No Proof Alzheimer’s Smell Test Works

By Daniel J. DeNoon

Do you have Alzheimer’s disease? TV’s Dr. Oz and others say a simple smell test can tell.

But the test may only be confirming what you already know: that you are getting older and your sense of smell is fading.

Dozens of studies show that people with Alzheimer’s disease lose their ability to tell one common odor from another. And smell tests — the 40-item University of Pennsylvania Smell Identification Test (UPSIT) and the simpler three-item Pocket Smell Test (PST) — reliably identify people whose sense of smell is fading.

For these tests to be useful, however, they must predict who will go on to develop Alzheimer’s disease and who won’t.

There’s no proof smell tests can do that, finds a new study by University of Michigan researcher Gordon H. Sun, MD, and colleagues.

Sun’s team found 125 studies in the medical literature on the topic of smell testing and Alzheimer’s disease. Only two of them actually tested people and then followed them to see if they developed Alzheimer’s. Another 30 studies, of variable quality, compared the smell-test results of Alzheimer’s patients to normal adults or to patients with mild cognitive impairment.

One of the two long-term studies found that a 41- to 85-year-old person with mild cognitive impairment had about a 50-50 chance of developing Alzheimer’s if she or he flunked the smell test.

The other study found that people who went on to get Alzheimer’s tended to do a bit worse on the smell test, although this could have been a statistical blip. People who did and did not develop Alzheimer’s were equally likely not to have noticed a difference in their sense of smell.

The other 30 studies did find that people with Alzheimer’s disease were more likely to flunk a smell test than normal people or people with mild cognitive impairment. But the studies failed to correct for things that might have affected sense of smell — age, decreased hydration, thinning of the mucous lining in the nasal passages, and/or exposure to toxic agents.

“The predictive value of olfactory dysfunction [i.e., flunking a smell test] in the development of Alzheimer’s disease is uncertain,” Sun and colleagues conclude.

But they note that the door remains open to clinical trials designed to show whether smell tests can predict Alzheimer’s disease. To date, nobody has done such a study.

Sun’s study, sponsored by the Robert Wood Johnson Foundation and the U.S. Department of Veterans Affairs, appears in the current issue of the journal The Laryngoscope.