Brief communicationLack of developmental improvement on a face memory task during adolescence in autism
Introduction
Autism Spectrum Disorders (ASD) affect not just the starting point of development but also how development unfolds (Minshew, Sung, Jones, & Furman, 2004). Such differences in the course of development may be more striking for skills that have a protracted development through adolescence. The current paper examines the developmental trajectory of one such skill, face memory, in individuals with and without ASD. In this paper, face memory refers to recognizing unfamiliar faces. While this task is complex and engages multiple processes (attention, perception, memory, and matching), it also has high ecological validity, approximating how we interact with faces in the environment. Face memory has been reported to be impaired in high-functioning children and adults with ASD (Faja et al., 2009, Scherf et al., 2008a, Wallace et al., 2008, Wolf et al., 2008), but this deficit is not always evident (Sterling et al., 2008). It is also controversial whether these memory deficits are specific to faces. Impairments in face memory have been reported along with typical, or even better than typical, performance with cars and houses (Wallace et al., 2008, Wolf et al., 2008). However, similar impairments have also been reported with non-faces (Behrmann et al., 2006, Scherf et al., 2008a).
Examining the development of face memory provides insight into the neural underpinnings of ASD, since research now indicates that the brain regions related to face representation mature late and that this maturation is impacted by ASD. Face-related activation in the fusiform face area (FFA) increases from adolescence to adulthood typically, as face memory improves (Golarai, Liberman, Yoon, & Grill-Spector, 2010). In contrast activation specific to faces in or near FFA appears both decreased and more posterior in young adolescents with ASD compared to matched controls (Scherf, Luna, Minshew, & Behrmann, 2010). In terms of brain structure Raznahan et al. (2009) report that cortical thinning in temporal regions occurs typically during adolescence but does not happen during adolescence in ASD. This differential development results in increased cortical thickness in adults with ASD compared to controls, potentially affecting face recognition and other high-level visual skills. Similarly, disruption in the development of white matter in the temporal lobe is evident in ASD. A Diffusion Tensor Imaging (DTI) study suggests that white matter tracts in temporal regions continue to mature into adolescence but that these changes do not occur in ASD (Lee et al., 2007). Reduced connectivity between temporal and frontal regions, reported to be late developing typically (on a phonological task; Bitan, Cheon, Lu, Burman, & Booth, 2009), also affects face memory in adults with ASD (Koshino et al., 2008).
Thus, limitations in face memory in ASD may not be fully apparent until maturity is reached. To examine this possibility, we used the Cambridge Face Memory Test (CFMT) developed to identify relatively subtle deficits in adults with prosopagnosia (Duchaine & Nakayama, 2006). Individuals with ASD, 9–29 years old, matched individually to typically developing individuals on age and IQ, were tested on the CFMT. The CFMT limits the use of non-facial information and feature matching strategies, which may decrease the sensitivity of other standardized face processing tasks. Importantly, the CFMT provides increasing levels of difficulty to control for ceiling and floor effects, which often complicate developmental studies (Crookes & McKone, 2009). The CFMT also exhibits a robust inversion effect (i.e., poorer performance when faces are upside-down), indicating that it tests a skill related to upright faces. One proposal of such a skill is the use of holistic/configural information (perceptual integration of elements into a whole and/or spacing between features), aspects of which may develop late (Diamond and Carey, 1986, Mondloch et al., 2002) and be impaired in ASD (Behrmann et al., 2006).
Given the accumulating evidence of atypical maturation of brain processes in ASD, we hypothesized that, while performance on the CFMT would improve with age typically, it would not in ASD. While the use of the CFMT does not allow us to disentangle exactly what is developing late, or whether it is specific to face memory, it seemed likely to be sensitive to late development. Indeed, the typical developmental literature indicates a lack of consensus on a particular element of face memory that develops late. While more traditional views suggest that late face development reflects face-specific perceptual skills and/or expertise coming online (Diamond and Carey, 1986, Mondloch et al., 2002), recent evidence indicates that development of face recognition reflects improvement in general cognitive and/or visual processes, which may impact a range of tasks (Crookes & McKone, 2009). In either case, differences in the development of these skills during adolescence provide insight into the limitations in visual processing and the dysmaturation of brain processes in ASD.
