Olfactory deficits predict cognitive decline and Alzheimer dementia in an urban community

Abstract

Objective: To determine the predictive utility of baseline odor identification deficits for future cognitive decline and the diagnosis of Alzheimer disease (AD) dementia.

Methods: In a multiethnic community cohort in North Manhattan, NY, 1,037 participants without dementia were evaluated with the 40-item University of Pennsylvania Smell Identification Test (UPSIT). In 757 participants, follow-up occurred at 2 years and 4 years.

Results: In logistic regression analyses, lower baseline UPSIT scores were associated with cognitive decline (relative risk 1.067 per point interval; 95% confidence interval [CI] 1.040, 1.095; p < 0.0001), and remained significant (relative risk 1.065 per point interval; 95% CI 1.034, 1.095; p < 0.0001) after including covariates. UPSIT, but not Selective Reminding Test-total immediate recall, predicted cognitive decline in participants without baseline cognitive impairment. During follow-up, 101 participants transitioned to AD dementia. In discrete time survival analyses, lower baseline UPSIT scores were associated with transition to AD dementia (hazard ratio 1.099 per point interval; 95% CI 1.067, 1.131; p < 0.0001), and remained highly significant (hazard ratio 1.072 per point interval; 95% CI 1.036, 1.109; p < 0.0001) after including demographic, cognitive, and functional covariates.

Conclusions: Impairment in odor identification was superior to deficits in verbal episodic memory in predicting cognitive decline in cognitively intact participants. The findings support the cross-cultural use of a relatively inexpensive odor identification test as an early biomarker of cognitive decline and AD dementia. Such testing may have the potential to select/stratify patients in treatment trials of cognitively impaired patients or prevention trials in cognitively intact individuals.

Figure 1. Baseline UPSIT quartile scores in patients who transitioned to AD dementia during follow-up

In 757 participants without dementia who were followed, percent transitioning to AD classified by baseline UPSIT quartile scores. Q1 represents the quartile with the lowest UPSIT scores (worst quartile test performance) and Q4 represents the quartile with the highest UPSIT scores (best quartile test performance) that was used as the reference group in the analyses of other quartiles. AD = Alzheimer disease; UPSIT = University of Pennsylvania Smell Identification Test.

Figure 2. Receiver operating characteristic curves for odor identification and episodic verbal memory in predicting cognitive decline

Cognitive decline. In the entire follow-up sample, 732 of 757 participants (25 missing) provided data for cognitive decline. Receiver operating characteristic curves for cognitive decline for the 40-item UPSIT, SRT-TR, and UPSIT plus SRT-TR. AUC: UPSIT = 0.638, SD 0.024; SRT-TR = 0.513, SD 0.025; UPSIT + SRT-TR = 0.638, SD 0.024. Comparison of AUCs: UPSIT > SRT-TR (p < 0.001), (UPSIT + SRT-TR) > SRT-TR (p < 0.001). AUC = area under the curve; SRT-TR = Selective Reminding Test–total immediate recall; UPSIT = University of Pennsylvania Smell Identification Test.

Figure 3. Receiver operating characteristic curves for odor identification and episodic verbal memory in predicting AD dementia

Transition to AD. In the entire follow-up sample (n = 757), receiver operating characteristic curves for transition to AD for the 40-item UPSIT, SRT-TR, and UPSIT plus SRT-TR. AUC: UPSIT = 0.65, SRT-TR = 0.73, UPSIT + SRT-TR = 0.77. Comparison of AUCs: SRT-TR > UPSIT (p = 0.02), (UPSIT + SRT-TR) > SRT-TR (p = 0.01). AD = Alzheimer disease; AUC = area under the curve; SRT-TR = Selective Reminding Test–total immediate recall; UPSIT = University of Pennsylvania Smell Identification Test.