More than Meets the Eye

A team of researchers from Yale set out to investigate eye contact and face-to-face behaviors in autism spectrum disorder (ASD) – a complex, neurodevelopmental condition with distinct behavioral, communicative, and social responses. With functional near-infrared spectroscopy (fNIRS), they found that the dorsal parietal regions of the brain are hypoactive during live eye-to-eye contact in participants with ASD (1) – establishing a focus area for diagnostic biomarkers. 

Lead authors Joy Hirsch, Elizabeth Mears and House Jameson Professor of Psychiatry, and James McPartland, Harris Professor of Child Psychiatry and Psychology, talk us through the main findings, challenges faced, and future directions of their research.

Why did you think this was an important area of research?
 

According to the Centers for Disease Control and Prevention, the incidence of autism is as high as 1 in 44 in the US (2). Investigations into the underlying neurobiology of ASD are conventionally performed by single, non-interactive participants. Therefore, the biological underpinnings of behavioral differences during live and interactive social processes remain under investigation. To bridge this knowledge gap, we applied recent developments in fNIRS to image two-people simultaneously during live interactions. These novel neuroimaging applications can investigate neural substrates associated with atypical social interactions.

Could you talk us through the analytical method that you used?
 

Analytical methods were specifically developed for the two-person neuroimaging paradigm – with a hemodynamic signal that represents task related neural activity. In fNIRS, that signal is detected by optical methods that identify task-related variations in oxyhemoglobin and deoxyhemoglobin. Brain signals were compared for neurotypical and ASD adults in a real-person interaction condition, where partners viewed each other’s faces directly, while sitting across a table from one another. Findings were compared with a condition where participants and their lab partners gazed at a comparable dynamic face video.

The contrast between these two conditions and direct comparisons between ASD and TD groups tested the hypothesis that neural processes responsive to real eye-to-eye contact are altered in ASD relative to neurotypical participants. Eye tracking and pupillometry were used to test the related hypothesis that gaze characteristics and pupil size variation altered during face processing between the two conditions. The social performance for ASD participants was also assessed by clinical interview, including administration of the Autism Diagnostic Observation Schedule. We employed multimodal techniques to relate the neural, behavioral, and clinical scores, and determine how the social metrics of behavioral function related to eye contact were linked to the underlying neurophysiology of the brain.

What were your main findings? 
 

During live eye contact, participants with ASD had significantly reduced activity in the dorsal parietal cortex region of the brain compared with those without ASD. Furthermore, the eye gaze patterns of ASD participants were more variable, and pupil diameters were increased in the presence of a live human face relative to a comparable simulated face video. Live faces in ASD are processed by ventral stream systems – including the lateral occipital cortex and temporal systems – whereas these same live faces are processed by dorsal parietal systems in neurotypical brains. 

Findings also show that, in ASD, the dorsal parietal brain region is hypoactive during social encounters. In the case of neurotypical individuals, this region is known to be associated with attention and eye-movements directed toward salient stimuli. Based on these results, an emerging theoretical framework relates social symptom severity in ASD to dorsal parietal processes and suggests a possible biomarker for ASD. 

Additionally, in neurotypical participants, the neural activity in these regions was synchronously coupled between participants during real eye-to-eye contact. However, this increase in person-to-person neural coupling between partners including an ASD individual was not observed – suggesting reduced sharing of sensory information between partners during eye contact when one of the partners is autistic. 

What were the advantages of using a live, two-person methodology?
 

The primary advantage to using a live, two-person paradigm is the match between the research question and the approach. A primary question in ASD is: What are the neural correlates of the specific social difficulties that are typical in ASD? Since live and interactive behaviors are diagnostic in ASD, it is critical to investigate them directly. Here, we have the advantage of using investigative conditions that represent the specific questions of the neurobiology that underlie live, face-to-face interactions.

What are the main implications of these findings?
 

These findings advance the theoretical understanding of social difficulties in ASD by providing a framework based on face-processing mechanisms. Live conditions of face processing in ASD are restricted to ventral stream systems, which are known for static object processing. This finding constrains models of social disability. These findings may also guide evidence-based approaches to treatment by focusing on possible plasticity of hypoactive dorsal parietal systems.

What are your future hopes for this research?
 

This investigation provides a “proof-of-principle” for live, two-person studies in ASD that focus on the social difficulties experienced in real life. In the future, we hope that these interactive studies will be expanded to provide neurobiological insight to other questions; for example, how are live facial expressions processed in ASD? Is emotion conveyed by facial expressions in ASD? If so, how do the neurobiological underpinnings compare with typical individuals? 

There is a fundamental need to identify an objective biomarker for the treatment of ASD. Our study suggests that the dorsal parietal regions of the brain are hypoactive during natural eye-to-eye contact; perhaps future studies could use this finding to develop biomarkers. Finally, the development of treatments for social difficulties in ASD is also urgent. We suggest that such research and findings could provide evidence-based neurobiological support to evaluate treatment progress.