History of Mental Illness Does Not Predict Gamma Band Deficits In First-Degree Relatives of Children with Autism

Background:  

Gamma band oscillatory activity, as measured using EEG and magnetoencephalography (MEG), has been associated with intrinsic GABA and glutamate activity in animal and computational models of the neocortex and is thought to play an important role in perceptual binding and central coherence. We have previously published data demonstrating that auditory steady-state response (ASSR) gamma activity is abnormal in people with autism. We have also recently established that transient gamma-band response deficits are seen in adult first-degree relatives, suggesting the utility of some of the findings as endophenotypes. As first-degree relatives of persons with ASD have an increased prevalence of various Axis I mental illnesses, this factor should be considered as an alternative explanation for the observed gamma-band deficits.

Objectives:  

The purpose of this study was to extend the ASSR gamma-band findings to 1^st degree relatives of persons with ASD and to examine the potential impact of history of mental illness on gamma-band deficits.

Methods:  

We examined MEG recorded ASSR in 21 parents of children with autism spectrum disorders (pASD) and 18 adult control subjects (CON). History of mental illness was assessed using the Structured Clinical Interview for DSM-IV (SCID); 8 persons in the HC group had a personal history of Axis I mental illness and 11 of the pASD group had such history (primarily mood and anxiety disorders). MEG data were acquired using a 248-channel neuromagnetometer system. Responses to repeated, 32, 40 and 48 Hz amplitude modulated white noise (100% depth, 500 ms duration, 50 dB HL) were examined in the time-frequency domain from source space projected auditory “virtual electrodes” for the left and right hemispheres. Measures of gamma-band phase-locking factor, normalized phase-locked power and normalized total power were derived for each subject and hemisphere.

Results:  

Evoked power was significantly higher at 40 and 48 Hz modulation rates than for the 32 Hz rate across all subjects. The pASD subjects had significantly lower phase-locking factor and phase-locked power across all modulation frequencies compared to the HC group, while normalized total power was not different between groups. Evoked, or phase-locked power was highly correlated with the phase-locking factor, as expected. There was no significant effect of history of mental illness noted for any MEG dependent measure.

Conclusions:

The preservation of total gamma-band power suggests a shift of power from phase-locked to non-phase-locked in the pASD group, consistent with prior results for ASD and pASD. Both the phase-locking factor and the normalized phase-locked power differences were greater in the left than in the right hemisphere. While ASSR gamma-band power may be significantly higher to 40+ Hz modulation rates, the phase-locking deficit appeared consistent across the range of frequencies employed in this study. History of mental illness, at least with respect to anxiety and mood disorders, does not appear to be related to the gamma-band findings. This study, together with previous results for the transient gamma-band response in first-degree relatives, highlights the potential of gamma-band deficits as a potential new ASD endophenotype.