Parkinson’s disease produces a subtle, musky odor that most people never consciously notice. In formal sensory testing, people who could detect it described the scent as musty, strong, and unpleasant compared to the body odor of healthy individuals. The smell originates in the oily substance called sebum that coats your skin, and it has attracted serious scientific attention since a retired nurse in Scotland demonstrated she could identify people with Parkinson’s by sniffing their worn T-shirts. What began as a curiosity has grown into an active area of diagnostic research, with trained dogs, electronic sensors, and mass spectrometry all showing they can pick up the disease’s chemical signature from a simple skin swab.
How the Smell Was Discovered
The story starts with Joy Milne, a retired nurse from Perth, Scotland, who noticed her husband Les had developed a different smell years before he was diagnosed with Parkinson’s disease. She described it as a changed, musky quality to his scent, concentrated around the upper back and neck. When Les joined a Parkinson’s support group, Milne realized other members of the group carried the same odor. She mentioned this to researchers at a Parkinson’s UK event, and the unusual claim eventually reached scientists at the University of Manchester.
In a now-famous pilot test, Milne was given T-shirts worn by people with Parkinson’s and by healthy volunteers. She correctly identified the Parkinson’s shirts and also flagged one control subject as having the disease. That person was diagnosed with Parkinson’s several months later. This striking result, while anecdotal, was enough to launch a formal investigation into whether Parkinson’s really does change the way the body smells and, if so, what chemical compounds are responsible.
What the Scent Actually Smells Like
Describing a disease odor precisely is tricky because most people lack the olfactory sensitivity to detect it at all. In a 2024 study that asked human volunteers to compare body odor samples from people with Parkinson’s to those from healthy controls, participants rated the Parkinson’s samples as significantly more musty, stronger, and more unpleasant overall. The odor did not smell like any single familiar thing. It was more of a shift in the character of normal body odor: heavier, greasier, and less clean.
1PubMed. Human perception of Parkinson’s disease body odor in comparison to the volatile organic compounds of Parkinson’s diseaseJoy Milne herself has described it in various interviews as a musky, yeasty smell, sometimes comparing it to the scent you might notice on someone who has been sweating heavily but with an added oily quality. Other people with heightened smell sensitivity have used words like “woody” or “slightly sweet.” The inconsistency in these descriptions makes sense when you consider that what people are detecting is not a single odor molecule but a shifted blend of dozens of volatile compounds, each present at very low concentrations. Your brain interprets that blend holistically, and the resulting impression differs somewhat from person to person.
The Chemistry Behind the Smell
The odor comes from volatile organic compounds released by sebum, the waxy, oily coating that sebaceous glands produce on your skin. Researchers at the University of Manchester collected sebum by swabbing the upper backs of 64 participants, then used gas chromatography and mass spectrometry to identify the chemical fingerprint. They found a distinct volatile signature in the Parkinson’s group, with altered levels of several compounds. Two stood out: perillic aldehyde, a molecule with a citrus-like, slightly herbal note, and eicosane, a long-chain hydrocarbon. When the “Super Smeller” was presented with synthetic versions of these compounds, she confirmed the smell was highly similar to the Parkinson’s scent she had originally noticed on her husband.
2PubMed Central. Discovery of Volatile Biomarkers of Parkinson’s Disease from SebumSubsequent work identified additional compounds of interest, including octanal and hexyl acetate. Octanal has an orange-peel, slightly waxy aroma; hexyl acetate smells fruity, like green apples or pears. These are not new or exotic molecules. They already exist in everyone’s sebum at low levels. What changes in Parkinson’s is the ratio: some go up, others go down, and the overall blend shifts enough that a sensitive nose or a calibrated instrument can tell the difference. The signature has held up across independent groups of subjects sampled in different parts of the UK, adding confidence that it reflects the disease itself rather than some local environmental factor.
3PubMed Central. Classification of Parkinson’s disease and isolated REM sleep behaviour disorder: delineating progression markers from the sebum volatilomeWhy Parkinson’s Changes Your Skin’s Oil
People with Parkinson’s often develop noticeably oilier skin, a symptom dermatologists call seborrhea. This happens because the disease affects the autonomic nervous system, which regulates gland activity throughout the body, including the sebaceous glands in your skin. The overproduction of sebum has been recognized clinically for decades, but only recently have researchers started looking at what that extra sebum is made of.
Metabolomic analysis of sebum from people with Parkinson’s has revealed broad disruptions in lipid metabolism. Pathways involved in fatty acid production, sphingolipid processing, and arachidonic acid metabolism are all altered.
4Nature Communications. Metabolomics of sebum reveals lipid dysregulation in Parkinson’s diseaseA likely driver of these changes is alpha-synuclein, a protein that misfolds and clumps in Parkinson’s disease, famously forming Lewy bodies in the brain. But alpha-synuclein aggregates also accumulate in the peripheral nervous system, including in the nerves that supply the skin.
