Trained Dogs Achieve Up to 98% Accuracy in Detecting Parkinson’s Disease from Skin Swabs

Trained Dogs Achieve Up to 98% Accuracy in Detecting Parkinson’s Disease from Skin Swabs

(IN BRIEF) A collaborative study by the University of Bristol, Medical Detection Dogs, and the University of Manchester found that two trained dogs can detect Parkinson’s disease from sebum skin swabs with up to 80% sensitivity and 98% specificity. In double‑blind trials, Golden Retriever Bumper and Black Labrador Peanut correctly identified PD samples among over 200 swabs, even when patients had other illnesses. This non‑invasive approach, detailed in The Journal of Parkinson’s Disease, highlights the potential of canine olfaction to serve as an early diagnostic biomarker—vital for timely intervention in a condition whose symptoms may precede diagnosis by decades. Lead researchers emphasize dogs’ performance as a foundation for developing quick, cost‑effective screening tools.

(PRESS RELEASE) BRISTOL, 15-Jul-2025 — /EuropaWire/ — Researchers from the University of Bristol, in partnership with Medical Detection Dogs and the University of Manchester, have demonstrated that trained canines can accurately identify Parkinson’s disease (PD) through odour samples taken from the skin. Published today in The Journal of Parkinson’s Disease (15 July), the study showed that two specially trained dogs could distinguish sebum swabs from people with PD against those without, achieving sensitivities up to 80% and specificities as high as 98%. Remarkably, their success extended to samples from patients with additional health conditions.

Bumper, a Golden Retriever Medical Detection dog
Image credit: Medical Detection Dogs

Over several weeks, Medical Detection Dogs conditioned the Golden Retriever “Bumper” and Black Labrador “Peanut” on more than 200 odour samples—both from PD‑positive individuals and control participants. During double‑blind trials, where only a computer knew sample positions, swabs were presented on a stand and dogs received rewards for correctly indicating PD samples and for ignoring non‑PD ones. To ensure thoroughness, each line of samples was presented twice (including a reversed order), and any unassessed swabs were regrouped and retested until every sample received a definitive decision.

Currently, no early diagnostic test exists for Parkinson’s, even though its symptoms can emerge up to 20 years before diagnosis. Identifying reliable biomarkers is therefore crucial for prompt treatment, which may slow disease progression and alleviate symptoms. Claire Guest, CEO and Chief Scientific Officer of Medical Detection Dogs, remarked, “Once again, we see dogs’ extraordinary ability to detect disease with remarkable accuracy. A non‑invasive early test for Parkinson’s could transform patient care.”

Nicola Rooney, Associate Professor at the University of Bristol Veterinary School and lead author, added, “Our findings confirm an odour signature unique to Parkinson’s. With sensitivities of 70–80% well above chance, dogs could pave the way for a rapid, low‑cost diagnostic tool.” Perdita Barran, Professor of Mass Spectrometry at the University of Manchester, noted, “Inspired by Joy Milne and our Nose2Diagnose programme, this research strengthens the case for skin swabs as a simple early detection method.”

‘Trained dogs can detect the odour of Parkinson’s Disease’ by Nicola Rooney, Perdita Barran, Claire Guest et al. in The Journal of Parkinson’s Disease [open access]

Further information

Medical Detection Dogs
Medical Detection Dogs is the world-leading organisation for research into canine olfactory diagnostics.  We train dogs to detect the odour of disease with the aim of developing faster, more efficient and less invasive diagnostics that lead to better patient outcomes.

Our Bio Detection research includes cancer, neurological disease and bacterial infections and has the potential to benefit millions.  We already apply what we know about the science of canine olfaction to benefit people by training Medical Alert Assistance Dogs, which help individuals manage complex, life-threatening medical conditions.

Media Contact:

Tel: +44 (0)117 928 9000
Email: press-office@bristol.ac.uk

SOURCE: University of Bristol