STOPBANG vs. STOPBWANG: Comparative predictive performance for obstructive sleep apnoea severity and oxygen desaturation in an Indian cohort.

BACKGROUND

Obstructive sleep apnoea (OSA) is underdiagnosed in India, necessitating context-specific screening tools. While STOPBANG (incorporating neck circumference) is widely used, waist circumference (reflecting visceral adiposity) may better predict OSA severity and oxygen desaturation in populations with the high abdominal obesity. This study compared STOPBANG, STOPBAWG (waist-modified STOPBANG), and STOPBWANG (STOPBANG + waist) for predicting OSA severity and hypoxemia in an Indian cohort.

METHODS

Retrospective analysis of 662 Indian adults undergoing polysomnography (PSG) for suspected OSA. Anthropometrics (neck/waist circumference, Body Mass Index [BMI]), Epworth Sleepiness Scale (ESS), and desaturation indices (DI, T90) were analysed. Predictive performance of STOPBANG, STOPBAWG, and STOPBWANG for OSA severity (Apnoea-Hypopnea Index [AHI-based]) was evaluated using logistic regression (Area Under the Curve [AUC] and Akaike Information Criterion [AIC]). Linear regression assessed associations between anthropometrics and oxygen desaturation.

RESULTS

Waist circumference >36 inches showed the strongest association with severe OSA (93.5% of severe OSA patients, P < 0.001) and oxygen desaturation ( R² =0.093 for T90). STOPBANG outperformed STOPBAWG (AUC: 0.849 vs. 0.819; AIC: 164.2 vs. 167.66). STOPBWANG (STOPBANG + waist) achieved the highest predictive accuracy (AUC: 0.88, AIC: 148.94). ESS correlated moderately with both tools (STOPBANG: r = 0.343; STOPBAWG: r = 0.316), but did not enhance model performance.

CONCLUSION

STOPBANG remains superior to waist-modified STOPBAWG for OSA screening in Indians. However, integrating waist circumference into STOPBANG (STOPBWANG) significantly improves risk stratification, reflecting visceral adiposity's role in hypoxemia. Clinically, waist circumference should complement-not replace-neck circumference in OSA screening tools for abdominal obesity-predominant populations.