Brain

Social and developmental disorders (SDD), which include the psychiatric diagnoses of autism and Asperger's syndrome and the neurological diagnosis of social-emotional processing disorder (SEPD), are characterized by impaired social function. Face perception per se and facial expression processing in particular are important for social interaction, and the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, Criteria for Autism and Asperger's Disorder emphasizes both facial expression processing deficits and social dysfunction as features of these disorders. It is therefore unsurprising that previous studies of individuals with SDD have focused on the examination of facial expression recognition deficits. Facial expression perception was impaired in autistic individuals, for example (e.g. Hobson et al., 1988). Furthermore, whilst many previous studies have highlighted the roles of specific neural regions, including the amygdala and fusiform gyrus, in normal facial expression and identity perception (e.g. Haxby et al., 2000; Phillips et al., 2003), neuroimaging studies in autistic individuals indicate impaired function in these regions (e.g. Critchley et al., 2000; Pierce et al., 2001; Hubl et al., 2003) and abnormal amygdalar volumes (e.g. Howard et al., 2000; Pierce et al., 2001; Salmond et al., 2003).

These data suggest dysfunction within neural systems underlying face and facial expression perception in SDD but there are several outstanding questions. First, the extent to which SDD is characterized predominantly by a facial expression processing deficit rather than by a generalized face processing deficit, that includes impaired face identity perception, remains unclear. Secondly, if there is a deficit in face identity processing, it is unclear whether this is predominantly a visual perceptual deficit or an impairment in associative processing, resulting in impaired face memory and imagery but intact face perception. Thirdly, it remains to be determined whether there is a unitary pattern of face processing deficits in SDD, or if this can be considered a heterogeneous syndrome characterized by different patterns of face processing impairments. Finally, do these face processing deficits occur as a primary deficit or are they secondary to the social dysfunction in individuals with SDD?

In an elegant study, Barton and colleagues (Barton et al., 2004) aimed to answer these questions by examining the nature of face identity processing abnormalities in 24 adults with SDD. These included Asperger's syndrome only (n = 2), SEPD only (n = 11), both Asperger's disorder and SEPD (n = 8), and high-functioning autism (n = 3). All individuals participated in five facial identity processing tasks. These included two standard neuropsychological tests of face identity recognition memory and matching, and three novel tests, including a test examining famous face identification and naming, a test of famous face imagery, in which individuals were asked to compare facial appearances of pairs of celebrities, and a test of internal facial configuration, measuring the ability to detect small alterations in facial feature position. In order to determine the nature of the neural deficits underlying face identity processing deficits in these individuals, the authors compared their performance with that of 12 prosopagnosic adults with well-described focal brain lesions.

A striking finding was that a subgroup of eight SDD individuals (SDD-1) had normal levels of performance on all five face processing tasks. The remaining 16 individuals (SDD-2) were impaired on all face processing tasks, although less impaired than the prosopagnosic individuals on the famous face identification task. SDD-2 individuals could be divided into three subgroups differing in the degrees of impairment on the face imagery and facial configuration tasks. Interestingly, SDD-2 individuals who were more impaired on the facial configuration task (SDD-A) had similar performance on the face imagery and configuration tasks to that of prosopagnosic individuals with occipitotemporal lesions, whilst SDD-2 individuals who were more impaired in face imagery resembled a prosopagnosic individual with bilateral anterior temporal lesions. There were no differences between SDD-1 and SDD-2 in frequency of the three different diagnoses.

The first conclusion from these findings is that facial identity processing impairments are a feature of SDD, and include both perceptual and associative deficits. Furthermore, there appear to be different patterns of facial identity processing impairments in SDD, indicating that SDD can be considered a heterogeneous syndrome characterized by differential patterns of impairment in neural systems underlying facial perception and memory. Importantly, however, the demonstration of a subgroup of individuals with SDD with intact facial identity processing suggests that face processing deficits in SDD are not necessarily a result of the social dysfunction demonstrated by individuals with these disorders: both SDD-1 and SDD-2 subgroups were socially impaired by definition, and also had low scores on a scale measuring perceived ability in social skills, whilst only the SDD-2 subgroup had impaired face processing. Rather, the findings suggest that the development of social dysfunction in SDD may occur secondarily to different types of face processing impairments and associated neural deficits in a subgroup of these individuals, whilst in other individuals the social dysfunction may occur secondarily to other primary perceptual deficits. Another possibility is that there is no causal relationship between facial processing impairments and social dysfunction in individuals with SDD. As the authors suggest, further study of the nature and development of social function in prosopagnosic individuals would help to clarify the nature of this relationship.

Neuroimaging findings of individuals with autism and Asperger's syndrome indicate dysfunction not only within neural systems important for face processing but also within regions underlying higher-order cognitive functions, including theory of mind, the ability to understand others' mental states (Castelli et al., 2002). Furthermore, given the emphasis in the diagnostic criteria for SDD on a combination of face expression processing deficits and social dysfunction, it is surprising that the nature of the relationship between these processes in normal development and in individuals with SDD has not received further attention. Barton and colleagues provide some of the first data suggesting a role, albeit not exclusive, for face identity processing impairments in the development of the social dysfunction in individuals with SDD, and will no doubt encourage further studies examining the relationship between the development of facial perception and that of social function in healthy and socially impaired groups.

• Brain Vol. 127 No. 8 © Guarantors of Brain 2004; all rights reserved