Use of coffee grounds to test olfaction for predicting cognitive dysfunction and decline.

Abstract Introduction Our objective was to determine whether non-standardized testing of olfaction may provide useful information for predicting cognitive dysfunction and decline in patients with neurobehavioral disorders. Methods We conducted cross-sectional and longitudinal analyses of 82 patients who presented to a Memory Clinic with a chief complaint of cognitive deficits using non-standardized odor identification testing (nSOIT). Each patient was classified as having intact or impaired olfaction based on the ability to identify and name the odor of coffee grounds. The cross-sectional study used Student's t-test to examine whether nSOIT results were related to cognitive dysfunction as approximated by Montreal Cognitive Assessment (MoCA) scores. The longitudinal study used mixed effects multiple regression with an interaction term to investigate whether nSOIT results were predictive of cognitive decline over a period of follow-up testing (0.4 to 4.0 years [mean 1.4, SD 0.8]) to compare patients who exhibited cognitive decline over the evaluation period (decliners) and those who did not (non-decliners). Results Analysis of the initial use of nSOIT in the cross-sectional study demonstrated no association between nSOIT performance and objective cognitive dysfunction. In the longitudinal study, impairment in follow-up nSOIT testing was found to be a sensitive but nonspecific predictor of cognitive decline. Conclusion These results suggest that routine olfactory testing may serve as a convenient and readily available method that can be used by clinicians to better predict cognitive dysfunction and decline in patients with a variety of neurobehavioral disorders.