Towards Identifying Phenotypic Subtypes In Autism: Fragile X Syndrome, A Disorder of Lower-Order Repetitive Behaviors

Background: Autism is well known to be an etiologically and phenotypically heterogeneous, behaviorally-defined syndrome. Genetics research has identified numerous distinct genetic markers associated with the diagnosis of autism and has come to refer to this group as constituting 'the autisms'. However, while structured, diagnostic instruments often lump these individuals into the same phenotypic category, there is good reason to expect that fine grained analysis will reveal a pattern of phenotypic differences that correspond to these distinct, etiologically-defined groups.

Objectives: In this study we undertake a more detailed characterization of restricted, repetitive behavior (RRB) in a cohort of 4 year old males with Fragile X syndrome and contrast them with an age and sex matched group of individuals with idiopathic autism (iAutism). Factor analyses have suggested a two factor structure underlying RRBs in idiopathic autism, including both so-called "lower" and "higher-order" RRBs. The aim of this study was to discern whether there exists a unique pattern of RRBs that characterize Fragile X individuals and distinguish autistic individuals with Fragile X Sydrome from those with iAutism.

Methods: Participants included 87 male children with a mean age of 4.3 years (SD = 1). Groups included autism (n = 39) and FXS (n = 48). A subset of the FXS group was comorbid for autism (FXS/autism) (n = 14). Diagnostic status was confirmed by joint ADI-R and ADOS assessments. Repetitive behavior was assessed using the Repetitive Behavior Scales-Revised (RBS-R; Bodfish, Symons, Parker, & Lewis, 2000). Developmental IQ (Mullen) and adaptive behavior (Vineland-II) measures were collected for all participants. 

Results: Rates of higher-order RRB were significantly higher for those with iAutism compared to FXS (p = .004), while lower-order RRB was not significantly different between groups. Higher-order factors of compulsivity (p < .001) and ritual behavior (p = .001) in particular strongly differentiated FXS from iAutism. Sub-group analyses (iAutism, FXS/no autism, and FXS/autism) indicated that groups differed by higher-order, but not lower-order RRB. Post-hoc analyses indicated that higher-order RBs (with the exception of sameness) distinguished iAutism from FXS/no autism, but did not distinguish FXS/autism from either iAutism or FXS/no autism. Correlational analyses indicated that stereotypy was highly related to daily living skills (r = -.42) and developmental IQ (r = -.5) for the autism group, but not for those with FXS, FX + autism or FX/ no autism. 

Conclusions: These data demonstrate that the profile of RRBs in four year old individuals with FX can be distinguished from that seen in individuals with iAutism by the predominance of lower order but not higher-order repetitive behaviors.  Converging evidence supporting phenotypic distinction comes from additional data showing differential correlations between RRBs and both daily living skills and developmental IQ in iAutistic individuals and autistic individuals with FX.  These data suggest that more refined behavioral analyses of 'the autisms' (i.e., individuals meeting diagnostic criteria for autistic disorder presumably resulting from distinct etiologies) will reveal unique behavioral profiles and provide increasingly fruitful targets for studies aiming to discern the biological factors underlying these overlapping but phenotypically distinct conditions.