Self-Reported Hypersensitivity to Smells: Objective Evidence for Hyposmia, not Hyperosmia

To compare, using state-of-the-art psychophysical tests, the olfactory function of patients complaining and not complaining of olfactory hypersensitivity.

Retrospective cross-sectional.

The Smell and Taste Center at the University of Pennsylvania.

University of Pennsylvania Smell Identification Test (UPSIT) scores were obtained from 148 patients complaining of hyperosmia and 494 patients with no such complaints; detection threshold test scores were obtained from 77 and 483 patients of these respective groups. The effects of subject group, age, and sex on the test scores were assessed using analyses of variance. Categorical variables were evaluated by χ². Responses to items within a detailed intake questionnaire, for example, the Beck Depression Inventory (BDI-II), were also evaluated.

Unexpectedly, those complaining of hyperosmia had lower olfactory test scores than those with no such complaints (respective UPSIT means [95% confidence interval [CIs]] = 27.86 (26.85, 28.87) and 32.19 (31.67, 32.71); P < .001; respective threshold means (log vol/vol) = -4.49 (-4.89, -4.09) and -5.22 (-5.36, -5.06); P < .001). Remarkably, 70.95% of the self-identified hyperosmics exhibited mild to severe microsmia. The hyposmia complainers also exhibited elevated BDI scores (11.02 [9.53, 12.51] vs 7.58 [6.80, 8.34]).

When objectively tested, many patients who complain of hypersensitivity to odors are actually less sensitive to them. The basis of this phenomenon is unclear. It could reflect the presence of emotionally disturbing altered smell sensations, or one or more comorbidities, such as hypochondria or osmophobia. These findings point to the importance of objective testing of persons with complaints of chemosensory dysfunction and reiterate the inaccuracy of self-reports.