5Journal of Neuropathology & Experimental Neurology. Skin nerve phosphorylated α-synuclein in the elderlyLab experiments exposing sebaceous gland cells to aggregated alpha-synuclein have shown that it pushes those cells toward a more mature, fat-producing state, ramping up the production of neutral lipids and the enzymes that make them.
6PubMed Central. Effects of α-Synuclein on the Lipid Phenotype of SZ95 Human Sebocytes: A Preliminary Study in the Context of Parkinson’s DiseaseThe result is skin that produces more sebum with an abnormal lipid composition. As that altered sebum sits on the skin and is broken down by bacteria and exposure to air, it releases a different mix of volatile compounds. That different mix is the smell.
Can Dogs Smell Parkinson’s Disease?
Yes, and with impressive consistency. Researchers have been training detection dogs to discriminate between skin swabs from people with Parkinson’s and those from healthy volunteers, drawing on the same canine olfactory abilities used in explosive and narcotics detection. In a 2025 double-blind trial, two dogs trained over roughly a year on more than 200 samples were tested on 100 new samples, 40 from drug-naïve Parkinson’s patients and 60 from controls. One dog achieved 80% sensitivity and 98% specificity; the other hit 70% sensitivity and 90% specificity.
7PubMed Central. Trained dogs can detect the odor of Parkinson’s diseaseA larger program involving 23 household companion dogs of various breeds reported even broader results. Over a two-year training period, the dogs averaged 89% sensitivity and 87% specificity across thousands of encounters with human donor samples.
8PubMed Central. From small to tall: breed-varied household pet dogs can be trained to detect Parkinson’s DiseaseThe fact that these were ordinary pet dogs rather than specialist working breeds is noteworthy. It suggests the odor signal is robust enough that most dogs can learn to detect it, not just unusually talented animals. These programs are still in the proof-of-concept stage, though; no one is deploying diagnostic dogs in a clinical setting yet. The value of the dog studies is as confirmation that the chemical signal is real and consistent enough for a biological nose to identify.
Electronic Noses and Laboratory Tests
If dogs can smell it, the thinking goes, a machine should be able to, too. Several groups have developed electronic nose devices, essentially arrays of chemical sensors paired with pattern-recognition software, aimed at picking up the Parkinson’s volatile signature from skin swabs. One early prototype identified three key odor compounds (octanal, hexyl acetate, and perillic aldehyde) that differed between Parkinson’s patients and controls. On a validation set of 24 new subjects, the device was 92% sensitive at correctly flagging true Parkinson’s cases, though its specificity was only 50%, meaning it generated a lot of false positives. Applying machine learning to the full odor profile rather than just three compounds improved overall accuracy to about 79%.
More sophisticated laboratory techniques have pushed the numbers further. A method using paper spray ionization coupled with ion mobility mass spectrometry could classify Parkinson’s versus control sebum in about three minutes per swab, using molecular classes of lipids rather than volatile compounds alone.
9PubMed Central. Paper Spray Ionization Ion Mobility Mass Spectrometry of Sebum Classifies Biomarker Classes for the Diagnosis of Parkinson’s DiseaseNewer work has focused on quantifying specific lipids from skin swabs using liquid chromatography-tandem mass spectrometry, moving toward the kind of standardized, reproducible measurement that a clinical diagnostic test requires.
10PubMed. Benchmarking Lipid Quantitation from Skin Swab Sebum-Rich Samples to Establish Mass Spectrometry-Based Diagnostic MethodologyThe appeal of a skin swab test is obvious. Right now, Parkinson’s is diagnosed based on clinical symptoms like tremor, rigidity, and slowness of movement, and by the time those symptoms are visible enough for a confident diagnosis, a substantial amount of brain cell loss has already occurred. A noninvasive test that could flag the disease earlier, from a simple gauze pad rubbed across the upper back, would be transformative. The technology is not there yet, but the trajectory of the research is promising.
Do Parkinson’s Medications Affect the Smell?
This is an important question for any potential diagnostic test. If the drugs used to treat Parkinson’s change the sebum volatile profile, it would be hard to tell whether you are detecting the disease or the treatment. The evidence so far is reassuring. In one study that compared the volatile compound profiles of drug-naïve Parkinson’s patients with those already on medication, the overall profile reliably separated Parkinson’s from healthy controls with about 88% accuracy, but the measured compounds did not reliably distinguish between medicated and unmedicated patients.
11npj Parkinson’s Disease. Detection of Volatile Organic Compounds as an emerging strategy for Parkinson’s disease diagnosis and monitoringThis suggests that the core chemical signature is driven by the disease process itself rather than by levodopa or dopamine agonists. The dog studies reinforce this interpretation, since the 2025 canine trial deliberately used only drug-naïve patients as targets and the dogs still identified them successfully. That said, this is still early-stage evidence, and larger studies across more diverse patient populations are needed before anyone can state definitively that medications play no role.
