Pupillary Light Reflex Constriction Amplitude's Correlation to Clinical Symptoms

Background:  Pupillary light reflex (PLR) provides a non-invasive model system for study of the nervous system in ASD. Pupillary response to a light flash includes a latency period, then pupil constriction and recovery. Daluwatte et al, (2012) found PLR parameters discriminated ASD children from typical developing (TD) controls with 87.7% specificity and 76.4% sensitivity. Four PLR measurements (constriction amplitude (CA), latency, constriction time, redilation time) were calculated to quantify PLR. Though latency was the strongest ASD predictor, each parameter strengthened the association. To determine significance of each parameter and to assess their potential usefulness as ASD biomarkers, we have undertaken an analysis of associations between PLR parameters and ASD symptoms. Constriction amplitude (CA) assesses autonomic function, since the constriction sphincter is innervated by the parasympathetic system. A small but consistent literature finds that children with autism have lower parasympathetic and higher sympathetic activity. Moreover, children with ASD commonly present with GI, urinary, sensory, sleep disturbances which are to some degree under ANS control. We question whether variation in CA might inform us about systemic ANS dysfunction. 

Objectives: Investigate the association between CA and clinical ASD symptoms.

Methods: PLR was measured in 152 children with ASD (age 10.7±3.4 years, 135 males and 17 females) and 107 TD children (age 10.9 ± 2.9 years, 79 males and 28 females). PLR induced by a 100ms green light was measured in light adapted and dark adapted conditions using a two channel binocular apparatus. PLR measurements were calculated to quantify PLR. A parent questionnaire designed to evaluate areas of ANS participation, including GI, GU, fever response, sleep and hyper-sensitivity was completed. From this group, 53 ASD children who had also completed the Simons Simplex Study were selected for preliminary analysis.

 Results: CA was in autism range or below for 53%, in normal range for 34% and equivocal for 13%. Using CA as dependent variable, we found low CA correlated significantly with lower IQ (FSIQ & NVIQ; p= 0.02, VIQ: p=0.03). In addition, children with the lowest CA were more than twice as likely to have parents report improvement in core ASD symptoms with fever (40% vs19%). This confirmed findings from our initial PLR subjects (Fan et al., 2009). Systemic symptoms influenced by the ANS (GI, sensory, sweating, salivation, urination, swallowing, sleep disturbances) were not associated with CA. Though not reaching significance the low CA group showed a trend toward further neurologic dysfunction based on higher toe walking and dysmorphology.

Conclusions: PLR CA is a measure of ANS activity. Our data show CA correlates inversely with heart rate, indicating low pupillary CA could be an indicator of systemic ANS dysfunction in ASD. In this small sample none of the clinical symptoms suggestive of ANS dysfunction correlate with CA. We did find children with the smallest CA had significantly lower IQs suggesting variation in CA may be a marker for general neurologic disruption. Clarification of these results will depend on our ongoing analysis in the entire sample of 152 ASD children and 107 TD children.