Poor sense of smell may be an early warning for poor health in older age that goes beyond neurodegenerative diseases, according to findings from a prospective cohort study1 published in Annals of Internal Medicine.
While dementia or Parkinson’s disease explained only 22 per cent of the higher death risk, and weight loss six per cent, more than 70 per cent of the higher mortality associated with poor sense of smell is unexplained.
Many studies with five years or less of follow-up have shown an independent association between olfactory impairment and increased all-cause mortality among older adults. However, these studies did not examine whether there are differences by race or sex for the factors that might explain the relationship.
Researchers from Michigan State University in the US studied more than 2,000 adults aged 71 to 82 to determine the relationship between poor sense of smell and mortality and to investigate potential explanations for the connection. Participants completed a brief smell identification test (BSIT) of 12 common odours and were asked to identify each odour from one of four options. Each correct response was given a point.
Using the BSIT scores, the researchers classified the participants as having good, moderate, or poor sense of smell. The researchers found that poor olfaction was associated with 46 per cent higher mortality at year 10 and 30 per cent higher mortality at year 13 compared with good olfaction. The association was robust and could not be explained by measured confounders, such as demographic characteristics, lifestyle, and comorbid conditions.
The authors of an accompanying editorial from Johns Hopkins University School of Medicine in the US city of Baltimore note that many older adults are unaware that their sense of smell is declining and they are rarely screened for poor olfaction in clinical practice. The authors hope that this study will inform whether and how assessment of smell should be incorporated into routine exams for older adults.
Reference:
1 Abstract: http://annals.org/aim/article/doi/10.7326/M18-0775
