RESUMO
PURPOSE: Hyposmia in childhood is poorly characterized. The "U-Sniff Test", validated for children with anosmia, can be used to objectify olfactory impairment but has not been used to distinguish between hyposmia and normosmia. Therefore, we investigated children with enlarged adenoids with respect to hyposmia, its correlation with adenoid size, and the sensitivity of questionnaires to predict olfactory impairment. METHODS: In a prospective comparison, olfaction was assessed by "U-Sniff Test" (score 0-12; <8 hyposmia) in 41 children (5-18 years) with adenoid hyperplasia and compared with 196 children without any respiratory affection (control) after exclusion of previous SARS-Cov2-infection from December 2020 to December 2021. ENT-related complaints were collected using a self-designed questionnaire. We were able to include 13 children in a follow-up examination to compare preoperative performance in the "U-Sniff Test" with postoperative outcome after adenoidectomy. STATISTICS: chi-square-test (p < 0.05), odds-ratio, Spearman's rho, ROC-, cluster analysis. RESULTS: Severe hyposmia was present in 36.6% of children with adenoid-hyperplasia compared to 3.1% of the control-group. Adenoid-children scored significantly more often between 8 and 10 points (58.5%) than the control (31.6%; p < 0.01). Adenoid size and olfactory performance correlate significantly (r: 0.83; CI -0.89 -0.72). Hyposmia in the adenoid group is characterized predominately by loss of the odors banana, butter and rose. None of children with hyposmia or parents reported impaired olfactory performance. Postoperatively, olfactory function improved significantly in 85% of cases (p 0.01, SD ± 1.71, Δ3.54points). CONCLUSION: Questionnaires are insufficient to detect hyposmia in this cohort. In contrast, the "U-Sniff Test" detects even reduced olfactory performance without reaching the cut-off value, which represents the majority of test results in the adenoid group. Therefore, we recommend the classification of moderate hyposmia (8-10 points) to be included for our study population.
