Patterns of olfactory impairment reflect underlying disease etiology

We aimed to determine whether the pattern of olfactory impairment seen in psychophysical testing reflects underlying disease etiology.

Retrospective cohort.

We performed a retrospective analysis of 1,226 patients from our tertiary referral center. Only hyposmic patients with the following conditions were included: postinfectious hyposmia, posttraumatic hyposmia, hyposmia secondary to sinonasal disease, and hyposmia secondary to Parkinson disease. Patients with anosmia were excluded.

Using a repeated measures analysis of variance (within subject factor "test": threshold [T], discrimination [D], identification [I]; between subject factor: "etiology") with posthoc Bonferroni corrected t tests, we found significant interaction between the factors "test" and "etiology" (F_6,2444 = 8.46, P < 0.001), indicating that different causes of hyposmia produce different patterns of olfactory loss with respect to the individual subtests T, D, and I . Specifically, patients with Parkinson disease performed relatively well in odor threshold testing, but poorly in odor identification and discrimination compared with the other etiology groups. Conversely, patients with sinonasal disease performed well in odor identification and discrimination but poorly in odor threshold. Patients with postinfectious and posttraumatic hyposmia performed relatively well in both thresholds and discrimination but poorly in identification. However, patients with posttraumatic hyposmia had globally reduced scores compared with the other groups.

This is the first study to comprehensively show that patterns of olfactory impairment reflect underlying disease etiology. We suggest that multicomponent olfactory testing should be performed, especially if there is diagnostic uncertainty. However, to clearly separate different patterns of olfactory loss to the various causes at an individual level, more work is needed.

4. Laryngoscope, 2016 127:291-295, 2017.