Analysis of the cause of retrolingual obstruction in patients with moderate-severe obstructive sleep apnea.

Purpose: To evaluate the percentage of obstructive sleep apnea (OSA) patients with retrolingual obstruction in all moderate-severe OSA patients and the proportions of different causes in all moderate-severe OSA patients with retrolingual obstruction and to discuss the accuracy of the Friedman tongue position (FTP) and retrolingual cross-sectional area (RCSA) in assessing the retrolingual obstruction.

Methods: Two hundred and twenty moderate-severe OSA patients were enrolled. After retrolingual obstruction was diagnosed, the percentage of OSA patients with retrolingual obstruction in all moderate-severe OSA patients was calculated. After that, the different causes of retrolingual obstruction were diagnosed based on different diagnostic criteria, and the proportions of different causes in all moderate-severe OSA patients with retrolingual obstruction were calculated. Finally, the correlations between FTP, RCSA, and apnea-hypopnea index after nasopharyngeal tube insertion (NPT-AHI) were analyzed, and the proportions of different causes of retrolingual obstruction based on different FTP and RCSA were observed.

Results: There were 128 patients with retrolingual obstruction, accounting for 58.2% of all moderate-severe OSA patients. In 128 patients with retrolingual obstruction, the proportions of glossoptosis (48.4%), palatal tonsil hypertrophy (28.1%), and lingual hypertrophy (8.6%) were relatively high. Both FTP and RCSA did not correlate with NPT-AHI. The proportion of lingual hypertrophy increased gradually with the increase of FTP and the proportions of glossoptosis in all FTP classifications were high. The patients with RCSA > 180 mm were mainly affected by glossoptosis and palatal tonsil hypertrophy, while patients with RCSA ≤ 180 mm were mainly affected by lingual hypertrophy.

Conclusion: The percentage of patients with retrolingual obstruction in all moderate-severe OSA patients is relatively high, mainly glossoptosis, palatal tonsil hypertrophy, and lingual hypertrophy. FTP classification and RCSA can only reflect the retrolingual anatomical stenosis, but cannot fully reflect the retrolingual functional stenosis, especially the presence of glossoptosis.