MedicalResearch.com Interview with:
Steven D. Hicks, M.D.,Ph.D
Department of Pediatrics
Penn State College of Medicine
MedicalResearch.com: What is the background for this study? What are the main findings?
Response: Since autism has both genetic and environmental underpinnings, my colleagues and I suspected that transcriptional elements (e.g. regulatory RNA molecules) might be different in the saliva of children with autism compared to peers without autism. We used a non-biased approach to analyze saliva from 372 children, and allowed machine learning techniques to inform which RNA elements best predicted autism status. To our surprise, microbial RNA levels and human RNA levels were equally powerful in predicting which children had autism. This may be because some children with autism eat restricted diets, resist tooth brushing, or put foreign objects in their mouths. The end result was a panel of 32 RNAs (20 human and 12 bacterial) that identified autism with 87% accuracy. Interestingly, when we tested the panel in a completely separate set of 84 children (including children from a different geographic region) the accuracy remained 88%.