The Role of FMR1 in the Language Profile Associated with the Broad Autism Phenotype

Background: Fragile X syndrome (FXS) is the most common known cause of autism. The gene causing FXS (FMR1) is associated with significant genetic risk for autism, with 21-50% of individuals with FXS meeting criteria for autism (Moss & Howlin, 2009). While the genetic basis of autism is well established, detection of susceptibility loci has proven challenging. Studies of autism and the broad autism phenotype in genetically defined syndromes such as FXS provide the rare opportunity to examine potential gene-behavior associations in an etiologically homogeneous condition.
This study investigated the role FMR1 in autism symptomatology through the study of 1^st degree relatives who are at increased genetic liability (and in the case of FXS, who are FMR1 premutation carriers). Specifically, we examined particular language phenotypes that may index genetic liability to autism, and which could be linked to FMR1. Language impairment is core to both autism and FXS, and is thought to be genetically mediated. Specifically, phonological processing and verbal fluency have been documented as weaknesses associated with autism. Phonological processing and verbal fluency skills seen in FMR1 premutation carriers and mothers of children with autism may constitute genetically meaningful features related to autism, and may therefore help to illuminate a possible role of FMR1 in the presentation of autism and the broad autism phenotype.
Objectives: This study characterized the language domains of phonological processing and verbal fluency in mothers of individuals with fragile X syndrome (who are FMR1 premutation carriers) in comparison to mothers of individuals with autism (who are at increased genetic liability to autism) and controls, in order to explore overlapping, potentially genetically-linked features associated with autism.

Methods: The rapid automatic naming (RAN) and non-word repetition (NWR) subtests of the Comprehensive Test of Phonological Processing (Wagner, Torgesen & Rashotte, 1999) were administered to mothers of individuals with FXS (n=52), autism (n=24), and typically-developing children (n=15) to assess verbal fluency and phonological working memory. RAN was scored live for completion time and number of errors. NWR was scored from video by a trained, blinded rater for total number of errors. Ten percent of NWR samples were randomly selected and scored for intra-rater reliability (95% agreement).

Results: Analyses indicated significantly longer RAN completion times for autism and FXS mothers in comparison to controls (p=.032, p=.005, respectively). No significant differences were detected between the FXS and autism parent groups (p= .900). No significant differences were found in RAN total number of errors (F(2,88)=.080, p= .923) or NWR total number of errors (F(2,57)=.698, p=.502).

Conclusions: Both mothers of children with FXS and mothers of individuals with autism presented with overlapping profiles in verbal fluency skills, relative to controls. Phonological processing skills were not found to be impaired in either group. These findings suggest that verbal fluency skills may be a component feature of the broad autism phenotype and that FMR1 could play a role in this phenotype associated with genetic liability to autism.