This study will explore a possible etiological mechanism for regression among children with autism spectrum disorders (ASD). By definition, children with ASD present with aberrant development in social skills, language, and restricted or repetitive behaviors/interests within the first 3 years of life. However, the trajectory of ASD is ot uniform across affected children. Nearly one third of children with ASD experience developmental regression, or a loss of previously acquired skills, most often in language and social skills, between the ages of 1 and 3. While some regain the skills lost, those who regress have poorer cognitive and adaptive-functioning outcomes by age 8. No one knows what causes regression. In the search for an ASD etiological mechanism, researchers have noted significantly reduced levels of carnitine among those with ASD. Mutations in the genes TMLHE and SLC22A5 lead to neuronal and systemic carnitine deficiency, respectively. New data from the Beaudet lab suggest that mutations in (a) TMLHE are present in at least 1/200 ASD cases and (b) SLC22A exist in 1/35 ASD cases with regression. We believe that children with ASD who have TMLHE and SLC22A5 mutations could experience regression because of reduced capacity for carnitine biosynthesis. This study will use the matched case-control design to investigate whether TMLHE and SLC22A5 mutations are more common among children with ASD and regression (cases) than those without regression (controls).
Specific aims are to (1) perform genotyping for the TMLHE and SLC22A5 genes on existing samples from children with ASD;(2) ascertain regression status for genotyped children;and (3) compare the frequencies of mutations in the TMLHE and SLC22A5 genes between cases and controls. We hypothesize that children with ASD and regression will have greater odds of having mutations in TMLHE and SLC22A5 than those with ASD who have not regressed. Participants will be 1200 children from the Simons Simplex Collection, a repository of genetic and phenotypic data from children with ASD. Regression data will be ascertained from the skill-loss items of the Autism Diagnostic Interview-Revised. Genotyping involves sequencing all exons and determining copy number for exons in both genes. Frequencies of mutations for TMLHE and SLC22A5 will be compared between groups (regressed vs. nonregressed) using conditional logistic regression statistics. Findings could lead to methods for predicting and preventing regression among children with ASD through dietary programs, subsequently enhancing their cognitive and adaptive-functioning outcomes.
Developmental regression, or a loss of previously acquired skills, occurs in about one third of children with autism spectrum disorders (ASD), most often between the ages of 1 and 3, and recent studies demonstrate that those with ASD who regress have poorer cognitive and adaptive-functioning outcomes by middle childhood than children with ASD who do not regress. Understanding genetic factors that are possibly associated with developmental regression could lead to methods for predicting and preventing regression, as well as targeted programs that mitigate the deleterious effects of regression, subsequently enhancing these children's long-term outcomes.