At one of the daily press conferences held by the Alzheimer’s Association during its AAIC14 Denmark conference in July, 2 studies were highlighted that were designed to better appreciate healthy cognitive aging and conversion from it to mild cognitive impairment (MCI) and to dementia due to probable Alzheimer disease. The studies by Growdon and colleagues1 from the Harvard Aging Brain Study and by Devanand and associates2 from Columbia University Medical Center used the same measure as part of their protocol: the “UPSIT,” shorthand for the University of Pennsylvania Smell Identification Test.
Olfaction as a Data Source
At first glance, olfaction might seem an odd source of data to collect for this type of work. But there is a sizable literature about the relation between olfactory recognition in aging and in the neurodegenerative disorders.
Growdon's study was cross-sectional in design. It was composed of 215 healthy elderly, aged 64 to 88 years, who were living independent lives without cognitive deficits. A subset of these participants were noted to show a lowered ability to recognize smells on the UPSIT. As a group, they also were found to show significant brain alterations relative to those with better smell recognition.
Poorer UPSIT performers showed reduced entorhinal cortex thickness, smaller hippocampal size, and evidence of greater amyloid deposits in their brains via PiB-enhanced PET imaging, all of which are potential biomarkers of neurodegenerative change. Among those individuals with greater amyloid levels, those who also had smaller entorhinal cortex thickness showed the worst UPSIT scores.
The authors proposed that using not only neuropsychological memory testing (a more conventional approach) but also the UPSIT might help identify individuals at greater risk for expressing Alzheimer disease at a later time and that this would be informing their study’s longitudinal component.
Devanand, in a longitudinal study built on earlier work (eg, Devanand et al3), reported the performances of over 750 healthy elderly with a mean age of 80.5 years. The sample was followed from baseline through 2 subsequent 2-year follow-ups to determine conversion to MCI and conversion to dementia due to probable Alzheimer disease.
Initial UPSIT performances showed 2 important predictions: (1) subsequent cognitive decline in those individuals cognitively asymptomatic at baseline and (2) conversion to dementia and Alzheimer disease. Interestingly, UPSIT predicted cognitive decline from normal to a greater degree than did baseline performance on a verbal selective-reminding task, a more traditional test used in such research. However, both smell recognition problems and verbal memory deficits were predictive of the transition to a formal diagnosis of dementia and probable Alzheimer disease.
What is UPSIT?
The UPSIT has been used in both research and clinical practice for over 30 years, yet it still has far from routine recognition in the neurological community. If at all, many physicians might think of smell testing as being limited to the peanut butter container that some colleagues might keep at hand for use during the cranial-nerve screening portion of a neurological examination.
So what is it? The UPSIT is a “scratch-and-sniff” manufactured instrument, created by Dr Richard Doty of the Department of Otorhinolaryngology at the Hospital of the University of Pennsylvania. It is composed of 40 items, presented 1 item at a time. That is, as a psychological test instrument, it challenges the person taking the task in the same manner 40 times to measure this single construct of smell recognition.
This structure provides for a wide and standardized range of variability of test performance to be examined and covers the broad normal range, borderline scores, and various levels of defective performance. For each trial, a microencapsulated stimulus is scratched and the odorant released, which the person inhales. Response is in the form of a 4-choice multiple choice, the choices being relatively distinct from one another.
The psychometric properties of the standardized UPSIT have been extensively studied and presented by Doty (eg, Doty, 19974) and, from the start, normative data have included the performances of elderly patients in a life-span scope. Reliability measured in terms of both internal consistency and test-retest reliability is above .90.5 Validational studies for the UPSIT are also described by Doty.5
The ABCs of Olfaction and Its Assessment
Consideration of UPSIT performances and olfaction in aging has recently been reviewed by Doty.6,7 To explore this literature and the potential role of testing smell recognition in older persons, one must keep in mind some basic facts about olfaction and its assessment.
First, olfactory recognition is not the same as olfactory threshold testing—identification and sensitivity need not be impaired to an equivalent degree in the very early stages of neurodegenerative disease.
Second, there are multiple reasons why someone might have an impairment or a reduction in smell identification.
Third, only severe deficits might be diagnostic of an anosmia, ie, the loss of olfactory functioning.
Fourth, smell identification ability changes over the course of healthy aging, absent any disease.
Fifth, smell identification deficits can be identified in Alzheimer disease but are not specific to that disease.
The ongoing research studies by the Growdon and Devanand groups, among other current efforts, continue to refine the appreciation of the health care and research benefits of including data about olfaction in our understanding of and the study methodologies for the neurodegenerative disorders. Reduced olfactory functioning can have an impact on the daily lives of patients. If smell identification has any predictive value for later Alzheimer disease and why this might be the case are questions with answers only now becoming clearer with studies like the 2 highlighted at AAIC14.
An Inclusion-criterion Role?
As pharmaceutical clinical trials look to recruit subjects at earlier times prior to the onset of Alzheimer disease, given the trial failures of disease-modifying investigational products in patients already having a diagnosis of probable Alzheimer, it is possible that the UPSIT could serve in an inclusion-criterion role. Identifying a subject pool who might be at greater risk for developing cognitive deficits and showing biomarker changes of the types that might be seen in patients destined for dementia of the Alzheimer type could make it easier to determine whether an investigational product has efficacy in treating that disease.
The UPSIT’s standardized administration makes it amenable for staff training for use at multiple clinical sites, administration times are short, and its availability in multiple languages (supported by cross-cultural research studies) makes it potentially suitable for consideration in CNS clinical trials examining investigation drugs to modify the course of Alzheimer disease.
1. Growdon M, et al. Abstract: Odor identification and Alzheimer’s disease biomarkers in clinically normal elderly. A paper presented at the Alzheimer’s Association International Conference 2014. Copenhagen, Denmark. Presented July 2014.
2. Devanand DP, et al. Abstract: Olfactory identification deficits predict cognitive decline and transition to Alzheimer’s disease in a multi-ethnic urban community. A paper presented at the Alzheimer’s Association International Conference 2014. Copenhagen, Denmark. Presented July 2014.
3. Devanand DP, Michaels-Marston KS, Liu X, et al. Olfactory deficits in patients with mild cognitive impairment predicts Alzheimer’s disease at follow-up. Am J Psychiatry. 2000;157:1399-1405.
4. Doty RL. Studies of human olfaction from the University of Pennsylvania Smell and Taste Center. Chem Senses. 1997;22:565-586.
5. Doty RL. The Smell Identification Test: Administration Manual. 3rd ed. Haddon Heights, NJ: Sensonics, Inc. 1995.
6. Doty RL. Smell and the degenerating brain. The Scientist. 2013;27;33-37.
7. Doty RL, Kamath V. The influences of age on olfaction: a review. Front Psychol. 2014;5:20.