Repeated Smell Testing May Help ID Those at Risk for Parkinson’s

Method appears to be efficient, cost-effective diagnostic tool

Lindsey Shapiro, PhD avatar

by Lindsey Shapiro, PhD |

Share this article:

Share article via email
An illustration of percent risk shows three individuals who each hold one part of an oversized percent sign.

People with persistent smell deficits are more likely to develop brain imaging abnormalities suggestive of Parkinson’s disease, or be diagnosed with the disease, than those whose deficits return to normal on a subsequent smell test, according to data from a Phase 2 study.

Deficits in two consecutive smell tests, separated by about a year, were associated with a 15 times greater risk of later developing such brain abnormalities, and an eightfold greater risk of a Parkinson’s diagnosis. However, the latter only came close to reaching statistical significance.

“Repeat testing may be an efficient and cost-effective strategy to identify at-risk patients for early diagnosis and disease modification studies,” the researchers wrote.

The study, “Serial olfactory testing for the diagnosis of prodromal Parkinson’s disease in the PARS study,” was published in Parkinsonism & Related Disorders.

Nerve cell loss begins years before the hallmark motor symptoms of Parkinson’s disease are evident, during the so-called “prodromal” period. Nonmotor signs of the disease, such as gastrointestinal or sleep problems, also may emerge in that period.

Recommended Reading
Scientists studying in detail data from a patient, with problem areas highlighted.

Rapid Skin Swab Test May Help Diagnose Parkinson’s

This provides a window of opportunity for identifying at-risk patients. If treatment is initiated during this prodromal phase, progression to full-blown Parkinson’s might be slowed or prevented.

Smell impairments, or hyposmia, are a common prodromal sign of Parkinson’s, with studies showing they may precede clinical diagnosis and motor symptoms by at least four years. This is consistent with the observation of lesions in areas of the brain involved in smell early in the disease.

“Olfactory [smell] testing is relatively low cost and easy to administer, and so has been proposed by many groups as a useful and scalable screening test for prodromal PD [Parkinson’s disease],” the researchers wrote.

However, smell can be affected by many other conditions, meaning that such testing is not very specific.  But when it’s associated with other causes, such as a respiratory virus, smell loss usually is temporary, while smell loss caused by a neurodegenerative disease would be expected to be long-lasting.

Therefore, repeated smell testing over time “may be a simple and effective way to increase specificity of olfactory testing with minimal effect on sensitivity,” the researchers wrote. A test’s sensitivity corresponds to its true-positive rate.

PARS study

The Phase 2 Parkinson Associated Risk Syndrome (PARS) study (NCT00387075), expected to recruit 3,000 people, was designed to investigate whether smell testing combined with dopamine transporter scan (DaTscan) imaging could be used to forecast Parkinson’s risk.

DaTscan is a widely used tool for assessing the loss of dopamine-producing nerve cells that is characteristic and highly specific to Parkinson’s. Dopamine is a major brain chemical messenger.

In the current analysis, the researchers assessed whether persistent hyposmia, evaluated by two consecutive smell tests, could predict a risk of developing DaTscan abnormalities or clinical Parkinson’s disease among PARS participants.

Participants, ages 50 or older, were recruited to the longitudinal study with an online screening questionnaire and an at-home, self-administered smell test. About half were first-degree relatives of Parkinson’s patients.

Participants were asked to undergo a second smell test, about 1.4 years after the first, and to complete up to 10 years of clinical and imaging evaluations.


Smell was assessed with the University of Pennsylvania Smell Identification Test, which is essentially a booklet of 40 scratch-and-sniff odors with a multiple-choice question to identify each smell.

A score at or below the 15th percentile for their age and sex meant a person was hyposmic. Such a score means that only 15% of the general population has lower smell scores than that person.

Ultimately, 186 hyposmic participants and 96 age- and sex-matched people with normal smell abilities were included in the analysis.

Results showed that among those initially deemed hyposmic, 133 (71.5%) continued to show hyposmia on a second test, while the remaining 53 (28.5%) reverted to a normal sense of smell (reverters).

Among persistently hyposmic participants, 29% went on to show abnormal DaTscan — suggestive of loss of dopamine-producing nerve cells — while 20% eventually were diagnosed with Parkinson’s. In contrast, no participant with normal smell at the study’s start, and only 2% of reverters, developed either condition.

Notably, 20 people had abnormal DaTscan scores at study’s start and an additional 20 developed such abnormalities during follow-up. Of these 40 people, all but one had persistent hyposmia.

Also, a clinician-determined emergence of Parkinson’s disease occurred in 27 people, 26 of whom had persistent hyposmia.

To analyze participants’ cumulative risk of developing DaTscan abnormalities or being diagnosed with Parkinson’s in the future, the researchers adjusted for potentially influential factors including age, sex, presence of first-degree relative with Parkinson’s, motor abilities, and gastrointestinal and sleep symptoms.

Significant differences

Results showed that people with persistent smell deficits were significantly more likely to develop abnormal DaTscans (by 44.1 times) and a clinical diagnosis of Parkinson’s (by 27.1 times) than those with normal smell and those who initially were hyposmic but reverted.

Compared with reverters alone, persistently hyposmic participants had a 15.2 times greater likelihood of showing DaTscan abnormalities in the future, and a 7.6 times greater chance of later being diagnosed with Parkinson’s. The latter association was only close to reaching statistical significance.

For participants who had normal DaTscans at the study’s start and later showed imaging or clinical signs of the disease, hyposmia was noted an average of 4.3 years before the deficit emerged, and 6.6 years before a clinical diagnosis of Parkinson’s. However, smell deficits could have pre-dated the initial assessment, the team noted.

Overall, “serial olfactory testing is associated with a higher relative risk of developing clinical or imaging evidence of [Parkinson’s], as compared to a single test,” the researchers wrote. “Further work with shorter intervals before repeat testing will be needed to define the optimal approach to serial olfactory assessment.”

Among the study’s limitations, the team noted about half of participants in PARS had a first-degree relative with Parkinson’s, which may limit generalizing the findings to the broader general population.