Relationship Between Poor Olfaction and Mortality Among Community-Dwelling Older Adults: A Cohort Study

Abstract

Background: Poor olfaction is common among older adults and has been linked to higher mortality. However, most studies have had a relatively short follow-up and have not explored potential explanations.

Objective: To assess poor olfaction in relation to mortality in older adults and to investigate potential explanations.

Design: Community-based prospective cohort study.

Setting: 2 U.S. communities.

Participants: 2289 adults aged 71 to 82 years at baseline (37.7% black persons and 51.9% women).

Measurements: Brief Smell Identification Test in 1999 or 2000 (baseline) and all-cause and cause-specific mortality at 3, 5, 10, and 13 years after baseline.

Results: During follow-up, 1211 participants died by year 13. Compared with participants with good olfaction, those with poor olfaction had a 46% higher cumulative risk for death at year 10 (risk ratio, 1.46 [95% CI, 1.27 to 1.67]) and a 30% higher risk at year 13 (risk ratio, 1.30 [CI, 1.18 to 1.42]). Similar associations were found in men and women and in white and black persons. However, the association was evident among participants who reported excellent to good health at baseline (for example, 10-year mortality risk ratio, 1.62 [CI, 1.37 to 1.90]) but not among those who reported fair to poor health (10-year mortality risk ratio, 1.06 [CI, 0.82 to 1.37]). In analyses of cause-specific mortality, poor olfaction was associated with higher mortality from neurodegenerative and cardiovascular diseases. Mediation analyses showed that neurodegenerative diseases explained 22% and weight loss explained 6% of the higher 10-year mortality among participants with poor olfaction.

Limitation: No data were collected on change in olfaction and its relationship to mortality.

Conclusion: Poor olfaction is associated with higher long-term mortality among older adults, particularly those with excellent to good health at baseline. Neurodegenerative diseases and weight loss explain only part of the increased mortality.

Primary funding source: National Institutes of Health and Michigan State University.

Appendix Figure 1.

Flow chart of analytic sample.

Appendix Figure 2.

Standardized survival of cause-specific death among different olfactory function groups (n = 2289).

Figure 1.

Directed acyclic graph for the mediation analysis. The bold arrow represents the direct effect of olfaction, the dotted arrows represent the natural indirect effect through M_1t, and the dashed arrows represent the partial indirect effect through M_2t. A = olfaction; M_1t = dementia or Parkinson disease; M_2t = average annual weight loss ≥2%; t = year (3, 5, 10, or 13); U = unmeasured confounding; X and W = baseline confounders; Y_t = all-cause mortality for year t.

Figure 2.

Difference in standardized survival of cause-specific death comparing good versus poor or moderate olfaction groups (n = 2289).