Scratch-and-Sniff Test Beats Imaging in Predicting Cognitive Decline

Pauline Anderson

August 03, 2016

TORONTO — A simple, inexpensive odor identification test is better than an MRI measurement of entorhinal cortex thickness in predicting cognitive decline, a new study suggests.

However, although the University of Pennsylvania Smell Identification Test (UPSIT) used in the study provides useful information, it's not a diagnostic tool by itself, said study investigator Davangere Devanand, MBBS, professor of psychiatry and neurology, Columbia University Medical Center, New York, New York.

"Clearly, odor identification impairment is an early sign of Alzheimer's disease and can be used to supplement a diagnostic workup," said Dr Devanand.

"The caution I would add is that it's not super specific because it can be affected by so many other things," including having a cold, being a heavy smoker, or having schizophrenia, Parkinson's disease (PD) or Lewy body dementia," he said.

The research was presented here at the Alzheimer's Association International Conference (AAIC) 2016.

Scratch-and-Sniff Test

The UPSIT involves scratching a surface, sniffing the odor that's released, and identifying it from a multiple-choice list. The test is scored from 0 (no correct answer) to 40 (all answers correct).

A low score indicates a decreased ability to correctly identify odors.

Dr Devanand emphasized that loss of odor identification, which is based on memory, is not the same as an impaired sense of smell. For the UPSIT, "you have to know what the odor is and put a name to it."

He added that the sense of smell does get impaired in AD, but not until "much later in the disease."

Neuropathology in the olfactory system occurs in the early stages of AD. Both the entorhinal cortex and hippocampus are known to be atrophied in AD, according to Dr Devanand.

An olfactory deficit is primarily linked to tau pathology; it's only later on in the disease that amyloid plaques accumulate.

"It looks as if an olfactory deficit is really a feature of tau pathology, at least certainly early on," said Dr Devanand.

Better AD Predictor

The study included 397 participants from Manhattan, New York, who had undergone MRI and UPSIT and had no dementia at baseline. They had a mean age of 79.85 years, entorhinal thickness of 3.07 mm, and UPSIT score of 27.09.

This means, said Dr Devanand, that odor identification was somewhat impaired in these people, "some of whom we considered normal or with very subtle cognitive problems."

Researchers then followed these participants for 4 years. They defined cognitive decline as a decrease in the average of three cognitive composite scores (memory, language, and visuospatial domains) of 1 standard deviation (SD) or greater at 4 years.

During the follow-up period, 50 participants (12.6%) developed dementia, and 19.8% were classified as having cognitive decline.

Researchers tested both average entorhinal cortical thickness and UPSIT score, as well as their interaction, as main predictors. They adjusted for age, sex, years of education, functional status, intracranial volume, and language (English or Spanish) in which the test was administered.

The analysis found that UPSIT (hazard ratio [HR], 1.74 per 1 SD; P < .001) and entorhinal cortical thickness (HR, 1.50 per 1 SD; P = .012) predicted transition to dementia.

UPSIT predicted cognitive decline (odds ratio, 1.49 per 1 SD; P = .005), but entorhinal cortical thickness was only at a trend level in this prediction (P = .154).

The study also showed that entorhinal cortical thickness was only significantly associated with UPIST among participants who transitioned to dementia (P < .0001) — not among those with cognitive decline.

"So odor identification seems to be a better predictor of who will get AD," said Dr Devanand.

The results also support the view that odor identification deficits are linked to neurodegenerative changes in the entorhinal cortex during the progression of AD.

Important Caveat

A "caveat" to this study is that while the UPSIT test is "very precise — you either get the odor identified or you don't — it's difficult to measure the entorhinal cortex and there is a lot of variability in its measurements," said Dr Devanand.

UPSIT, he stressed, improves diagnostic accuracy but is not by itself diagnostic. The reason for this, he said, is that there are many false-positives in, for example, heavy smokers, those with a respiratory infection, and those with certain other conditions.

"So it is useful where you already know that a patient doesn't have one of those other problems," said Dr Devanand.

When these are ruled out and a case is still in doubt, clinicians have the option of ordering neuroimaging, but this is expensive.

"Or they can do this odor identification test, which costs about $20 and gives them some idea as to what's happening," said Dr Devanand.

Interestingly, in the same study sample, among participants who scored very well on the UPSIT (score of more than 35 out of 40), "almost nobody went on to get AD and their average age is 80 years old," said Dr Devanand.

"So if you score really well on this test, even if you have some memory impairment, you are probably okay."

Odor identification varies across the ages. It improves until about age 20 to 30 years, then gradually declines but drops off more rapidly after about age 70 years.

As well as the full UPSIT, which takes about 20 minutes to complete, there is a shorter version with 12 items that takes only about 6 to 7 minutes.

Tight Link

Commenting on the study, Suzanne Craft, PhD, Wake Forest School of Medicine, Winston-Salem, North Carolina, and a member of the Alzheimer's Association Medical & Scientific Advisory Council, said it was "very interesting."

The "new part" of this research, she said, is linking odor identification "very tightly" to a "very specific aspect" of AD pathology, cortical thinning, said Dr Craft, who chaired the press briefing where the study was presented.

The research highlights the importance of the UPSIT test as a screening tool before initiating further workup, she said.

But she pointed out that odor identification is impaired in several neurodegenerative diseases, not just AD.

"So perhaps it's the specificity that we need to figure out," she said. "It may be that the test is a general brain health indictor, which is still very valuable."

The research was supported by grants from the National Institute on Aging. Dr Devanand has disclosed no relevant financial relationships.

Alzheimer's Association International Conference (AAIC) 2016. Abstract 03-12-02. Presented July 26, 2016.

Follow the AAIC Twitter feed using #AAIC16. For more Medscape Neurology news, join us on Facebook and Twitter


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.