of Oslo, Norway.
We have previously shown genetic association/linkage between autism and Reelin gene alleles resulting in reduced protein synthesis. Reelin’s serine protease activity, critical to neuronal migration, is inhibited by organophosphates. These compounds, routinely used as pesticides, could thus act upon vulnerable individuals during critical periods in prenatal neurodevelopment, yielding an autistic syndrome. Organophosphates are inactivated by serum paraoxonase (PON1), an LDL-bound enzyme physiologically acting as a lipid oxidizer and against bacterial endotoxins. The PON1 gene is located on chromosome 7q21.3, in a gene cluster also encompassing PON2 and PON3, whose function is less clear. We are currently genotyping the functional PON1 -108 C/T, PON1 192 Q/R and PON2 311 C/S polymorphisms for TDT and haplotype association analyses. Initial results from the PON1 -108 C/T polymorphism yield no evidence of linkage/association (TDT chi sq=0.15, 1df, P=0.70, n.s.).
SLIDE SESSION 2.4: JOINT ATTENTION
S2.4.1
JOINT ATTENTION, SOCIAL REFERENCING AND REQUESTING ABILITIES IN YOUNG CHILDREN WITH AN AUTISM SPECTRUM DISORDER. P. Warreyn and H. Roeyers Research Group Developmental Disorders, Ghent University. Ghent, Belgium, B-9000.
Introduction: During the last decade, there has been growing scientific interest in the early social-communicative abilities of young children with an autism spectrum disorder (ASD). Especially the ability to share attention with others is seen as crucial for the later development of language and perspective taking. Aim: The aim of this study was to investigate the different aspects of sharing attention: joint attention, requesting and social referencing and their interrelation in young children (2y1m – 5y6m) with an ASD. Method: The children were invited with their mothers to the lab. The early social-communicative abilities were elicited in a 40-minute session. During the trials, mothers were instructed through a wireless earphone. All sessions were videotaped, and coded afterwards with the Observer Video Pro, a computer program enabling very detailed frame-by-frame coding. The children with ASD were compared with a group of age-matched children, who had been referred to an infant psychiatric unit, but who did not show any signs of ASD. Results: The children with ASD looked, pointed and vocalized almost as much as did the control children. Especially the older, more verbal children with an ASD initiated many instances of social interaction with their mothers, while some of the (clinical) control children communicated very little. However, qualitative differences in the coordination of looking behavior were found between both groups.
S2.4.2
CONCURRENT AND LONGITUDINAL ASSOCIATIONS BETWEEN JOINT ATTENTION, IMITATION AND PLAY IN INFANCY AND LANGUAGE IN THE PRE-SCHOOL YEARS IN CHILDREN WITH AUTISM SPECTRUM DISORDER Tony Charman, Simon Baron-Cohen, John Swettenham, Gillian Baird, Auriol Drew, Antony Cox
Primary objective: To examine concurrent and longitudinal associations between joint attention, play and imitation abilities and language from infancy to the pre-school years in children with autism spectrum disorder. Research design: A within-group cross-sectional and longitudinal correlational design was adopted. Methods and procedures: Experimental measures of joint attention, play and imitation were conducted with a sample of infants with autism spectrum disorder at age 20 months. Language competence was assessed at age 20 months and again at age 42 months. Main outcomes and results: Only a measure of joint attention ability was concurrently associated with language ability at age 20 months. Longitudinally, measures of both joint attention and imitation were associated with later expressive and receptive language ability, even when the effects of language and IQ at the initial timepoint had been partialled out. Conclusions: The results confirmed findings from previous studies with older pre-school children with autism spectrum disorder. Current and longitudinal associations between infant social-communication abilities and language development suggest that they may be important prognostic factors and targets for intervention.
S2.4.3
A LONGITUDINAL STUDY OF PRETEND PLAY IN AUTISM M.D. Rutherford and S.J. Rogers, JFK Partners University of Colorado Health Sciences Center, Denver CO 80262
In this study theories that consider either theory of mind (ToM) or executive function (EF) important deficits in autism and important factors in the development of pretend play are considered and compared. Each of these two theories posits a cognitive precursor to pretense which should be present in typical development, and whose absence could explain pretend play deficits in children with autism. We tested which of these theories better predicts a child’s production of pretend play in a longitudinal study. At time 1 (approx. 2 yrs MA), the subjects were 28 children with idiopathic autism (IA), 24 children with other developmental disorders (DD), and 26 typically developing children. At time 2 (approx. 4 yrs MA) there were 21 in the IA group, 13 in the DD group, and 9 in the typical group. The Fewell play scale was used to measure pretend play. Joint attention measures were our index of ToM development. Measures of EF were a generativity task and a spatial reversal task. With data from all three groups combined, joint attention at time 1 was correlated with pretend play at time 2 significantly and more strongly than any other time 1 measure. In a regression analysis, joint attention at time 1 was the strongest predictor of pretend play at time 2; time 1 generativity scores also significantly predicted time 2 pretend play scores. Considering data from the autism group alone, joint attention was the only time 1 variable to correlate significantly with time 2 pretend play. A joint attention deficit, but no EF deficit is measurable early enough to explain the pretend play deficit in autism. One of the EF measures, the generativity task, had a strong relationship with pretend play scores at both time 1 and time 2, so one cannot rule out the idea that EF contributes to pretend play performance. At time 1, joint attention scores, our index of ToM development, best predicted pretend play at time 2, favoring the ToM theory of the development of pretend play in autism.
