Factors Associated with Inaccurate Self-Reporting of Olfactory Dysfunction in Older US Adults

Abstract

Self-reported olfactory function has poor sensitivity (i.e., people with measured olfactory dysfunction are unlikely to accurately report it). We aimed to identify factors associated with lack of awareness of smell dysfunction. Objective odor identification was evaluated using a validated 5-item test in respondents from the National Social Life, Health, and Aging Project, a representative sample of home-dwelling, US adults ages 57-85 (n = 1468). Self-reported olfaction was assessed with a 5-point Likert scale. Using multivariate logistic regression, we tested factors that might influence inaccuracy of self-reported olfaction, including age, gender, race/ethnicity, education, marital status, cognition, comorbidity, smoking, depression, anxiety, self-rated mental and physical health, and social activity. Among older US adults, 12.4% reported their sense of smell as fair or poor, while 22.0% had objective olfactory dysfunction (≤3 items correct out of 5). Among those with measured olfactory dysfunction, 74.2% did not recognize it; these individuals were more likely to be older, Black, never married, and to have worse cognitive function compared to individuals who recognized their dysfunction (P < 0.05, all). Individuals who lacked awareness of their olfactory dysfunction had the greatest cognitive impairment at 5-year follow-up, followed by individuals aware of their dysfunction and finally normosmics (P < 0.001). Older Americans with measured olfactory dysfunction are unlikely to report it, and those who lack awareness of this dysfunction have distinct demographic, social, and cognitive characteristics. Therefore, clinicians should objectively test patients. Individuals who lack awareness of their olfactory dysfunction have poor cognitive outcomes and should receive additional clinical scrutiny.

Keywords: aged; cognition; epidemiology; olfaction disorders; self report; smell.

Figure 1.

Prevalence of errors in self-reported olfaction (n = 1468).

Figure 2.

Estimated MoCA-SA score at 5-year follow-up is lowest in false negatives, higher in true positives, and highest in normosmics (n = 1324). MCI, mild cognitive impairment; MoCA-SA, Montreal Cognitive Assessment-Survey Adaptation; N, normosmics.

Figure 3.

Estimated percent with new dementia diagnosis at 5-year follow-up is highest for false negatives, followed by true positives, and finally normosmics (n = 1288).