A surprising shift in sniffing and odor perception could be a key sign of Parkinson’s disease

A new test that measures both how people describe odors and how they physically sniff them provides a way to distinguish Parkinson’s disease from other causes of smell loss. The findings suggest that people with Parkinson’s do not just lose their sense of smell, but experience a shift in how they perceive odors. This research was published in the journal npj Parkinson’s Disease.

A reduced ability to detect and identify odors often appears years or even decades before the motor symptoms of Parkinson’s disease emerge. Because of this connection, standard smell tests are sometimes used to support a diagnosis. However, these traditional tests measure overall performance, such as whether a person can detect or name a scent.

This creates a diagnostic challenge because general smell loss is common. People often lose their sense of smell due to healthy aging, viral infections, or sinus issues, making it difficult to know if poor test results point to Parkinson’s or something else entirely. Michal Andelman-Gur, a medical doctor and researcher who conducted the study while working with Noam Sobel’s Weizmann Olfaction Research Group at the Weizmann Institute of Science, wanted to address this diagnostic gap.

“As a medical doctor, I have always been interested in neurological diseases,” Andelman-Gur said. “Before joining the lab, I worked in the Neurology Department at Tel Aviv Medical Center, where I saw firsthand the impact these diseases can have on patients and their families.”

“When I joined the lab, I became especially interested in early detection, because I truly believe that detecting a disease as early as possible is key to developing better treatments and improving patients’ lives,” Andelman-Gur added.

Free daily newsletter

“Smell loss may seem like a minor, common symptom, but in some cases it can be an early warning sign of Parkinson’s disease – years before motor symptoms begin,” Andelman-Gur noted. “Today, smell tests can tell us that someone’s sense of smell is impaired, but not why.”

“Our test aims to fill this gap by identifying a specific pattern of smell changes associated with Parkinson’s disease,” she explained. “We hope it could eventually help detect Parkinson’s disease earlier and lead to earlier intervention and better treatment.”

Recent studies in vision and hearing suggest that Parkinson’s disease alters how sensory information is processed, rather than just dulling the senses. For instance, individuals might see colors differently or perceive rhythms in an altered way. The authors of the new study reasoned that a similar perceptual shift might happen with smell.

Andelman-Gur and her colleagues recruited 94 participants to test this idea. The sample included 33 individuals diagnosed with Parkinson’s disease, 33 healthy control participants matched for age and sex, and 28 participants experiencing smell loss from causes unrelated to Parkinson’s. The average age across the groups ranged from about 58 to 66 years old.

First, all participants completed a standard clinical smell test that measures the ability to detect, discriminate, and identify odors on a 48-point scale. The standard test successfully identified smell loss but failed to specify the cause. Both the Parkinson’s group and the non-Parkinson’s smell loss group performed poorly compared to healthy controls.

Out of a possible 48 points, healthy controls scored a median of 30.5. The Parkinson’s group scored a median of 16.5, and the group with other types of smell loss scored a median of 12.4. When controlling for variables like age, sex, and Parkinson’s medication, the test still could not tell the two groups with smell loss apart.

Next, the researchers asked the participants to smell jars containing either a pleasant lemon scent, an unpleasant fecal scent, or no odor at all. Participants rated the intensity and pleasantness of these jars on a scale from 1 to 100. People with non-Parkinson’s smell loss rated the intensity of the pleasant odor at a median of just 16, showing a steep drop compared to healthy controls, who gave it a 78.

Interestingly, individuals with Parkinson’s rated the intensity of the pleasant odor at a 78 as well, matching the healthy controls. When it came to pleasantness, however, the Parkinson’s group matched the group with other types of smell loss. Healthy controls rated the pleasant scent at a median of 81.5 out of 100.

In contrast, both the Parkinson’s group and the other smell loss group rated it at a median of 50. This separation in ratings indicates that people with Parkinson’s do not experience a generalized dulling of the senses, but rather a specific change in how they assign pleasantness to smells.

