Six years before her husband was diagnosed with Parkinson’s disease, a progressive neurodegenerative disorder marked by tremors and movement difficulties, Joy Milne detected a change in his scent. She later linked the subtle, musky odor to the disease when she joined the charity Parkinson’s UK and met others with the same, distinct smell. Being one of the most common age-related disorders, Parkinson’s affects an estimated seven million to 10 million people worldwide. Although there is currently no definitive diagnostic test, researchers hope that this newly found olfactory signature will lead help create one.

Milne, a super-smeller from Perth, Scotland, wanted to share her ability with researchers. So when Tilo Kunath, a neuroscientist at the University of Edinburgh, gave a talk during a Parkinson’s UK event in 2012, she raised her hand during the Q&A session and claimed she was able to smell the disease. “I didn’t take her seriously at first,” Kunath says. “I said, ‘No, I never heard of that, next question please.’”

But months later Kunath shared this anecdote with a colleague and received a surprising response. “She told me that that lady wasn’t wrong and that I should find her,” Kunath says. Once the researchers found Milne, they tested her claim by having her sniff 12 T-shirts: six that belonged to people with Parkinson’s and six from healthy individuals. Milne correctly identified 11 out of 12, but miscategorized one of the non-Parkinson’s T-shirts in the disease category. It turned out, however, she was not wrong at all—that person would be diagnosed with Parkinson’s less than a year later.

Milne may not be the only person with this ability. Since her story was revealed on BBC late last week Tilo and his collaborator, Perdita Barran, a chemist at the University of Manchester, have received e-mails from around 20 people who claim they also have this ability. Kunath and Barran hope to test these individuals to further validate the Parkinson’s scent. In the meantime they are analyzing the sebum samples to identify its source.

What’s that smell?

When the researchers asked Milne about the source of the odor, she pointed to the shirt collars. This was surprising, says Kunath, as they were initially expecting the scent to originate from the armpits, where sweat glands are abundant. Identifying the neck area provided an important clue—sebaceous glands, which secrete an oily, waxy substance called sebum, are concentrated around your back and T-zone area—the forehead, nose and chin.

Once they pinpointed sebum as the source, a paper trail began to emerge. It turned out researchers had identified changes in sebum production on Parkinson’s patients’ skin as early as the 1920s. This leads to parkinsonian patients’ skin having a waxy appearance, a condition that dermatologists had long documented, but neurologists largely ignored, as a classic sign of the disease.

Other studies have found that alpha-synuclein, the telltale clumps of protein that appear in the brains of those with Parkinson’s, also accumulate on the skin. This may also be a potential cause of the change in scent. Another possibility relates to dysfunction in the autonomic nervous system that occurs before motor symptoms appear. According to Kunath, these disturbances may affect the way the sebaceous glands are functioning. At this point, however, all potential mechanisms behind the odoriferous changes are merely guesses.

Chemical signatures

Barran is responsible for the technical side of things. Her group is working to identify the molecular signature responsible for the Parkinson’s scent. Initial analysis with mass spectrometry, a technique used in chemistry to identify the amount and type of constituents in a sample, has revealed around 9,000 different molecules from the skin secretions. And not all these originate from humans—some belong to bacteria that live on the skin, pointing to yet another possible source of the scent: a certain bacteria that likes to colonize the skin of people with Parkinson’s.

Further tests are needed, and not just comparing those with and without the disease. Parkinson’s involves a collection of symptoms that do not always appear in the same way between individuals or through the course of the disorder. One of the researchers’ next steps is to compare individuals at the early and late stages of the disease. “We don’t yet know how detailed the metabolic signature will be but we certainly do know that diseases can be graded, and a different and a more powerful test would be able to distinguish the different types of [Parkinson’s] and how it might progress,” Barran says.

New tools for diagnosis

What would this test look like? According to Barran, either a dipstick biosensor—which would work something like a pregnancy test, indicating whether specific biomarkers are present or absent—or a portable version of a mass spectrometer, which could not only identity what is there, but how much. The latter would be especially useful, given anecdotal evidence that biomarkers change with the progression of a disease. “With my husband, I found when his symptoms worsened, the smell became stronger, and when he was balanced with his pills, the smell was less,” Milne says.

Other researchers have identified signatures of Parkinson’s in breath. Hossam Haick, a chemical engineer at the Technion–Israel Institute of Technology, recently developed a breath test to detect Parkinson’s. A study by his group in 2013 found that it effectively classified parkinsonian versus healthy states with 78 percent accuracy. Haick did not expect to work with Parkinson’s either. But although he was testing breath samples for lung cancer, one of his participants who showed no signs of cancer insisted that something was wrong with him. It turned out that person had Parkinson’s, and they were able to identify a unique molecular signature in his breath. “We have now tested the device on more than 600 people with Parkinson’s disease,” Haick says. “The device is ready but it still needs to be commercialized.”

Whether there are similar molecules in Parkinson’s patients’ sebum and breath is unclear, and according to researchers, unlikely. The cause of the changes in breath also remains a mystery, and unlike sebum, there are no documented differences in the respiratory system. “There’s no evidence that the air pathways of the lungs of people with Parkinson’s change whereas there’s a lot of evidence that the skin changes,” Barran says.

Another, potentially faster method for developing a diagnostic tool is training so-called detection dogs. Dogs have a much keener sense of smell than humans do and have been used to sniff out explosives and drugs and even differentiate between identical twins. Detection dogs are also used as an early-detection sensor for some forms of cancer and diabetes. “We’ve been in discussion with Medical Detection Dogs, a charity that uses dogs to detect cancer,” Kunath says. They hope to eventually train dogs to smell Parkinson’s, which would be ideal as a fast-track diagnostic tool. Dogs do not need to know the exact molecular signature of a disease and are extremely accurate at diagnostics. For example, dogs are currently more accurate than the prostate-specific antigen (PSA) test, the primary screening method for prostate cancer.

Hope for sufferers

Tests using skin swabs, breath analysis and detection dogs are attractive because they are noninvasive. To date, no early diagnostic tools for Parkinson’s exist. The forthcoming breath test and future sebum tests may help fill this gap. Detecting the disease early can make a world of difference, especially with the emergence of new drugs that may be effective in slowing progression if administered during the beginning stages of the disease.

“My husband suffered from the disease for 21 years after his diagnosis but he had it many years before that,” Milne says. “I would like to see that people don’t suffer the way he suffered.” Milne’s husband died last June.