Section snippets
Participants
Participants included 34 individuals with autism spectrum disorders (ASD) (31 males) from 9 to 29 years old and 34 typically developing control subjects (30 males) matched individually to the participants with ASD on age (±1 year in children, 2 years in adults) and IQ on the Wechsler Abbreviated Scale of Intelligence (±15 points) (Table 1). These included children (9–12, N = 8/group), adolescents (13–17, N = 12/group), and adults (18–29, N = 14/group). Individuals with ASD were recruited through the
Results
Percent correct (Fig. 3, Fig. 4) was examined using repeated measures ANOVA with condition (Conditions 1–3) as a within subject factor, group (TD, ASD) and age (children, adolescents, adults) as between subjects factors, and spatial span as a covariate.1 This ANOVA revealed main effects of group (F(1,61) = 14.68, p < .001, partial eta2 = .19); condition (F(2,122) =
Discussion
The CMFT task was used to characterize the development of face memory in individuals with and without ASD. Results indicate that performance on this face memory task continues to improve into adulthood typically but not in individuals with ASD similar in age and IQ. This pattern of results lead to impairments in adulthood. The level of deficit documented in the adults with ASD did not differ from that previously reported in adults with congenital prosopagnosia, while the typically developing
Acknowledgements
This work was done at the University of Pittsburgh and supported by Autism Speaks Grant 04593 to BL, NIH HD055748 to NM, NIMH 5 R01 MH067924 to BL, and NIMH KO1 081191 to KOH. We are very grateful to the participants and their families, Catherine Wright, the staff at the ACE project, and the Autism Speaks organization. We also would like to thank Brad Duchaine, Ken Nakayama and their collaborators for making the CFMT available to the public.
References (25)
- et al.
Configural processing in autism and its relationship to face processing
Neuropsychologia
(2006) - et al.
Early maturity of face recognition: No childhood development of holistic processing, novel face encoding, or face-space
Cognition
(2009) - et al.
The Cambridge Face Memory Test: Results for neurologically intact individuals and an investigation of its validity using inverted face stimuli and prosopagnosic participants
Neuropsychologia
(2006) - et al.
Diffusion tensor imaging of white matter in the superior temporal gyrus and temporal stem in autism
Neuroscience Letters
(2007) - et al.
Maturation of executive function in autism
Biological Psychiatry
(2007) - et al.
When is the brain enlarged in autism? A meta-analysis of all brain size reports
Biological Psychiatry
(2005) - et al.
Developmental increase in top-down and bottom-up processing in a phonological task: An effect connectivity, fMRI study
Journal of Cognitive Neuroscience
(2009) - et al.
Why faces are and are not special: An effect of expertise
Journal of Experimental Psychology: General
(1986) - et al.
Brief report: Face configuration accuracy and processing speed among adults with high-functioning autism spectrum disorders
Journal of Autism and Developmental Disorders
(2009) - et al.
Differential development of the ventral visual cortex extends through adolescence
Frontiers in Human Neuroscience
(2010)
fMRI investigation of working memory for faces in autism: Visual coding and underconnectivity with frontal areas
Cerebral Cortex
Autism diagnostic interview-revised: A revised version of a diagnostic interview for caregivers of individuals with possible pervasive developmental disorders
Journal of Autism and Developmental Disorders
Cited by (77)
Replicable Patterns of Memory Impairments in Children With Autism and Their Links to Hyperconnected Brain Circuits
2023, Biological Psychiatry: Cognitive Neuroscience and NeuroimagingProsopagnosia and disorders of face processing
2021, Handbook of Clinical NeurologyRepresentational similarity analysis reveals atypical age-related changes in brain regions supporting face and car recognition in autism
2020, NeuroImageCitation Excerpt :The L and R hemispheres were analyzed separately, on the basis of theoretical and empirical work indicating that activation to faces in the regions of interest differs across hemispheres (Rossion et al., 2000). We examined the behavioral performance in the scanner, and the pattern of performance was similar to our previous work but less robust (O’Hearn et al., 2010, Fedor et al., 2018). TD participants still tended to improve from adolescent to adulthood, while participants with ASD did not.
The development of face expertise: Evidence for a qualitative change in processing
2018, Cognitive Development