NEW ORLEANS – A test of smell identification seems to predict cognitive decline in elderly patients who score normally on memory tests.
Each 1-point decline on the 40-item smell test was associated with a 7% increase in the risk of cognitive decline over 2 years, Dr. Davangere P. Devanand reported at the annual meeting of the American Association for Geriatric Psychiatry.
Dr. Davangere Devanand
The results of his retrospective cohort analysis, combined with previous studies on smell and cognition, suggest that an inability to identify smells might be a very early sign of neuropathology related to Alzheimer’s disease.
“In the Alzheimer’s disease field, we are moving more and more to looking at people who are cognitively normal,” said Dr. Devanand, professor of clinical psychiatry in neurology at Columbia University, New York. “A memory test is great when people already have a memory deficit but not that good when people are scoring normal on cognitive measures. We should be looking at olfaction identification deficit as a potential indicator of incipient pathology that a cognitive battery simply may not pick up.”
Dr. Devanand and his colleagues have previously shown that the University of Pennsylvania Smell Identification Test (UPSIT) can predict conversion from mild cognitive impairment (MCI) to Alzheimer’s disease.
But screening for cognitive decline before any memory symptoms develop is a tough challenge, he said. The olfactory nerves might just be a literal window deep in the brain, where memory deficits begin.
Second-order neurons project from the olfactory bulb directly into the piriform cortex, amygdala, entorhinal cortex, and hippocampus. In Alzheimer’s patients, those neurons seem particularly susceptible to the very early development of neurofibrillary tangles, although they do not show evidence of amyloid plaques. These neuropathology findings might explain why olfactory memory is impaired very early in the disease process and is associated with MCI conversion, Dr. Devanand said.
To investigate these changes and their correlation with cognition, he looked at data from the ongoing Northern Manhattan Study. The study is following a large multiethnic cohort for stroke risk factors and stroke. All of the 1,037 participants completed the UPSIT as part of their baseline evaluation. The cognition study comprised 757 of these participants, who were followed for 2 years. The mean age at baseline was 80 years.
During the study period, 109 subjects transitioned to dementia; of those, 101 converted to Alzheimer’s disease. Lower UPSIT scores at baseline were associated with a significant risk of eventual progression to AD – almost a 10% increased risk for each point decline from age-adjusted norms.
Dr. Devanand then looked at the UPSIT as a predictor of cognitive decline among those subjects who were cognitively normal at baseline. He defined cognitive decline as a decline of 0.5 standard deviation in composite measure of memory, language, or visuospatial ability over 2 years.
In a logistic regression model that controlled for demographics, lower UPSIT scores were significantly associated with cognitive decline, with a 7% increase in risk for each point lost. “It wasn’t a big effect, but it is clearly there,” Dr. Devanand said.
In fact, he added, “Perhaps the most interesting finding was the negative one.” The smell test was a significantly better predictor of 2-year cognitive decline than was the Selective Reminding Test – a very accurate measure of episodic verbal memory. Adding the SRT to the UPSIT did not improve its predictive value.
“If someone is scoring normal or close to normal memory, the olfactory test may predict if that person will decline cognitively,” said Dr. Devanand, who indicated that he had no financial disclosures.
On Twitter @alz_gal