Abnormalities In Neuronal Gamma-Band Synchronization and M100 Latency Delays In First-Degree Relatives of Children with Autism Spectrum Disorders (ASD)

Background: Current diagnoses and treatment of Autism Spectrum Disorders (ASD) are based entirely on behavioral observations, so identification of biomarkers and endophenotypes in ASD would provide substantial benefits to diagnostic subtyping, preclinical development of therapeutics, treatment monitoring and genetics.  An endophenotype would be seen at higher rates in those with ASD than unaffected family members or the general population, but also more frequently in unaffected family members than in the general population.  Synchronized neuronal activity in the gamma band (30-80 Hz) has been found to be abnormal in both children and adults with ASD during presentation of auditory tone stimuli (Wilson et al., 2007; Rojas et al., 2008).  These abnormalities have also been observed in unaffected parents of children with ASD (pASD), suggesting the gamma-band response may be a useful ASD endophenotype.  However, the gamma-band response to language in ASD has not been investigated as an endophenotype marker, and would be of specific interest for this population given the language and communication impairments commonly seen in ASD. 

Objectives: The current study aims to investigate differences in gamma-band responses to language between a group of adults who have a child with ASD and a healthy adult control group, using magnetoencephalography (MEG).  In addition, since children with ASD have shown latency delays in the M100 response (an early evoked field component of the MEG waveform) (Roberts et al., 2010), M100 latency will be assessed.

Methods: Parents of a child with ASD (N = 17) and a group of adult control participants (N = 23) listened to 90 abstract English nouns presented binaurally during MEG recording.  Dipole fits were used to construct virtual electrodes from which to conduct time-frequency transformation, and averaged evoked (i.e., phase-locked) waveform data between 30-100 ms post-stimulus was analyzed to capture evoked transient gamma-band activity.  In addition, phase-locking factor (PLF), a direct measure of phase-locking to stimulus, was calculated.  M100 peak latency was measured from the time of stimulus onset.  Independent samples t-tests were used to assess group differences in mean transient evoked power, PLF, and M100 latency. 

Results: The pASD group showed an increased mean PLF (p = .012) and increased M100 peak latency (p = .005) compared to controls.  No group differences were found for transient evoked gamma-band activity, p > .05.

Conclusions: Previous studies showed decreased PLF in pASD during tone presentation (Rojas et al., 2008), but it seems this gamma-band abnormality is distinct for language stimuli, as the current study found pASD to have an increased PLF compared to controls.  Additionally, pASD showed an M100 latency delay compared to controls, which has been found previously in children with ASD (Roberts et al., 2010), but has not previously been investigated in first-degree relatives of individuals with ASD.  These results suggest M100 latency and gamma-band responses to language stimuli are heritable, and are potential ASD endophenotypes.