Parkinson’s disease produces a distinctive smell, carried in the oily substance that coats human skin. The odor comes from changes in sebum, the waxy secretion of sebaceous glands, which takes on an altered chemical profile in people with Parkinson’s. A woman in Scotland first brought this to researchers’ attention when she noticed a musky scent on her husband years before he was diagnosed, and the science that followed has turned a curious anecdote into a genuinely promising avenue for early, noninvasive detection.
The Woman Who Could Smell Parkinson’s
Joy Milne, a retired nurse from Perth, Scotland, noticed a change in her husband Les’s scent roughly six years before he received a Parkinson’s diagnosis. She described it as a musky, oily smell, different from normal body odor. After Les was diagnosed and she attended a Parkinson’s support group, she realized other patients carried the same scent. She mentioned this to researchers at the University of Edinburgh and later the University of Manchester, who were understandably skeptical but agreed to test her.
In a controlled study, Milne was given T-shirts worn by people with Parkinson’s and healthy volunteers. She correctly identified the Parkinson’s subjects by smell alone. She also flagged one control subject’s shirt as carrying the scent. That person was diagnosed with Parkinson’s several months later, an outcome that stunned the research team. The researchers then moved from anecdote to chemistry, analyzing sebum samples from the upper back of 64 participants. Using mass spectrometry, they identified a distinct volatile signature for Parkinson’s, with altered levels of compounds including perillic aldehyde and eicosane. When Milne was asked to smell synthesized versions of those molecules, she described the scent as highly similar to the smell she associated with the disease.1PubMed Central. Discovery of Volatile Biomarkers of Parkinson’s Disease from Sebum
Why Parkinson’s Changes the Way Skin Smells
The smell originates in sebum, the oily film your sebaceous glands produce to lubricate and protect your skin. People with Parkinson’s tend to have noticeably oilier skin than average. Clinicians have long recognized this as “seborrheic” skin, and conditions like seborrheic dermatitis are more common in people with Parkinson’s than in the general population.2PubMed Central. Dermatological Manifestations and Sebum Composition in Parkinson’s Disease The excess oil is not just cosmetically different; its chemical makeup shifts in ways that produce a new set of volatile compounds, and those compounds are what create the smell.
The leading explanation for the sebum overproduction ties back to the same dopamine loss that causes Parkinson’s motor symptoms. When dopamine drops, a chain of hormonal events follows. Levels of a hormone called alpha-MSH rise, which stimulates sebaceous glands to pump out more oil. Prolactin also increases, and it too pushes sebaceous glands to work harder.2PubMed Central. Dermatological Manifestations and Sebum Composition in Parkinson’s Disease The result is not just more sebum, but sebum with a different lipid composition, including higher levels of certain short-chain fatty acids and disrupted sphingolipid metabolism.3Dermatology Practical & Conceptual. Dermatological Manifestations and Sebum Composition in Parkinson’s Disease
There is also a deeper molecular story. Alpha-synuclein, the protein that clumps abnormally in the brains of people with Parkinson’s, appears to affect sebaceous gland cells directly. When researchers exposed lab-grown sebocytes to a form of alpha-synuclein, the cells shifted toward a more mature, lipid-producing state. They accumulated more triglycerides, ceramides, and certain phospholipids, and they ramped up genes involved in fat production. The same pattern showed up in actual skin biopsies from people with Parkinson’s compared to controls.4PubMed Central. Effects of α-Synuclein on the Lipid Phenotype of SZ95 Human Sebocytes: A Preliminary Study in the Context of Parkinson’s Disease This suggests that the altered skin chemistry is not just a downstream consequence of dopamine loss but may be directly tied to the core protein pathology of the disease.
What the Odor Is Made Of
When sebum sits on the skin, its lipids break down and release volatile organic compounds, the gaseous molecules that your nose can detect. In Parkinson’s, the altered lipid profile means a different set of volatiles, and researchers have been working to catalog exactly which ones change. The original Manchester team identified perillic aldehyde and eicosane as two compounds whose levels shifted in Parkinson’s sebum, and whose combined smell Milne recognized as the disease’s signature.1PubMed Central. Discovery of Volatile Biomarkers of Parkinson’s Disease from Sebum
Later work has expanded the list. A Chinese study using gas chromatography with ion mobility spectrometry found a cluster of volatile biomarkers in sebum from Parkinson’s patients, distinct enough that a machine-learning model built on those markers could distinguish patients from healthy controls with about 84% accuracy.5Microchemical Journal. Volatile organic compounds profile of sebum from patients with Parkinson’s disease by gas chromatography-ion mobility spectrometry Another study found that the volatile peaks in sebum headspace were consistently higher in Parkinson’s subjects, concentrated in a narrow time window during analysis, which fits with the idea that specific compounds are elevated rather than the whole volatile profile being uniformly different.6PubMed Central. A fast and non-invasive artificial intelligence olfactory-like system that aids diagnosis of Parkinson’s disease
The picture that emerges is not a single “Parkinson’s molecule” but a constellation of compounds whose relative amounts shift. This is part of why the smell is hard to pin down in everyday language. It is not like the fruity breath of uncontrolled diabetes or the ammonia scent of kidney failure, conditions where a single dominant molecule defines the odor. The Parkinson’s scent is subtler, a composite signature that most human noses cannot consciously parse but that stands out to someone with an unusually acute sense of smell.