The Irony of Smell Loss in Parkinson’s
There is a striking paradox at the heart of this whole field. Parkinson’s disease produces a detectable smell, yet one of its earliest and most common symptoms is a loss of the patient’s own sense of smell. Hyposmia, or reduced ability to smell, is recognized as a frequent and early feature of the disease, often appearing years before the movement symptoms that lead to a formal diagnosis.
12PubMed Central. Hyposmia in Parkinson’s disease; exploring selective odour lossThis means the person whose body is emitting an altered scent is typically the last person who would be able to notice it. Joy Milne’s husband, Les, had no idea he smelled different. The disconnect makes biological sense when you consider where the disease strikes. The olfactory bulb, the brain structure that processes smell, is one of the earliest regions affected by alpha-synuclein pathology. Meanwhile, the sebaceous glands in the skin are driven by peripheral nerve changes that develop on a parallel but separate track. Two different tissues are affected by the same underlying disease process, producing two completely independent symptoms: one that you lose (your sense of smell) and one that emerges for others to detect (your changed body odor).
Smell loss on its own is not specific to Parkinson’s. Plenty of other conditions, from sinus problems to normal aging, can dull your sense of smell. But when smell loss is combined with other early features like sleep disturbances or constipation, clinicians already use it as a red flag. Adding a skin-swab test for the other side of the equation, the sebum volatile profile, could eventually make the prodromal picture much clearer.
Skin Biopsies and Alpha-Synuclein Testing
Sebum volatiles are not the only skin-based approach to Parkinson’s diagnosis being explored. A separate line of research has focused on detecting misfolded alpha-synuclein directly in skin nerve fibers using a technique called RT-QuIC, which amplifies tiny amounts of the abnormal protein until they are measurable. In a two-laboratory validation study, this skin biopsy method correctly identified roughly 91% of Parkinson’s patients and correctly excluded about 90% of controls.
13npj Parkinson’s Disease. Diagnostic value of skin RT-QuIC in Parkinson’s disease: a two-laboratory studyThe skin biopsy approach is more invasive than a swab, requiring a small punch biopsy usually taken from the leg or the back of the neck, but it measures the disease’s hallmark protein directly rather than relying on a downstream metabolic consequence. These two approaches, volatile profiling and alpha-synuclein detection, are not competitors so much as complementary tools. A swab-based volatile test could serve as a quick, cheap initial screen, with alpha-synuclein confirmation reserved for cases where the screen is positive. Neither has been validated at the scale needed for clinical deployment, but both are moving through the pipeline.
Can You Smell It on Yourself or a Family Member?
The honest answer is almost certainly not. Joy Milne appears to have an unusually sensitive sense of smell, a trait sometimes called hyperosmia, and even she did not immediately recognize it as a sign of disease. She noticed her husband smelled different and found it mildly unpleasant but did not connect it to Parkinson’s until years later. For most people, the change is too subtle to detect consciously. It exists at a level that requires either a trained animal nose, an extremely sensitive human nose, or analytical chemistry equipment.
If you have noticed that a family member with Parkinson’s has a stronger or more unpleasant body odor, that could reflect the increased sebum production that accompanies the disease rather than a conscious detection of the specific volatile signature. Oilier skin naturally smells stronger, and increased seborrhea is a well-known dermatological finding in Parkinson’s patients. But translating a vague “something smells different” impression into any kind of diagnostic judgment is not realistic and is not something anyone in the research community is suggesting laypeople try to do. The science is aimed at building objective, reproducible tools, not sharpening individual noses.
What Else Changes About the Skin in Parkinson’s
Sebum overproduction is the most studied skin change, but it is not the only one. People with Parkinson’s frequently develop seborrheic dermatitis, a condition characterized by red, flaky, itchy patches in oily areas of the skin like the scalp, forehead, and sides of the nose. The connection appears to be that excess sebum provides a rich food source for Malassezia yeasts that naturally live on the skin, leading to inflammation. Some people find that seborrheic dermatitis shows up years before any neurological symptoms and is later recognized in hindsight as an early sign.
Excessive sweating, or hyperhidrosis, is another common complaint. Like sebum production, sweating is regulated by the autonomic nervous system, which Parkinson’s progressively disrupts. Changes in sweating patterns may contribute their own volatile compounds to the overall body odor profile, though this has received far less research attention than sebum. The skin microbiome, the community of bacteria and fungi living on the skin surface, is also likely to shift in response to altered sebum composition, and those microbes play a large role in generating the volatile compounds that constitute body odor. Teasing apart the contributions of host metabolism, microbial metabolism, and environmental exposure remains one of the open challenges in this field.
The broader takeaway from all this skin research is that Parkinson’s is not purely a brain disease. It is a systemic condition that leaves traces across the body, from the gut to the skin to the olfactory system. The skin just happens to be the most accessible tissue for sampling, which is why it has become such a focus for diagnostic innovation. A gauze pad on the upper back is about as noninvasive as a medical test can get, and if the volatile or lipid signature proves reliable enough in large-scale trials, it could eventually join the toolkit clinicians use to catch the disease before its most damaging phase.