We read with interest the letter of Fitzgerald and Bellgrove outlining their views on the possibility of an overlap between Alexithymia and Asperger syndrome (Fitzgerald & Bellgrove, 2006). It is refreshing to see that others are noting a similarity between certain aspects of these two conditions. However, a number of relevant studies were not mentioned. Consequently we feel that it is important to provide information concerning direct studies of this topic, as well as of recent neuroimaging studies of alexithymia, in order to provide a broader, and more direct base of evidence with which to inform these discussions.

We have reported data from two studies investigating the presence of alexithymia-type symptoms in a group of high-functioning adults with ASD, many of whom had a specific diagnosis of Asperger syndrome. In the first of these studies, we assessed levels of alexithymia using the TAS-20 (Bagby, Parker, & Taylor, 1994; Bagby, Taylor, & Parker, 1994) in a group of high-functioning adults with ASD. We reported the responses of 27 ASD adults in comparison to 35 controls (matched for age and gender) and 49 relatives of people with ASD on this self-report measure (Hill, Berthoz, & Frith, 2004). As far as we are aware, this was the first study to address directly the issue of the existence of such aspects of emotion regulation difficulties in ASD. In addition to showing that adults with high-functioning forms of ASD are able to report their own emotional processes, we also showed that these individuals had significantly higher alexithymia scores than the control group (ASD mean 60.44, SD 10.84; control mean 42.51, SD 9.09). Furthermore, the individuals with ASD had significantly higher alexithymia scores than the group of relatives (who did not differ from the control group; relatives mean 42.4, SD 10.07). When placing participants into non-impaired, slightly impaired and severely impaired alexithymia categories using previously established cut-off scores (Bagby et al., 1994), adults with ASD fell overwhelmingly into the slightly and severely impaired categories, while few of the controls or relatives fell into the slightly impaired group, and none of these groups were categorised as severely impaired. The hypothesis that the results may be explained by between-group differences in depression scores was ruled out (Hill et al., 2004).