Intranasal insulin for COVID-19-related smell loss

Background: Quantitative (hyposmia and anosmia) and qualitative (phantosmia and parosmia) olfactory disorders are common consequences of COVID-19 infection found in more than 38% of patients even months after resolution of acute disease. SARS-CoV-2 has tropism for angiotensin-converting enzyme 2 (ACE2) in the respiratory system, suggesting that it is the mechanism of damage to the olfactory neuroepithelium and of involvement at the central nervous system. The olfactory bulb is the organ with the highest insulin uptake in the central nervous system. Insulin increases the production of Growth Factors (GF); therefore, in this study, the administration of intranasal insulin is proposed as a viable treatment for olfactory disturbances. The aim of this study was to obtain improvement in olfaction after 4 weeks of intranasal insulin administration in a group of patients presenting chronic olfactory disturbances secondary to COVID-19 infection, quantified using the Threshold, Discrimination, and Identification (TDI) score based on the Sniffin Sticks®. Methods: Experimental, longitudinal, prolective and prospective study of patients with a previous diagnosis of COVID-19 in the last 3-18 months and who persisted with anosmia or hyposmia. The sample size was calculated with "satulator". The intervention was performed from January to May 2022. Throughout four appointments, a baseline olfactory measurement was obtained using the TDI score based on the Sniffin Sticks® test. In the first three appointments, Gelfoam® cottonoids soaked in 40 IU of NPH insulin were placed on the nasal roof of each nostril for 15 min. Descriptive statistics, student's paired t test and a multiple linear regression were utilized to ascertain statistical significance of the outcome on the TDI score obtained on the fourth and final appointment. Results: 27 patients were included in the study. Table 1 summarizes the sample characteristics. The results exhibit that 93% of the sample had an improvement. The initial mean TDI score was 67% (63-71) compared to the final mean of 83% (80-86, p < 0.01). TDI subsection analysis is shown in Table 2. There was no significant difference in pre-intervention and post-intervention glucose measurements after the intranasal insulin administration.