Olfaction, the Olfactory ‘Stress’ Test and cognition in community dwelling elders

Purpose (Aims): To determine 1) if there is an association of the Olfactory ‘Stress’ Test (OST) or conventional olfaction with cognition; 2) if baseline olfactory scores and the OST predict a change in cognition (CC); 3) if the relationship between cognition and the OST (or conventional olfaction) is influenced by brain reserve or cognitive reserve measures; and 4) if performance on the OST (or conventional olfaction) is associated with subjective memory complaints. Methods: The OST study included two phases: I (baseline, mixed clinical trial and cross-sectional study) including 295 community-dwelling elderly Australians and II (follow-up on average of 4.018 years later). Olfaction was assessed using the University Of Pennsylvania Smell Identification Test (UPSIT). The OST result (atropine effect, AE) is the score on the second 20-odor set of the UPSIT (post-intranasal atropine) minus that on the first 20-odor set (pre-intranasal atropine (preup)). Conventional olfaction included preup and hyposmia rating scale (HRS). Cognitive measures included the Audio Recorded Cognitive Screen and Standardized Mini-Mental Status Examination; CC is (phase II minus phase I) cognitive scores. Main algorithms and/or statistical methods used: Objectives 1 and 2 – linear and logistic regression, propensity score analysis; objective 3 – linear regression; and objective 4 – linear and logistic regression. Other methods and/or algorithms used include K-Means clustering, K-Nearest Neighbor classification, naïve Bayes and decision tree analysis. Results: The higher the preup, but the lower the AE and HRS, the higher the cognition was. AE < 0 (cholin = 0) was associated with higher phase I and II OST study cognitive scores than AE ≥ 0 (cholin = 1). Education level modifies the association of cholin, preup-75 and modified UPSIT classes (MC) with cognition. Clusters of AE or preup with ARCS scores have stronger associations with cognition than either AE or preup alone. Preup and cholin are associated with a single question history of memory loss (SML). History of depression modifies the association of cholin with SML. Conclusion: Conventional olfaction and cholin are predictors of CC. The OST (through cholin) is a potential biomarker for underlying AD, cholinergic neurodegeneration and SML pathology.