Objective:

To examine olfaction following traumatic brain injury (TBI).

Background:

Longitudinal assessment of olfactory function is understudied in TBI. We hypothesized that olfaction varies by initial head CT, Glasgow Coma Scale (GCS) score, and smoking status and improves over time.

Design/Methods:

In this ongoing, prospective observational study, hospitalized individuals with TBI were recruited from a Level 1 trauma center. Demographic, clinical, and injury data was collected at enrollment. Objective olfactory function was assessed at 2-weeks and 6-months post-injury using Sniffin’ Sticks (16-item odor identification test). Individuals subjectively rated their sense of smell on a 5-point Likert scale. Change in olfaction over time was analyzed with Wilcoxon signed rank tests and group differences in olfaction score by demographic and clinical variables were examined with Mann-Whitney U-tests and Kruskal-Wallis tests. Percent agreement between subjective (poor/fair versus good/very good/excellent) and objective (defined using published age- and sex-specific thresholds for normosmia/anosmia) olfaction was examined.

Results:

Thirty participants (mean age 38.4 [SD=14.7] years, 70% male, 47% Black race, 41% current smokers) were included. Participants had a median GCS of 15 (IQR=14-15) and 53% had trauma-related abnormalities on head CT. Thirty-nine percent of individuals self-reported their smell poor/fair; 33% had anosmia on objective testing. At 2-weeks, median (IQR) olfaction score was 12 (10-13) and at 6-months, was 12.5 (10.5-13.5). In nine individuals with paired data, olfaction scores did not significantly change over time (difference=0.5; p=0.71). There were no differences in olfaction scores by sex (p=0.06), head CT status (p=0.12), GCS score (p=0.43), or smoking status (p=0.08). Agreement between subjective and objective olfaction was 36.7%.

Conclusions:

Olfactory dysfunction is present in a subset of individuals after TBI and remains stable over 6-months. This work also suggests that certain demographic factors may be important variables to consider and that subjective and objective olfaction measures are inconsistent.