To build a more detailed profile, the researchers asked participants to rate 10 different chemical odors across 11 descriptive categories, such as sweet, spicy, and burnt, on a 100-point scale. This process created a mathematical representation of how similar or different various smells are to each individual.

When the researchers analyzed these profiles, they found that people with Parkinson’s shared a specific pattern of odor perception. Their ratings correlated more closely with other individuals who have Parkinson’s than with healthy controls or people with different types of smell loss.

The study also measured involuntary physical reactions to smells using a miniature nasal flow sensor. “One thing I find especially exciting is that this approach uses a very simple, natural behavior – how we sniff – to reveal information about the brain,” Andelman-Gur told PsyPost. “I hope this can eventually help us identify disease much earlier than we can today.”

Typically, humans take longer, deeper sniffs of pleasant odors and shorter sniffs of unpleasant ones. The researchers measured sniff duration for the pleasant and unpleasant scents in normalized units of time. Healthy controls inhaled the pleasant odor for 1.12 units and the unpleasant odor for 0.98 units, demonstrating the expected behavior.

Participants with non-Parkinson’s smell loss showed a similar pattern, sniffing the pleasant odor for 1.32 units and the unpleasant one for 1.17 units.

People with Parkinson’s disease behaved differently. They inhaled the pleasant odor for 1.18 units of time and the unpleasant odor for 1.20 units. This means they did not adjust their sniffing behavior based on how pleasant the odor was, and frequently took longer sniffs of the foul-smelling scent.

“At first, I was mainly interested in characterizing sniffing patterns and the natural tendency to reduce or briefly hold your breath in response to unpleasant odors,” Andelman-Gur said. “I did not expect patients with Parkinson’s disease to take longer sniffs when smelling unpleasant odors.”

“It was completely counterintuitive,” she added. “That was the moment I realized we might be seeing something really interesting, something that could reveal meaningful information about the disease itself.”

This abnormal sniffing pattern provided another independent marker of the disease. Finally, the researchers combined the subjective odor ratings with the physical sniff measurements to create a combined statistical model. This model analyzed the data and attempted to categorize each participant based solely on their perceptual and physical responses to smell.

The combined approach identified Parkinson’s disease from healthy controls with 89 percent accuracy, correctly categorizing 83 percent of actual Parkinson’s cases and 94 percent of the healthy cases. When distinguishing Parkinson’s from other types of smell loss, the model achieved 88 percent accuracy. After researchers analyzed a smaller, tightly matched group of 16 participants per condition to completely eliminate any slight age or sex differences, the accuracy rose to 94 percent.

The current model tested people who already had a confirmed diagnosis of Parkinson’s disease. Because smell loss usually begins long before motor symptoms appear, future studies need to evaluate if this type of test can identify the disease in its earliest, pre-symptomatic stages. Evaluating this tool as a predictive screening measure will require tracking healthy individuals over many years.

“The main limitation is that our cohorts are still relatively small, so the findings will need to be replicated in larger and more diverse groups,” Andelman-Gur noted.

The researchers plan to expand this work by testing individuals who possess specific genetic traits associated with the condition. Some people carry a mutation in a gene called LRRK2, which substantially increases their chances of developing Parkinson’s.

“Our next step is to test the method in groups at higher risk of developing Parkinson’s disease, such as LRRK2 mutation carriers,” Andelman-Gur said. “Ultimately, we hope to develop it into a simple screening tool that could be used more broadly in the general population.”

The study sample also included substantially more men than women. Men generally develop Parkinson’s disease at higher rates than women, but testing this specific odor profiling method in a more balanced demographic sample would help ensure the results apply evenly across the population. Finally, the researchers suggest exploring the relationship between altered breathing patterns and brain changes, raising the question of whether unusual sniffing behavior might contribute to sensory decline over time.

The study, “The World Smells Different in Parkinson’s Disease,” was authored by Michal M. Andelman-Gur, Sagit Shushan, Kobi Snitz, Gal Pinchasof, Danielle Honigstein, Lior Gorodisky, Aharon Ravia, Adi Ezra, Neomi Hezi, Tanya Gurevich, and Noam Sobel.