Autism is not associated with the fragile X syndrome.

We provided a controlled test of the hypothesis that individuals with the Martin-Bell or Fragile X [fra(x)] syndrome are more autistic than mentally retarded control individuals. A sample of fra(x) individuals was obtained from the register of a clinical genetics unit and compared with mentally retarded control individuals selected from an assessment centre who were individually matched for age, sex, and IQ. A comparison of 45 pairs of fra(x) cases and control individuals on the DSM-III (R) criteria for autism and two standardized "instruments" for the assessment of autism (ABC and ADC) failed to find a higher prevalence of autism among fra(x) individuals. There was no statistically significant difference between fra(x) individuals and control individuals, and the mean differences on both the ABC and ADC scales were in the direction contrary to the hypothesis. An analysis of the study's statistical power suggested that it is unlikely the investigation failed to detect a large to medium difference between fra(x) individuals and control individuals in autistic behaviour. Subsidiary analyses of case-control differences showed that two autistic-like behaviours occurred at a higher rate among fra(x) individuals than other mentally retarded children, namely, gaze avoidance and hand flapping. These abnormalities may have misled clinicians into thinking that autism and fra(x) are associated.