Sniff Test Fails to Predict Alzheimer's Disease Risk

Do we really know the best way to test for Alzheimer's disease? A grantee's new research shows that there's more work to be done.

    • June 18, 2012

Alzheimer’s disease affects 5.4 million Americans. As of 2012, the estimated cost of caring for people with Alzheimer’s, or other forms of dementia, is approximately $200 billion a year, according to the Alzheimer’s Association. As ever increasing numbers of baby boomers pass the golden age of 65, “those numbers are going to explode,” explains Gordon Sun, MD, a 2011-2013 Robert Wood Johnson Foundation (RWJF) Clinical Scholar. “By 2050,13 million people will be living with the disease.”

The statistics underscore the pressing need for an accessible, inexpensive, highly accurate method of testing for Alzheimer’s risk and prevalence. To advance science in the field, Sun, a general otolaryngologist, and his colleagues, decided to investigate the effectiveness of a test ear, nose and throat specialists frequently use to predict Alzheimer’s.

Revisiting a Common Assumption

“There are tests that are regularly used to diagnose smell disorders through the inability to identify common scents,” Sun explains. “Because people with Alzheimer’s have poor smell function, people in the medical profession thought that smell dysfunction tests could also indicate whether a person was going to develop Alzheimer’s in a few years.”

To test the hypothesis, Sun and an interdisciplinary team that included RWJF Clinical Scholar (2010-2012) James Burke, MD, a neurologist, Cyrus Raji, MD, PhD, a radiologist, and Mark MacEachern, a medical librarian, conducted a systematic review of approximately 1,200 articles, finding only two longitudinal studies and 30 observational, cross-sectional studies that used olfactory identification tests (smell tests) to evaluate study participants’ risk of developing Alzheimer’s. The results were reported in the May 2012 issue of The Laryngoscope.

“All of the studies reported a link between poor smell function and Alzheimer’s risk,” Sun says, “but closer analysis showed that there was no evidence of an association or causal relationship between having a poor sense of smell and developing Alzheimer’s, as previously thought. Two of the studies found that olfactory identification tests had some value in predicting Alzheimer’s risk, but only if the smell test was conducted as part of a battery of Alzheimer’s tests, including a full neurological evaluation.”

Preventing Mistakes in Testing

One of the key reasons for carefully investigating the predictive value of the olfactory identification test is that it could so easily be misused or misinterpreted, Sun says. “Three olfactory identification tests were used in the study. The most common one, the University of Pennsylvania test, has been used hundreds of thousands of times around the world. These are scratch and sniff tests using very common, everyday scents—fruits, grass or household items. There are even different versions for different cultures.”

Administering this simple test to people over 60, the prime Alzheimer’s disease risk group, is bound to turn up many cases of poor smell function for a number of reasons. “First, the elderly have a normally declining sense of smell. It’s part of aging. Second, they may have any of a number of other health conditions that may impair their ability to accurately interpret different scents,” Sun says.

The tests are also very easy to use. “While the tests are commonly performed by specialists at centers that test smell dysfunction and sinus disorders, the test is also very inexpensive, easy to obtain and administer and it requires very little training,” Sun explains. This would make it easy for the test to be performed in almost any setting.

But even within the profession, Sun notes, “people in my field should not be using this test, at this point, for this purpose. It’s not a useful predictor of Alzheimer’s risk. We need more longitudinal research that shows a real, predictive link between smell dysfunction and the disease before we can say that this works.”

“We do not want to use a test that can spread false fear or false hope,” adds Sun, who says that he advanced his systematic review skills during his first year of RWJF Clinical Scholar training. “I had great curiosity about this particular type of research before becoming a Clinical Scholar, but going forward, I hope to improve the quality of research in this field.”


The Robert Wood Johnson Foundation Clinical Scholars program advances the development of physicians who are leaders in transforming health care through positions in academic medicine, public health and other roles. The program trains clinicians in the program development and research methods that will enable them to find solutions to the many challenges posed by the health care system, community health and health services research.

Most Requested