Diminished Variability of Neural Circuits in Autism: MEG Studies of Tactile Evoked Response

Background: A consensus is rapidly emerging that atypical neural connectivity is a central characteristic of the autism phenotype. However, there is far less consensus as to what exactly is atypical about the circuitry. One proposal is that the neural circuits of persons with autism are noisy or highly variable. Imbalance in the excitation/inhibition ratio and suppression of cortical inhibition support the hypothesis of noisy synapses. Empirical tests of this influential theoretical proposal are required.

Objectives: The noisy synapses hypothesis predicts reduced reliability or increased variability in the evoked responses of ADs. We tested this prediction by examining, in the brains of individuals with autism, the response to tactile stimulation using magnetoencephalography (MEG). To further dissect the precise nature of the putative difference in variability in individuals with autism versus control, we also examined the evoked MEG responses to touch of individuals with spina bifida and elderly controls.

Methods: Using a whole-head neuromagnetometer containing an array of 248 gradiometers in a MEG recording setup, we recorded the neural response to passive tactile stimulation of the thumb (RD1) and index finger (RD2) of the dominant hand of young adult control participants (17 high-functioning persons with autism and 18 typically developing persons, matched for gender and age) while they remained awake in an eyes-closed supine posture. We also used data recorded from 21 individuals with spina bifida (Myelomeningocele) and 8 elderly control participants, obtained under similar stimulation and recording conditions as for the autism group.
We measured the mean and variability of the evoked response to tactile stimulation in all four groups. For each participant in our sample, the sensor in the contralateral somatosensory cortex that exhibited the largest evoked response relative to baseline was automatically selected. For this sensor, we computed across all trials and times (40-275ms following stimulus onset; 290 Hz sampling rate) for each body part the a) mean evoked response, b) variance in the evoked response, and c) the ratio of the variance to the mean of the response, a measure that is analogous to Fano factor.

Results: The evoked responses to tactile stimulation were less, not more, variable in the brains of persons with autism as compared to control (p<<0.001). The evoked responses of the spina bifida group was significantly less variable as compared to both controls as well as persons with autism (p<<0.001). The responses in the elderly controls were significantly less variable than those of the young adult controls (p<<0.001), whereas no difference was observed in comparison with the autism group (p>0.1).

Conclusions: The results thus far run counter to the hypothesis of noisy synapses in autism. Rather, diminished variability of neural response in autism and in other groups could be what distinguishes atypical brains from those of young, healthy control individuals.