International Meeting for Autism Research (IMFAR)
May 7-8, 2004
Sacramento, California

Abstracts presented by the Center for Autism and Developmental Disabilities Epidemiology

*NOTE: Findings in poster presentations are based on preliminary data analyses.  Final results may differ after completion of analyses.

National data collected by the US Department of Education’s Office of Special Education Programs on children receiving special education services were analyzed by birth cohort for prevalences of autism and three other special education classifications (hearing impairment, speech/language impairment, and other health impairments).  Age-specific prevalences were calculated for these special education classifications for annual birth cohorts from 1975 to 1998 using denominators from US Census Bureau estimates.  For autism, there are clear birth cohort differences - prevalence is increasing with successive (younger) birth cohorts. This effect is greatest for cohorts born between 1987 and 1992. For cohorts born after 1992, the prevalence increases with each successive year but does not appear to be as great as in previous years, although there are fewer data points available within cohorts.  The curves for other health impairment are notable both because this is the special education typically including children with attention deficit hyperactivity disorder and because strong cohort differences are also present here. There are no birth cohort differences for speech language impairment or hearing impairment prevalence.  Cohort graphs reveal that autism prevalence is increasing with time, as evidenced by higher prevalence in younger birth cohorts. Whether the narrowing in vertical separation of the cohort curves marks a slowing autism prevalence increases remains to be seen.

View poster.

Increased incidence of maternal HLA-DR4 in single-birth, but not multiplex families with autism
L-C. Lee, A. Zachary, S. Leffell, C. Newschaffer, K. Matteson, J. Tyler and A. Zimmerman

Autoimmune disorders (ADs) are observed with increased frequency among parents of autistic individuals, particularly mothers. Since ADs are associated with specific alleles of the HLA system, we examined HLA in families to determine their incidence and parental inheritance. Two groups were studied: a geographically defined sample, including 18 families with one autistic child; and a genetically loaded sample from all areas of the USA, which included 33 families with multiple autistic children. The HLA-DR4 frequencies in these two groups were compared to 475 normal, unrelated Caucasians. Findings of low resolution HLA typing indicated that DR4 was not significantly higher in individuals with autism or in their fathers, compared to the normal controls. Mothers in the geographically defined sample had significantly higher frequency of DR4 than normal controls (Odds Ratio=3.96; 95% Confidence Interval= 1.39, 11.57); but in the genetically loaded sample no significant difference was found. High resolution typing suggested no sharing of alleles, and no significant change in the distribution of DR4 alleles among those with DR4. No distortion in the segregation of the maternal DR4-bearing haplotype was seen among the autistic children.  There was a normal distribution of HLA homozygotes and heterozygotes and no evidence of antigen sharing between parents. Increased HLA-DR4 in mothers, but not children or fathers, in the single-birth families is consistent with a hypothesis that prenatal maternal-fetal immune interaction can affect fetal brain development in autism.  Such immune interactions may involve HLA and related genes in both genetic and epigenetic mechanisms during pregnancy.

View poster.

Investigating Developmental DElays Study: comparison of SCQ and PDDST
C.J. Newschaffer, L-C. Lee, A.B. David and N.L. Lee

Objective: To compare the effectiveness of the Social Communication Questionnaire (SCQ) and the Pervasive Developmental Disorders Screening Test (PDDST) in identifying autism spectrum disorder (ASD) among children age 3-5 years receiving special education services.
Methods: A brief, self-administered questionnaire consisting of demographic questions and the SCQ and PDDST was mailed to all parents of children age 3-5 years receiving special education services through the Howard County School District in Maryland  (n=740). In these preliminary findings, parent’s report of the child having an autism special education classification determined case status. Clinical assessments using the ADOS and ADI-R will validate these administrative classifications and the results will be presented at the meeting.
Results: Thirty-five percent (n=258) returned completed surveys, and 15% of these (n=38) scored positive for ASD on either the SCQ (score >15) or PDDST (score >13). When compared to autism special education classification, the sensitivity and specificity of the SCQ was 55.2% and 90.8%, respectively, with a positive predictive value (PPV) of 43.2% and a negative predictive value (NPV) of 94.1%. The sensitivity and specificity of the PDDST were 20.7% and 98.7%, respectively (PPV 66.7% and NPV 90.8%).
Conclusion: Sensitivity of these two brief screening tools appears limited. While specificity is higher, PPV is still fairly low even in this target population with enriched autism prevalence.

View poster.