A Nomogram Model for Predicting Moderate to Severe OSA in Western China: A Retrospective Analysis.

OBJECTIVE

To develop an interpretable nomogram model for predicting moderate-to-severe obstructive sleep apnea (OSA).

METHODS

This retrospective study analyzed 11,030 OSA patients who underwent polysomnography between 2012 and 2020. Patients were randomly divided into a modeling cohort (n=7,721) and a validation cohort (n=3,309) in a 7:3 ratio. Based on disease severity, patients were further classified into mild and moderate-to-severe subgroups. General clinical data were collected. Independent risk factors were identified using multivariate logistic regression, and a nomogram prediction model was constructed using the R "rms" package.

RESULTS

Univariate analysis showed significant differences in age, BMI, neck circumference, gender, hypertension, diabetes, smoking status, alcohol consumption, and coronary heart disease between the mild and moderate-to-severe groups in both cohorts (all P < 0.05). Multivariate logistic regression identified age, BMI, neck circumference, male gender, hypertension, and diabetes as independent risk factors for moderate-to-severe OSA (all P < 0.05). The nomogram model demonstrated good fit (Hosmer-Lemeshow test: modeling cohort χ 2=10.754, P=0.216; validation cohort χ 2=8.383, P=0.397). Decision curve analysis confirmed the model's clinical utility within specific risk thresholds. The area under the ROC curve was 0.676 (95% CI: 0.664-0.688) for the modeling cohort and 0.613 (95% CI: 0.593-0.632) for the validation cohort, indicating strong predictive performance.

CONCLUSION

The nomogram model based on age, BMI, neck circumference, gender, hypertension, and diabetes provides a valuable tool for predicting and assessing moderate-to-severe OSA, offering important strategic guidance for clinical practice.