Dogs Can Smell It Too
If a human with hyperosmia can detect Parkinson’s by smell, dogs should be able to do it even more reliably, and the research bears this out. Dogs have roughly 300 million olfactory receptors compared to around six million in humans, making their sense of smell orders of magnitude more sensitive to many compounds.
A 2024 study trained 23 pet dogs of various breeds to distinguish between sebum samples from people with Parkinson’s and healthy controls. Across two years of testing, the dogs averaged 89% sensitivity and 87% specificity as a group. The best-performing dogs exceeded 90% on both measures. When exposed to completely novel samples they had never encountered before, the dogs still averaged 86% sensitivity and 89% specificity.7PubMed Central. From small to tall: breed-varied household pet dogs can be trained to detect Parkinson’s Disease A separate study using two dogs tested against drug-naive patients, people who had never taken Parkinson’s medication, confirmed the finding with sensitivity of 70% and 80% and specificity of 90% and 98% respectively.8PubMed Central. Trained dogs can detect the odor of Parkinson’s disease
The drug-naive detail matters because a recurring question in this field is whether the smell comes from the disease itself or from the medications used to treat it. The fact that dogs can detect the odor in patients who have never taken levodopa or dopamine agonists strongly suggests the scent is intrinsic to the disease process, not a pharmaceutical artifact. The canine research is still experimental and unlikely to become a standard clinical tool, but it provides powerful biological confirmation that the volatile signature is real and consistent enough for a trained nose, human or otherwise, to pick up.
Building an Electronic Nose
Dogs and super smellers are fascinating proof of concept, but neither is practical for screening millions of people. The real diagnostic potential lies in devices that can analyze sebum volatiles quickly, cheaply, and at scale. Several research groups are working on exactly this.
One approach combines fast gas chromatography with a surface acoustic wave sensor and machine-learning algorithms to create what researchers call an artificial intelligent olfactory system. In a study of 87 sebum samples from a Chinese hospital, the system could distinguish Parkinson’s patients from healthy controls based on odor profiles.9PubMed Central. Artificial Intelligent Olfactory System for the Diagnosis of Parkinson’s Disease A similar system tested in a separate group found that volatile peaks measured in the first few seconds of analysis already carried diagnostic information.6PubMed Central. A fast and non-invasive artificial intelligence olfactory-like system that aids diagnosis of Parkinson’s disease These devices are still laboratory prototypes, but their speed is encouraging. If the analysis takes seconds rather than hours, population-level screening becomes imaginable.
Another approach skips volatile analysis entirely and goes straight to the lipid fingerprint. Paper spray ionization coupled with ion mobility mass spectrometry can analyze a sebum skin swab in about three minutes, pulling roughly 4,200 molecular features from each sample. Researchers used this method on 150 subjects and found patterns of lipid regulation that were diagnostic for Parkinson’s.10PubMed Central. Paper Spray Ionization Ion Mobility Mass Spectrometry of Sebum Classifies Biomarker Classes for the Diagnosis of Parkinson’s Disease Work is also underway to standardize quantitation methods for lipids collected from skin swabs, using liquid chromatography paired with mass spectrometry, to make these measurements robust and reproducible enough for routine clinical use.11PubMed. Benchmarking Lipid Quantitation from Skin Swab Sebum-Rich Samples to Establish Mass Spectrometry-Based Diagnostic Methodology
Why Early Detection Matters So Much
Parkinson’s is currently diagnosed based on motor symptoms: tremor, rigidity, slow movement. By the time those symptoms are obvious enough for a diagnosis, a substantial proportion of the dopamine-producing neurons in the brain have already been lost. There is no blood test, no imaging scan in routine use, and no biomarker that a GP can order from a standard lab. This means that any experimental treatment aimed at slowing or stopping neurodegeneration has to work on a brain that is already significantly damaged. If you could identify people earlier, before motor symptoms fully develop, you could intervene at a point where there are more neurons left to protect.
This is where the sebum smell becomes especially interesting. A condition called isolated REM sleep behavior disorder, where people physically act out their dreams, is one of the strongest known predictors of future Parkinson’s. The majority of people with this sleep disorder eventually develop Parkinson’s or a related condition. When researchers asked Joy Milne to smell sebum from people with isolated REM sleep behavior disorder alongside Parkinson’s patients and healthy controls, she could distinguish all three groups. She classified three of the sleep disorder subjects as smelling like Parkinson’s, and two of those three went on to show clinical signs of conversion to Parkinson’s at follow-up appointments.12PubMed Central. Classification of Parkinson’s disease and isolated REM sleep behaviour disorder: delineating progression markers from the sebum volatilome
That result, while based on small numbers, is striking. It suggests the sebum signature does not just reflect established Parkinson’s but may track the progression from a pre-clinical state. If this holds up in larger studies, a skin swab could eventually serve as a screening tool for people at elevated risk, flagging them for closer monitoring or enrollment in prevention trials long before they develop tremor or shuffling gait.