S2.4.4
RESPONSE TO JOINT ATTENTION MODERATES THE RELATION BETWEEN AMOUNT OF INTERVENTION AND GAIN IN LANGUAGE AGE IN YOUNG CHILDREN WITH AUTISM. M. A. Bono, T. Daley, L. Travis, & M. Sigman UCLA/NPI, Los Angeles, CA 90024.
Joint attention skills were hypothesized to predict language gain and moderate the effect of amount of intervention on language gain in 29 young children with autism. All children met DSM-IV criteria for a clinical diagnosis of autism and had above-cutoff scores on either the ADI-R or the ADOS. Children were assessed at the beginning and the end of a one-year interval. Children were, on average, 46.68 months of age (SD = 9.64) at the first evaluation and 59.06 months of age (SD = 9.98) at the second. Initiation and response to joint attention were assessed at the first visit with the Early Social Communication Scales. Intervention data were collected from parents’ reports of the amount of time per week children participated in structured group and individual settings. Language abilities were assessed twice with either the Reynell or the CELF-R and total language age was employed as the dependent variable. Difference scores were calculated as a measure of change. More initiation of joint attention and a more consistent response to joint attention were associated with greater gain in language age, r = .37, p < .05 and r = 64, p < .001, respectively. The relation between amount of intervention and gain in language age was weak, r = .18, p = .33. Linear regression analyses suggested that the relation between intervention and language gain depended upon children’s ability to respond to bids for joint attention, F (3, 25) = 11.9, p < .001, R2 = .59. The findings suggest that autistic children with more consistent responses to bids for joint attention gain more language as a result of interventions than others with less consistent responses.
S2.4.5
COMPARISON OF INDIVIDUALS WITH ASPERGER SYNDROME AND AUTISM ON MEASURES SENSITIVE TO THE ABILITY TO SHIFT ATTENTION E. Hanson,* A. Lincoln Children’s Hospital, Harvard Medical School, 300 Longwood Avenue, Boston, MA 02115
There have been a number of studies showing the difficulty that individuals with Autism (AD) have in shifting attention rapidly, especially when required to do so in a very short (less than 2.5 second) time frame. (Akshoomoff & Courchesne, 1994; Casey, Gordon, Mannheim & Rumsey, 1993; Townsend & Courchesne, 1994). This method has been shown to be sensitive to cerebellar abnormalities. The primary goal of this study was to replicate these earlier findings and to evaluate whether individuals with Asperger Syndrome (AS) also showed impairment in their ability to rapidly shift attention. Findings from the present study supported the replication hypothesis that AD individuals would have a lower percent accuracy in the shifting conditions at the fastest time interval than normal controls (NC). AD individuals performed similarly to the other groups in longer time bins. Their sustained attention appeared intact as did their ability to shift attention given more time. The study did not support the hypothesis that AS participants would perform differently from AD and NC individuals in the fastest time bins and shifting attention tasks did not appear to be sensitive to differences between these groups. However, there were a number of interesting trends in the data. In particular, AS individuals tended to perform better on auditory shifting tasks in the fastest time bin but not in the visual shifting modality. Other trends were for AS individuals to have more difficulty with sustained attention vs. AD individuals who had difficulty with shifting attention.
S2.4.6
TOWARDS A MODEL OF SHARED ATTENTION IN INFANCY: TYPICAL DEVELOPMENT AND AUTISM G. O. Deák, J. Triesch, J. Movellan, I. Fasel, Dept. of Cognitive Science & INC, UCSD, La Jolla, CA 92093-0515
A deficit in attention sharing is among the first behavioral signs of infantile autism. To understand this we need a plausible model of the emergence of shared attention during normal development. We propose that gaze following, an early form of attention sharing, emerges from natural social interactions. Through generic learning mechanisms, and perceptual and motivational biases, infants learn that the parent’s gaze direction predicts the location of rewarding stimuli. We are combining behavioral evidence and computational models to test the plausibility of this proposal. In an observational study of healthy 3- to 12-month-old infants, we are deriving patterns of parents’ and infants’ attention shifting during social interactions. The observed patterns of parental behavior are used as input to train a computational model. This model is in the form of a computer-simulated virtual “infant” endowed with a typical 3-month-old’s perceptual, motivational, and learning capacities. Using this model we can determine under what conditions gaze following is learned, given a pattern of realistic social input. By perturbing parameters of learning and perceptual bias in the model, we can describe the conditions under which gaze following fails to emerge. This will offer clues to possible learning and attentional underpinnings of autism, which can be tested in behavioral experiments.