The Neural Basis for Atypical Pupillary Light Response in Autism Spectrum Disorder

Background:  Pupillary light reflex (PLR) refers to the involuntary pupillary restriction that is induced by a luminance change. Recent studies (e.g., Daluwatte et al., 2013; Fan et al., 2009) have documented atypical PLR (i.e., longer PLR latency, smaller constriction amplitude, and lower constriction velocity) in individuals with Autism Spectrum Disorder (ASD).  The primary neurological pathway subserving the PLR response is well-established and comprises the retina, pretectal nucleus, Edinger-Westphal nucleus, and ciliary ganglion (Lowenstein & Loewenfeld, 1950).  There is also evidence, however, of tertiary cortical and cerebellar contributions to PLR (e.g., Tsukahara et al., 1973). Within this context, the neural locus of ASD-related abnormalities in PLR remains unclear.

Objectives: Functional magnetic resonance imaging (fMRI) was used to examine the neural disruption(s) that contribute to atypical PLR in ASD.

Methods: A sample of 25 individuals with ASD (mean age = 16.0 years) and a demographically-matched comparison group of 19 neurologically intact individuals without ASD (mean age = 16.0 years) participated. Scans were obtained on a 3T Siemens Trio scanner with a standard 8-channel head coil. Stimuli were displayed using an LCD projector, and pupillary responses were recorded using a MRI-compatible ASL long range optic eye tracking system.  Participants performed a passive viewing task in which they were shown a series of red-filtered, emotionally-neutral images (e.g., landscapes) that changed every 5 s to maintain the interest of the participant.  Every 20 s, the participant was presented a green-filtered light stimulus superimposed over the current image for 100 ms.  The light stimulus was designed to induce PLR.  For each participant, PLR and neural responses were recorded for a total of 96 light stimulus trials.  Trials were presented over the course of 8 functional MRI runs, each of which lasted approximately 4 1/2 minutes.

Results: As anticipated, both groups showed robust PLR-related activation in primary visual sensory areas including lateral geniculate nucleus [F(1,40) = 16.3, p < .0005] and striate cortex [F(1,40) = 17.8, p < .0005].  PLR-related activation was also observed in association areas including superior parietal cortex [F(1,40) = 17.4, p < .0005] and right lateral prefrontal cortex [F(1,40) = 9.0, p < .05]. Most importantly, group differences in PLR-related activation were evident in the cerebellum as well as anterior insula and superior frontal gyrus, [F(1,40) > 20, p< .00005 in all instances].

Conclusions: These results are consistent with the hypothesis that prolonged PLR latency observed in individuals with ASD is associated with cerebellar and prefrontal dysfunction.