Abstract
Parkinson's Disease (PD) is a clinically diagnosed disease that carries a reported misdiagnosis rate of 10-20%. Recent scientific discoveries have provided evidence of volatile organic compounds in sebum that are unique to patients with PD. The primary objective of this study was to determine if companion dogs could be trained to distinguish between sebum samples provided by PD-positive patients and PD-negative human controls. This was a randomized, handler-blind, controlled study. Twenty-three canines of varying breeds, ages, and environmental backgrounds were included. The study period encompassed 200 total working days from 2021 to 2022. Factors investigated included donor gender and levodopa drug affectivity, as well as canine breed, age, and duration of training time. The findings in this study were compiled from data collected during the final two years of a seven-year research program. For this two-year reporting period, when averaged as a group, the 23 dogs were 89% sensitive and 87% specific to olfactory distinction between PD-positive and PD-negative human donor samples. Ten of the twenty-three dogs averaged 90% or higher in both sensitivity and specificity. In 161 separate trials, a dog was presented with both novel PD-positive and PD-negative samples. For these novel exposures, the dogs collectively averaged 86% sensitivity and 89% specificity. PD medication was also investigated and was found to have no discernible impact on canine sensitivity or specificity results. Study findings support the application of companion dogs, trained with force-free, reward-based methodologies, for the detection of PD-positive and PD-negative samples under controlled conditions.