Pro- and Anti- Saccade Abnormalities in First-Degree Relatives of Individuals with Autism

Background: Saccades are rapid eye movements that shift the focus of gaze.  Cortico-cerebellar feedback loops modulate the accuracy of saccades and control for motor error.  Inhibitory control of reflexive saccades engages fronto-striatal pathways and can be assessed using anti-saccade tasks in which participants make saccades away from targets presented in their visual field.  Decreased accuracy of saccades and poor inhibitory control (i.e. increased rates of saccades made towards instead of away from targets) on anti-saccade tasks have been observed in individuals with autism.  These deficits and their underlying pathophysiology could serve as useful intermediate phenotypes, but saccade performance has not yet been systematically investigated in unaffected family members of individuals with autism. 

Objectives: To investigate sensorimotor and cognitive control of saccades in first-degree relatives of individuals with autism.

Methods:   Fifty-nine first-degree relatives (parents and siblings) of individuals with autism and 31 age- and IQ-matched healthy control individuals between 8-55 years of age performed visually guided pro-saccade and anti-saccade tasks. During the prosaccade task, participants made reflexive saccades to peripheral targets.  During the antisaccade task, participants were instructed to inhibit saccades towards novel peripheral targets and instead make a saccade to the mirror location of the target.  Each task included gap and overlap conditions.  ‘Gap’ trials were characterized by extinction of a cross-hair at central fixation 200 ms prior to presentation of a peripheral stimulus. ‘Overlap’ trials were characterized by extinction of a cross-hair at central fixation 200 ms after presentation of a peripheral stimulus.  The latency and accuracy of prosaccades were examined.  Anti-saccade error rates (the failure to inhibit prepotent responses) and latency of anti-saccades were examined. 

Results: Saccades of family members during the prosaccade task were less accurate and more variable in their accuracy than those of controls during gap trials only.  Family members also made more anti-saccade errors than controls in the gap condition.  No group differences in the latency or peak velocity of pro- or anti-saccades were observed in any task. 

Conclusions: First-degree relatives of individuals with autism show impairments in sensorimotor and inhibitory control of saccades that parallel those previously observed in probands.  The saccadic dysmetria and poor inhibitory control of reflexive saccades observed in family members implicate cerebellar circuitry involved in fine motor control and frontostriatal circuitry critical to the top-down control of motor behavior.  These findings in unaffected family members suggest that disturbances within these neural systems may be useful intermediate phenotypes for  research into the genetic bases of autism.