Anthropometric indices with high sensitivity for moderate-to-severe obstructive sleep Apnea: A sex-stratified diagnostic accuracy study based on the ESADA cohort

Obesity is a major risk factor for obstructive sleep apnea (OSA), but body mass index (BMI) may not adequately capture the regional adiposity most relevant to upper-airway collapse. Novel anthropometric indices may identify moderate-to-severe OSA (ms-OSA) more accurately and improve selection of uncomplicated patients for home sleep apnea testing (HSAT).

We conducted a retrospective diagnostic accuracy study using baseline anthropometric and polysomnographic data from the European Sleep Apnea Database. Patients with major comorbidities or other conditions precluding HSAT were excluded. The reference standard was apnea-hypopnea index ≥15 events/h. Receiver operating characteristic analyses were performed separately by sex, and areas under the curve (AUCs) of various anthropometric parameters were compared with BMI. The cohort was split into training and testing sets to derive and validate sex-specific cutoffs targeting at least 90% sensitivity.

Among 10,916 participants, 76.2% had ms-OSA. In males, BMI yielded an AUC of 0.715 (95% confidence intervals [CIs]: 0.701, 0.729), whereas the highest AUC was observed for the Deurenberg body fat percentage (BF%) formula: 0.741 (95% CIs: 0.728, 0.754). In females, BMI yielded an AUC of 0.681 (95% CIs: 0.662, 0.701), whereas the Lean BF% formula performed best: 0.739 (95% CIs: 0.721, 0.758). In the testing set, BF%_Deurenberg >25.94% in males yielded 90.5% sensitivity and 79.4% overall accuracy; BF%_Lean >43.15% in females yielded 88.0% sensitivity and 72.5% overall accuracy.

Several anthropometric indices outperformed BMI for identifying ms-OSA, with BF% estimators showing the best discrimination. Validated sex-specific cutoffs may support selection of uncomplicated patients for HSAT.