Background and objectives: Poor olfaction is common but underrecognized in older adults. This sensory deficit has broader health implications beyond being a prodromal symptom of neurodegeneration. Although biologically plausible, its cardiovascular health implications are unclear. Therefore, we aimed to investigate the associations of poor olfaction with incident stroke, coronary heart disease (CHD), and heart failure (HF), as well as subclinical cardiac biomarkers, by using two well-established community-dwelling cohorts of older adults in the US.Methods: In the Health Aging, and Body Composition (Health ABC) Study, we analysed data of 2,537 participants (aged 75.6±2.8 years) who completed a 12-item Brief-Smell Identification Test in 1999-2000. We defined good olfaction as a test score of 11-12, moderate olfaction as 9-10, and poor olfaction as ≤8. We followed at-risk participants from baseline until the date of the first cardiovascular outcome of interest, death, last contact, or the end of the 12-year follow-up, whichever occured first. We used the cause-specific Cox regression to estimate the associations of olfaction with incident stroke, CHD, and HF, respectively. Further, we leveraged data from the Atherosclerosis Risk in Community (ARIC) Study which was designed for cardiovascular health research to independently investigate the associations of olfaction with risks of stroke, CHD, and HF. Olfaction was assessed using the 12-item Sniffin’ Sticks odor identification test in 2011-2013 and defined categorically the same as in the Health ABC Study. We followed at-risk participants to the date of the first cardiovascular event of interest, death, last contact, or December 31, 2020, whichever came first. We used the discrete-time sub-distribution hazard model to estimate the marginal absolute risk of each outcome of interest across olfactory statuses and adjusted risk ratios (aRRs), accounting for competing risk of death and covariates. The cross-sectional associations of olfaction with subclinical HF markers were estimated using the quantile regression for N-terminal pro-B-type natriuretic peptides (NT-proBNP) and high-sensitive cardiac troponin T (hs-cTnT) and using the logistic regression for electrocardiography-defined structural heart disease. Results: In the Health ABC Study, we identified 353 incident CHD, 258 strokes, and 477 HF events during up to 12 years of follow-up. Poor olfaction was significantly associated with HF, but not with CHD or stroke. In the ARIC Study, among 5,799 participants who were free of stroke at baseline, we identified 332 incident stroke events (256 ischemic) during up to 9.6 years of follow-up. Compared with good olfaction, poor olfaction was robustly associated with higher stroke risk throughout the follow-up, albeit the association was modestly attenuated after 6 years. Among 5,142 participants free of CHD at baseline, we identified 280 incident CHD events during up to 9.6 years of follow-up. Poor olfaction was associated with a higher CHD risk during the first 6 years of follow-up, but not beyond. Among 5,217 participants without clinical HF at baseline, we identified 622 incident HF hospitalizations during up to 9.6 years of follow-up, including 212 HF with reduced ejection fraction (HFrEF), 250 HF with preserved EF, and 160 with unknown left ventricular EF. Compared with good olfaction, poor olfaction was associated with a modestly higher risk of HF for 8 years. The association was largely limited to HFrEF. Participants with poor olfaction had higher median levels of NT-proBNP and hs-cTnT, and higher odds of structural heart disease than those with good olfaction. Conclusions: Among community-based older adults in the US, we found preliminary evidence that poor olfaction assessed by a single smell test is associated with the risk of major adverse cardiovascular outcomes. The data from both cohorts are consistent for HF, supported by subclinical HF biomarkers. However, associations of olfaction with stroke and CHD were observed only in the ARIC Study. We suggest future studies be conducted to confirm our findings and investigate the underlying mechanisms.
Read
