Etiology-Specific Olfactory Phenotypes for COVID-19 and Mild Cognitive Impairment

ABSTRACT

Introduction: Olfactory dysfunction is a common symptom associated with COVID-19 infection. While often transient, nearly 1 in 8 patients experience persistent dysfunction after initial infection resolution. Given the known association between impaired olfaction and mild cognitive impairment (MCI), this persistent COVID-19 olfactory dysfunction may impede early detection of cognitive decline. Method(s) Patients with confirmed COVID-19-associated hyposmia (n=73), MCI (n=58), and normal controls (n=86) were prospectively enrolled. Demographic data were collected alongside formal olfactory testing via AROMA (Affordable Rapid Olfaction Measurement Assay) at time of initial enrollment. MCI was assessed via MoCA (Montreal Cognitive Assessment). Multivariate logistic regressions were utilized to evaluate for associations between variables and etiology of olfactory dysfunction. Result(s) After controlling for age and gender, when compared against normal controls, the inability to smell licorice, cinnamon, and lemon at the lowest 3 concentrations increased odds of COVID-19 hyposmia by 10.8 (95% CI, 4.6-25.6), 5.7 (95% CI, 2.7-11.7), and 5.3 (95% CI, 2.6-10.8), respectively. While the inability to smell coffee (9.9 odds ratio [OR];95% CI, 2.02-48.1), eucalyptus (6.7 OR;95% CI, 2.2-20.0), and rose (4.0 OR;95% CI, 1.7-9.7) were associated with MCI, decreased ability to smell licorice, cinnamon, and lemon were not. When combined into a composite score and compared against controls, decreased detection of licorice, cinnamon, and lemon was associated with a 16.5 OR (95% CI, 6.6-41.3) for COVID-19 hyposmia. This composite score was not significantly associated with MCI (1.2 OR;95% CI, 0.6-2.2) and, as such, performed well at discriminating between COVID-19 and MCI patients (receiver operating characteristic area under the curve=0.76). Conclusion(s) Distinct patterns of impaired olfaction were noted for COVID-19. We show that this etiology-specific phenotype has good discriminative performance when differentiating from MCI-associated hyposmia, which may allow for continued utilization of olfactory screening for MCI even among those with previous COVID-19 infection.