Olfaction and diabetes among older adults

Both poor olfaction and diabetes are common in older adults. It is biologically plausible that they may be related and interact to affect the health of older adults. We examined the association between poor olfaction and diabetes and their joint associations with mortality among 2,416 older adults from the Health, Aging, and Body Composition Study. Olfaction was assessed at year 3 (1999 to 2000) using the Brief Smell Identification Test (B-SIT). We used year 4 (2000 to 2001) as the study baseline and followed participants to year 11 (2007 to 2008) to identify incident diabetes and year 14 (2010 to 2011) to assess mortality. We used logistic regression to analyze the association of poor olfaction with prevalent diabetes and Cox proportional hazard models to assess its relationship to incident diabetes and its joint association with diabetes on mortality. Of the 2,416 participants, 611 (25.3%) had diabetes at baseline and 138 (7.6%) developed incident diabetes during 6.4 ± 1.7 yr of follow-up. Compared to those with good olfaction, the odds ratio of prevalent diabetes was 1.11 (95% confidence interval/CI: 0.87 to 1.42) for those with poor olfaction, and the corresponding hazard ratio (HR) for incident diabetes was 1.01 (95%CI: 0.66 to 1.57). During 8.2 ± 2.8 yr of follow-up, 1007 (41.7%) participants died. Compared with participants without poor olfaction and diabetes, those with both were twice likely to die during the follow-up (HR = 2.16, 95%CI: 1.71 to 2.73). However, we found no evidence for synergistic interaction (P = 0.97). In conclusion, poor olfaction is not associated with the risk of diabetes, and these two conditions independently predict mortality in older adults.