Neural Correlates of Olfactory Dysfunction in ASD: Preliminary Results

Background: Sensory abnormalities are prevalent in autism spectrum disorder (ASD) and may be an important predictor of ASD severity in various domains of dysfunction. Although olfaction is not the most commonly affected sense in ASD, atypical olfactory processing is the most predictive of ASD severity and is more prevalent in ASD than in other neurodevelopmental disabilities. Despite its shared neural substrates with primary emotion areas including the amygdala and orbital frontal cortex (OFC), the olfactory system is under-studied in fMRI research in relation to other sensory systems.

Objectives: To gain further understanding of the neural basis behind olfactory processing differences in ASD we hypothesized that school-aged children with ASD would exhibit increased activation of primary and secondary olfactory areas (OFC, amygdala, piriform cortex) compared to typically developing (TD) children and children with sensory processing challenges without ASD (SPC). Our second hypothesis was that greater Blood-Oxygen-Level Dependent (BOLD) responses in primary and secondary processing areas, independent of diagnosis, would be related to poorer olfactory function.

Methods:  Eight children with ASD (Age M = 11.24, SD=1.6), ten typically developing (TD) (Age M = 9.61 SD=1.34) and eleven children with sensory processing challenges (SPC; Age M = 10.61 SD=1.5) were studied. Olfactory function was tested using the University of Pennsylvania Smell Identification Test (UPSIT). T1-weighted 3DMPRAGE and fMRI data were acquired on a 3T Philips Achieva scanner. During the fMRI, we exposed our participants to a rose-like odorant (phenyl ethyl alcohol), in four, 9 second blocks separated by at least 30 second intervals of air using a respiration-triggered odor paradigm. We tested for whole-brain group differences in activation in response to olfactory stimuli, and compared the relationship between brain activation and olfactory function using a fixed-effects model in FSL’s FEAT Software.

Results:  Participants with ASD yielded significantly greater OFC activation than TD participants (ASD>TD: Left OFC, Z-Max = 4.69, p=0.0085; Right OFC, Z-Max = 5.31, p<0.001; Figure 1A) as well as participants with SPC (ASD > SPC: Left OFC, Z-Max = 5.27, p<0.001; Right OFC, Z-Max = 5.34, p<0.001; Figure 1B). No significant group differences were observed in the amygdala or piriform cortex. We found a negative correlation between OFC activation and UPSIT score across the groups (Left OFC, Z-Max = 4.61, p=0.0017; Right OFC, Z-Max = 5.34, p=0.0014; Figure 2).

Conclusions:  Children with ASD displayed greater bilateral OFC activation than children in both comparison groups, partially supporting our first hypothesis of ASD hyperactivation in olfactory areas. Additionally, we found a relationship between greater OFC activation and poorer olfactory function. Together, these results highlight the possibility that measures of olfactory functioning are sensitive to OFC dysfunction in ASD, which is consistent with the OFC’s role in odor identification. Further, these results indicate that our method shows promise as an early diagnosis biomarker, as odor perception can be measured safely using fMRI with neonatal subjects. Data collection is ongoing and we aim to validate the current preliminary findings and further characterize the role of OFC dysfunction, as well as related neural circuitry, in ASD.