Smell Test May Track Early Alzheimer’s Progression

Summary: Researchers at McGill University have developed a simple odor identification test that could help clinicians identify people at higher risk of Alzheimer’s disease and monitor early progression of the condition.

Source: McGill University.

Smell test shows promise as an early indicator of Alzheimer’s pathology

Alzheimer’s disease (AD) causes brain changes that can begin decades before memory problems become apparent, making early detection a critical goal for researchers and clinicians. A team at McGill University now reports that a straightforward smell identification test correlates with biological markers of AD in older adults who are at elevated risk. These findings suggest odor testing may serve as an affordable, noninvasive screening tool to help identify preclinical AD and monitor disease evolution.

“Although effective treatments for Alzheimer’s remain elusive, delaying symptom onset even by a few years would dramatically reduce the number and severity of cases,” said Dr. John Breitner, director of the Centre for Studies on Prevention of Alzheimer’s Disease at the Douglas Mental Health Research Centre, and one of the study’s authors. The research was published in the journal Neurology.

How the study was conducted

Nearly 300 cognitively normal participants, with an average age of 63 and a parental history of AD, completed a standardized multiple-choice scratch-and-sniff test that asked them to identify everyday scents—examples included bubble gum, gasoline and lemon. A subset of 101 volunteers also consented to lumbar punctures so researchers could measure Alzheimer’s-related proteins in cerebrospinal fluid (CSF), including total tau (t-tau), phosphorylated tau (P181-tau) and Aβ1-42 (beta-amyloid).

The investigators used the University of Pennsylvania Smell Identification Test to score odor identification (OI) performance. Cognitive status was assessed with the Repeatable Battery for the Assessment of Neuropsychological Status. Analyses adjusted for factors such as age, education, sex, APOE ε4 genotype and cognitive test scores.

Image shows a woman sniffing a flower.
Participants who had greater difficulty identifying odors also showed stronger biological indicators of Alzheimer’s disease in cerebrospinal fluid. Image for illustrative purposes.

Key findings

Participants with poorer performance on the smell identification test tended to show higher ratios of CSF t-tau and P181-tau relative to Aβ1-42—biomarker patterns commonly associated with Alzheimer’s pathology. Reduced odor identification was also linked to older age and lower cognitive scores in the overall sample. However, when analyses were limited to the subset who provided CSF and included the biomarker data, the apparent relationships with age and cognition were accounted for by the underlying AD pathology measured in CSF.

The study also found limited association between odor identification and CSF Aβ1-42 alone, except among APOE ε4 carriers with the lowest Aβ1-42 concentrations. These observations support the interpretation that diminished smell recognition reflects the degree of preclinical AD-related brain changes rather than serving as an independent marker of normal aging.

Clinical implications and cautions

The results indicate that a simple odor identification test could potentially provide information similar to more invasive and costly CSF analyses, helping to identify people who may already be developing AD pathology despite having no clear cognitive symptoms. Marie-Elyse Lafaille-Magnan, the study’s lead author, noted that brain regions involved in smell perception—the olfactory bulb and the entorhinal cortex—are among the earliest affected by Alzheimer’s disease, which helps explain the link between smell loss and preclinical pathology.

However, the researchers emphasize important caveats. Impaired odor identification can result from many medical conditions beyond Alzheimer’s, so smell testing should not replace established diagnostic procedures. Additional longitudinal research is required to determine how changes in odor identification over time relate to the progression from preclinical pathology to symptomatic AD. For now, odor testing represents a promising, low-cost tool to complement other screening and research approaches.

About this research

Funding: The study was supported by Pfizer Canada, the Canada Fund for Innovation, Fonds de recherche du Québec–Santé, the Levesque Foundation, and the Douglas Mental Health University Institute Foundation.

Source: McGill University

Abstract (study summary)

Objective: To evaluate odor identification (OI) as a marker of preclinical Alzheimer’s disease (AD) in cognitively normal older adults at increased risk for AD dementia.

Methods: Among 274 participants in the PREVENT-AD cohort who had a parental or multiple-sibling history of AD dementia, the researchers examined cross-sectional associations between OI and potential indicators of presymptomatic AD. One hundred one participants provided cerebrospinal fluid, allowing measurement of total tau (t-tau), phosphorylated tau (P181-tau), and β-amyloid (Aβ1-42). Models adjusted for age, cognitive performance, APOE ε4 carrier status, education and sex.

Results: Lower OI scores were associated with reduced cognitive performance, older age, and increased CSF ratios of t-tau and P181-tau to Aβ1-42 (all p < 0.02). When analyses were restricted to CSF donors and included biomarker measures, the associations of OI with age and cognitive test scores were largely explained by the presence of AD-related biomarkers. OI showed minimal association with CSF Aβ1-42 alone except among APOE ε4 carriers with the lowest Aβ1-42 levels.

Conclusions: In this cohort of healthy older individuals at increased genetic or familial risk for AD, odor identification appears to reflect the extent of preclinical AD pathology. Diminished OI may offer a practical, affordable biomarker for AD-related brain changes, although it should be used in conjunction with other assessments and biomarkers.

“Odor identification as a biomarker of preclinical AD in older adults at risk” by Marie-Elyse Lafaille-Magnan et al., published in Neurology (online July 25, 2017).