Practical Hurdles for a Skin Swab Test
Turning any biomarker discovery into a real-world clinical test is a long road. The sebum approach has a few things working in its favor and a few challenges to overcome.
On the favorable side, collecting sebum is about as noninvasive as a medical test can get. A gauze pad rubbed on the upper back or behind the ear captures enough material for analysis. There is no needle, no fasting, no preparation. And the samples appear to be remarkably stable. A study that tested the effects of storage temperature and storage time on sebum samples found that neither significantly altered the results, drawing on data from over 650 participants analyzed across multiple platforms.13PubMed Central. How storage post sampling influences the stability of sebum when used for mass spectrometry metabolomics analysis? This stability is a big deal because it means you could realistically collect a sample at home and mail it to a testing lab, the same model that has made at-home genetic testing and colon cancer screening kits successful.
On the challenging side, the studies so far have been relatively small, typically dozens to low hundreds of participants. Diagnostic accuracy around 84-90% is promising for a research setting but needs to be validated in much larger, more diverse populations before anyone would rely on it for screening. Parkinson’s also exists on a spectrum with other conditions that cause tremor, stiffness, or cognitive changes, and a useful test has to distinguish not just Parkinson’s from healthy but Parkinson’s from conditions that mimic it. Few studies have tested the sebum approach against these lookalike conditions.
There is also the question of what other factors might muddy the results. Skin microbiome composition, use of topical products, ambient temperature, medication status, and co-existing skin conditions like seborrheic dermatitis could all potentially affect sebum composition. Researchers are aware of these confounders and are beginning to control for them, but large-scale validation under real-world conditions is still needed.
A Longer History of Diagnostic Smells
The idea that disease changes the way a person smells is ancient. Physicians in the pre-modern era were taught to use their noses as diagnostic tools. The breath of someone with uncontrolled diabetes has a fruity, acetone-like quality. Liver failure produces a characteristic sweetish odor sometimes called fetor hepaticus. Certain metabolic disorders in newborns are named for their smells, like maple syrup urine disease. A 1976 review catalogued dozens of conditions with recognized odor profiles, from infections to organ failure, long before modern analytical chemistry made it possible to identify the molecules responsible.14PubMed Central. Smell as a diagnostic marker
What makes the Parkinson’s discovery unusual in this tradition is both the subtlety of the smell and the location. Most disease-associated odors come from breath or urine, and they tend to be noticeable to anyone standing nearby. The Parkinson’s scent lives on the skin, embedded in sebum, and it takes either an unusually sensitive nose or analytical equipment to detect it. That subtlety is probably why it went unrecognized for so long despite the disease being described over two centuries ago. It also explains why clinical diagnosis based on smell alone is not practical for Parkinson’s the way it might be for, say, a urinary tract infection that a nurse can identify from across the room. The research value lies not in training clinicians to sniff patients but in building devices that can read the molecular fingerprint behind the scent.
Skin as a Window Into the Brain
Perhaps the most conceptually interesting part of the Parkinson’s odor research is what it reveals about the disease itself. Parkinson’s has traditionally been understood as a brain disease, centered on the loss of dopamine neurons in a region called the substantia nigra. But the finding that alpha-synuclein directly alters the behavior of skin cells adds to a growing body of evidence that Parkinson’s is a whole-body disease with manifestations far beyond the brain.4PubMed Central. Effects of α-Synuclein on the Lipid Phenotype of SZ95 Human Sebocytes: A Preliminary Study in the Context of Parkinson’s Disease
Alpha-synuclein deposits have been found in the gut, the salivary glands, the skin, and the autonomic nervous system of people with Parkinson’s. The smell of the disease may be one expression of this broader peripheral pathology. If so, the skin is not just a convenient sampling site; it is a genuine tissue of disease involvement, one that happens to be far more accessible than brain tissue. That accessibility makes it useful not only for diagnosis but potentially for monitoring disease progression or treatment response over time. A patient could provide a skin swab at each clinic visit, and changes in the sebum profile might track with how their disease is evolving, offering information that currently requires expensive brain imaging or subjective clinical scales to estimate.
The lipid-rich environment of Parkinson’s skin also feeds microbial overgrowth. A yeast called Malassezia, which normally lives on everyone’s skin in low numbers, tends to proliferate in the oilier conditions created by excess sebum. This overgrowth contributes to the skin conditions frequently seen in Parkinson’s patients, including seborrheic dermatitis and its characteristic flaky, reddish patches.3Dermatology Practical & Conceptual. Dermatological Manifestations and Sebum Composition in Parkinson’s Disease Whether the microbial shift itself contributes to the volatile signature, by metabolizing lipids into different breakdown products, is an open question that researchers have started to explore. The skin, in other words, is not a passive surface in Parkinson’s. It is an active site where the disease’s molecular machinery plays out in ways you can sometimes see, and apparently smell.