TORONTO – A scratch-and-sniff test that asks subjects to identify 40 odors and ranks olfaction is almost as powerful a predictor of Alzheimer’s disease as is a positive test for amyloid.
Low scores on the University of Pennsylvania Smell Identification Test (UPSIT) are linked to a thinning of the entorhinal cortex – the brain region where amyloid plaques are thought to first appear as Alzheimer’s disease takes hold, Seonjoo Lee, PhD, reported at the Alzheimer’s Association International Conference.
“The findings indirectly suggest that impairment in odor identification may precede thinning in the entorhinal cortex in the early clinical stage of AD,” she concluded.
According to William Kriesl, MD, poor UPSIT sores are also related to brain levels of amyloid beta and are almost as predictive of cognitive decline.
These sensory changes appear to be one of the earliest manifestations of Alzheimer’s disease, the researchers said at the Alzheimer’s Association International Conference. Their studies also suggest that the test has a place in the clinic as an easy and inexpensive screening tool for patients with memory complaints, Dr. Kreisl said at the Alzheimer’s Association International Conference.
The amyloid biomarker tests currently available are not suitable for wide dissemination. Amyloid brain scans are currently investigative; they are also invasive, expensive, and not covered by Medicare or any private insurance. Lumbar punctures are also invasive and expensive, and almost universally disliked by patients. Additionally, there is little consensus on how to interpret CSF amyloid levels.
“We need easy, noninvasive biomarkers that can be deployed in the clinic for patients who are concerned about their risk of memory decline,” said Dr. Kreisl of Columbia University Medical Center, New York. “Odor identification testing may provide to be a useful tool in helping physicians counsel patients who are concerned about this.”
His study concluded that the UPSIT predicted Alzheimer’s disease almost as well as invasive amyloid biomarkers. The scratch-and-sniff test asks subjects to identify 40 odors and ranks olfaction as normal, or mildly, moderately or severely impaired.
Dr. Kreisl examined the relationship between UPSIT and brain amyloid beta in 84 subjects, 58 of whom had mild cognitive impairment (MCI) at baseline. All of these subjects had either an amyloid brain scan or a lumbar puncture to measure amyloid in cerebrospinal fluid. They were followed for at least 6 months.
At follow-up, 67% of the group of participants showed cognitive decline. After correcting for age, gender, and education, patients who were amyloid-positive on imaging or in CSF were more than 7 times as likely to have experienced cognitive decline [Odds Ratio (OR) 7.3]. Overall, UPSIT score alone didn’t predict cognitive decline, Dr. Kreisl said. However, when it was imputed as a continuous variable, patients with a score of less than 35 on the 40-item test were four times more likely to show cognitive decline than those with a score of 35 or higher (OR 4).
In fact, these low UPSIT scores were much more common among amyloid-positive patients. Of the 38 patients who were positive for amyloid beta on either diagnostic test, 32 had an UPSIT score of less than 35 while six had a score of 35 or higher. Among the 46 amyloid-negative patients, 28 had low UPSIT scores and 18 had normal UPSIT scores.
Combining amyloid status and UPSIT in a single predictive model didn’t increase accuracy above that of either variable alone, which suggests olfactory dysfunction is not being completely driven by amyloid brain pathology.
“This makes sense because other factors like neurofibrillary tangle burden and other neurodegeneration are also involved in influencing how the UPSIT score predicts memory decline,” he said.
In a separate study, Dr. Lee examined the relationship of UPSIT performance to entorhinal cortical thickness in 397 cognitively normal subjects who were involved in the Washington Heights-Inwood Columbia Aging Project. These subjects took the UPSIT and had magnetic resonance brain imaging both at baseline and at 4 years’ follow-up. Over that time, 50 transitioned to dementia, and 49 of them were diagnosed with Alzheimer’s disease. Another 79 subjects experienced cognitive decline, which was defined as a decline of at least one standard deviation in the average of the three cognitive composite scores of memory, language and visuospatial domains.
In comparing the groups with and without dementia, Dr. Lee found significant differences in the follow-up UPSIT score (23 vs. 27) and entorhinal cortical thickness (2.9 vs. 3.1 mm).
One standard deviation in performance on the UPSIT score was associated with a significant 47% increase in the risk of dementia, while one standard deviation in entorhinal cortical thickness was associated with a 22% increase in the risk. However, Dr. Lee said, the interaction of entorhinal thickness and UPSIT score was only significant in the group of subjects who transitioned to dementia.