16898
Delineating the Nature, Severity and Frequency of Face Processing Abnormalities in Autism Spectrum Disorders
Given the pivotal role of face processing in modulating social-communication, face processing deficits in ASD are increasingly viewed as potential “intermediate phenotypes” or “cognitive risk factors”. However, previous studies have reported mixed results. From a treatment/ intervention perspective, a better understanding of the nature, frequency and severity of deficits of face processing abnormalities is still needed. Moreover, identification of underpinning factors is vital in order to inform interventions that go beyond the effort to increase performance levels and will instead target abnormalities in the underlying processes themselves.
Objectives:
(1) To systematically test whether people with ASD have deficits in different stages and aspects of face processing, including face detection, face perception, face memory, gender recognition, basic and complex emotion recognition, as well as object recognition.
(2) To investigate whether face memory could be improved by altering (or “normalising”) the way people with ASD encode faces. Typically developing adults readily make social judgements when viewing novel faces (“is he nice or nasty?”). We hypothesised that in ASD this spontaneous tendency may be reduced and contributes to face memory deficits. Therefore, the explicit instruction to rate the trustworthiness of novel faces was predicted to increase face memory specifically in the ASD group.
Methods:
40 adults with ASD and 20 typically developing adults matched on sex, age and IQ completed a battery of computerised face processing tests. The battery comprised tests of face detection, the Cambridge Face Perception Task (CFPT), The Cambridge Face Memory Task (CFMT), a gender morph task, the Fleeting Films task, perceptual tests of happy/ angry emotion expressions, and a simple object recognition task. A modified version of the CFMT was created in which participants were asked to rate the trustworthiness of novel male faces during the encoding stage.
Results:
The ASD group showed significant impairments across all three stages of face processing (face detection p=.012, face perception, p=.008, face memory p=.0001), as well as in recognizing complex “fleeting” emotions from faces (p <.000001). Impairments were also found in recognizing two of three object categories but not in basic emotion perception (angry and happy faces) or gender matching ( all p >.08) .
Analyses of individual scores revealed that deficits were most frequent and profound on the fleeting films task with 89% of people with ASD performing below 2 SDs of the TD control group mean. Furthermore, severity of deficits were dimensionally correlated with autistic symptoms (r = -.44, p=.003). 36.8% of adults with ASD also showed strong deficits in face memory ( < 2SDs) and 68% had moderately-strong deficits (between 1-2 SDs). In addition, the trustworthiness-rating instruction yielded significantly improved face memory in both the ASD and TD groups.
Conclusions:
Using sensitive tests, people with ASD showed significant deficits across all stages of face processing and profound impairments in recognizing fleeting emotions. While improvements in face memory after the trustworthiness manipulation were not specific to the ASD group, our findings indicate the potential usefulness of interventions that emphasise social judgments during face encoding.