Impaired Olfactory Detection and Identification in Children and Young Adults with ASD

Background: Individuals with autism spectrum disorder (ASD) often experience significant sensory processing differences, including atypical chemosensory (i.e., taste and smell) functioning. Carefully characterizing olfactory differences in ASD is important given the functional impact they may have on the significant feeding challenges in this population, as well as shared neural circuitry between olfactory and social regions. Extant olfactory findings in ASD have varied, with some studies finding lower detection and/or identification in individuals with ASD and others finding no differences. Further research is needed to reconcile these findings, including controlling for other factors affecting performance, and to examine whether olfactory differences are specific to ASD or present in the broader phenotype.

Objectives: This study aimed to clarify and extend the ASD olfaction literature by examining detection and identification abilities in a large, well-characterized sample of children and young adults with ASD, while also assessing other factors that might influence performance (e.g., exposure, receptive language). To examine shared genetic and environmental factors, participants with ASD were compared to both unaffected siblings and typically developing (TD) controls.

Methods: Participants were 85 children and young adults with well-characterized ASD, 49 unaffected siblings of the participants with ASD, and 66 unrelated TD controls. Ages ranged from 10-24 years, and groups did not differ by age. IQs were > 70. Receptive language was also measured with the PPVT-4, which has a similar response format to the identification task.

Olfactory detection thresholds were determined using the Smell Threshold Test, in which varying concentrations of phenyl ethyl alcohol were presented bilaterally in a two-interval forced-choice adaptive paradigm. Odor identification was assessed using the University of Pennsylvania Smell Identification Test (UPSIT), in which participants identified a target odor amongst four choices (presented with photographs in this study to decrease verbal demands). Finally, each participant’s prior exposure to UPSIT target odors was assessed via parent-report questionnaire.

Results: Kruskal-Wallis analysis of variance (ANOVA) revealed an overall group difference in olfactory detection, X² (2)=19.65, p<.001. Follow-up comparisons indicated that individuals with ASD had higher (worse) thresholds than siblings (p<.001) and TD controls (p=.001), who did not differ. Olfactory identification, analyzed via ANOVA, also revealed an overall group difference, F(2,195)=16.65, p<.001. The ASD group showed worse identification compared to siblings (p<.001) and controls (p<.001), who did not differ. To account for differences in exposure, we calculated adjusted scores based only on target odors each participant had previously experienced. After adjusting for exposure history, the ASD group remained significantly worse than both siblings (p<.001) and controls (p=.02). A similar pattern of results emerged after controlling for age, gender, receptive language, and olfactory detection abilities.

Conclusions: These results support reduced olfactory detection and identification abilities in children and young adults with ASD, even after controlling for other factors influencing task performance. Furthermore, unaffected siblings demonstrated intact olfactory abilities, which were statistically comparable to TD controls, suggesting a unique olfactory phenotype in individuals with ASD, which does not extend to full siblings without the diagnosis.