Friday, May 21, 2010
Franklin Hall B Level 4 (Philadelphia Marriott Downtown)
Background: A significant body of evidence has accumulated in support of the cortical hypo-connectivity hypothesis of autism; the hypothesis states that individuals with autism spectrum disorders (ASD) have weaker than normal long-range cortical functional connectivity that may contribute to their cognitive abnormalities. Our own studies show that individuals with ASD make more errors than controls on an antisaccade task.
Objectives: Given compelling evidence of cortical hypo-connectivity in autism, we investigated long-range coherence in ASD during fixation and in preparation for prosaccades and antisaccades. We hypothesized that coherence in the network subserving volitional ocular motor control would be reduced in ASD. In particular, given the higher error rate in the ASD group for antisaccades, we investigated whether group differences in coherence were more pronounced during preparation for a more cognitively demanding task (an antisaccade) than for a simple task (fixation or a prepotent prosaccade), and if so, which cortical regions were most affected.
Methods: We studied 10 high functioning adults with ASD and 11 age and gender matched neurotypical controls ( NTC) using whole head Magnetoencephalography (MEG). We looked at three conditions – 1) fixation with no immediate associated task (‘fixation’), and fixation in preparation for 2) a saccade towards a suddenly appearing visual stimulus (‘prosaccade’) or 3) a saccade away from the stimulus (‘antisaccade’). For each subject and condition, we analyzed the MEG amplitude and coherence in source space (i.e., cortical space during the preparatory interval in between the instructional cue and stimulus appearance.
Results: In sensor space ASD participants showed significantly reduced coherence relative to NTC participants, most prominently in the alpha band, with the right hemisphere showing significantly greater reduction in coherence relative to the left hemisphere. Accordingly, in source space, we focused on alpha coherence and used the superior and inferior frontal eye field (FEF), the key cortical region for saccade generation, as the seed region. In the NTC group, the FEF region had significant coherence values primarily with the intraparietal sulcus (IPS), dorsolateral prefrontal cortex (DLPFC), and the dorsal anterior cingulated cortex, both ipsilaterally and contralaterally, depending on the region and the condition. In the ASD group, FEF showed significant coherence values with many but not all of the same regions, but those were often reduced in the alpha band relative to the NTC group. In particular, the left FEF to left IPS and left FEF to right FEF showed condition dependent group differences, with the strongest effects between the antisaccade and prosaccade conditions.
Conclusions: Our data indicates that when preparing for a saccade, participants in the NTC group show increased coherence in the alpha band between the FEF bilaterally and other cortical regions, relative to the ASD group. The extent of the group differences was task dependent. These findings likely correspond to the neural substrates of deficient inhibition of prepotent saccades observed in ASD.