Examining the association between CT-defined low tongue posture and CPAP adherence

CPAP therapy is the mainstay treatment for OSA, but adherence remains challenging. This study evaluated whether CT-defined low tongue posture (LTP), operationalized as a tongue-palate air gap on sagittal CT, is associated with CPAP adherence.

Retrospective analysis of 430 adults with OSA (AHI ≥ 20) who underwent PSG, CT imaging, and CPAP initiation at our sleep center. CT-defined LTP was assessed on mid-sagittal CT. CPAP adherence at 3 months was obtained from device-recorded data. Logistic regression identified factors associated with good adherence, defined as ≥ 70% and ≥ 50% of nights with ≥ 4 h of use.

CT-defined LTP was observed in 115 patients (26.7%). Patients with CT-defined LTP had lower average CPAP adherence (39.6% vs. 54.1%, p < 0.001). In multivariate analysis, CT-defined LTP was associated with lower odds of good adherence at both the 70% (adjusted OR 0.447, 95% CI 0.272-0.734, p = 0.001) and 50% (adjusted OR 0.496, 95% CI 0.319-0.772, p = 0.002) thresholds. The model showed limited specificity at the 70% threshold (22.5%) and improved balance at the 50% threshold (specificity 68.3%).

CT-defined LTP was associated with poorer CPAP adherence. Given its non-standardized CT definition and limited specificity at the 70% threshold, it should be interpreted as an adjunctive risk marker rather than a standalone screening tool.