Fragile X, Intermediate, and Premutation Alleles in the Autism Genetic Resource Exchange (AGRE)

Background:  AGRE is an autism family registry and resource that includes primarily multiplex families, having two or more children affected by autism spectrum disorders. Family pedigree, phenotypic and genotypic data are available online, and genetic material is available.  The families were recruited by ads and contact with local support groups by Cure Autism Now and Autism Speaks. They represent a relatively unbiased sample. ~35% of the first 480 families were noted to have had some genetic testing prior to entering the registry. The individuals receive ADI-R and ADOS.

Objectives:   We have screened the FMR1 locus in one proband from each of 1742 families. 

Methods:  Standard Molecular Methods were employed

Results:  We found 8 families that had the fragile X mutation present. Among these, 6 were found among the first set of 480 families. The prevalence of fragile X among the ~312 AGRE families that had had no prior genetic screening was ~ 1.9%. An estimate of the IQ score of the autistic subjects was 80±35 with range 34-144, based on Raven and Stanford-Binet testing. Thus, the AGRE sample is likely to have a higher IQ distribution than typical for fragile X subjects (mean ~40±25). Previous prevalence studies of fragile X in autistic samples range from 0 to 16%; with a mean of ~4%; (Feinstein 98). Our 1.9% is similar to a report of 1.6% among 123 unrelated autistic individuals (Bailey 93), but lower than the 13% we found on an earlier multicenter study of 183 individuals (Brown 86).  A growing awareness of fragile X syndrome has decreased the probability of fragile X in these multiplex autism families due to screening and exclusion from AGRE. The observed frequency of 1.9% is lower than the expected 4%, perhaps due to higher IQs in AGRE subjects than is typical for fragile X.  This finding confirms an association of fragile X and autism.

Conclusions:  We tested to see if there is an association of autism and permutations or intermediate alleles. Among the 1535 male probands tested, there were 2 with premutation (59 & 64 CGGs) and 12 with intermediate (45-54 CGGs) alleles (46, 47, 47, 47, 48, 48, 48, 49, 50, 50, 51, 54) for an intermediate prevalence of 0.78%.  Among the 206 female probands tested there were 2 with premutations (55, 59) and 7 with intermediate alleles (45, 46, 48, 50, 52, 53, 54). Since females have two alleles, dividing by 2 gives an intermediate allele prevalence of 1.7% in female alleles or an overall intermediate allele prevalence of 0.98%.  Our published control value for 2500 X chromosomes was 1.7% (Brown 96), and our more recent control value based on carrier screening of 9064 X chromosomes was 1.15%. Thus, there was no excess of intermediate or premutation alleles among the AGRE registry autistic probands. This finding suggests autism is NOT associated with intermediate (45-54) or premutation (55-200) alleles.