The impact of body roundness index (BRI) on obstructive sleep apnea (OSA) severity: an exploratory sequential mixed-methods study

Figures

• 

  FIGURE 1. Scatterplot of BRI and OSA severity. The scatterplot illustrates the positive correlation between the Body Roundness Index (BRI) and the Apnea-Hypopnea Index (AHI)-based OSA severity groups. The red solid line represents the linear fit, and the gray shaded area denotes the 95% confidence interval. R represents the correlation coefficient (r = 0.2853, P = 0.000).

• 

  FIGURE 2. Forest plot and interactions for subgroup analysis of BRI correlation with OSA severity. Odds ratios (OR) and 95% confidence intervals (CI) are illustrated. The blue square represents the subgroup OR, and the horizontal lines represent the 95% CI. Minimum blood oxygen saturation classifications are defined as: Mild hypoxia: > 90%; Moderate hypoxia: 85%–90%; Severe hypoxia: < 85%. Interaction tests were performed to assess effect modification across subgroups.

• 

  FIGURE 3. Interaction effect diagram of key modifying factors (gender, age, minimum blood oxygen level, and BMI) with BRI. (A) BRI and gender: Predicted probability of OSA severity across BRI values in males and females. (B) BRI and age: Association between BRI and OSA severity probability across age groups (<45, 45–60, >60 years). (C) BRI and minimum blood oxygen level: BRI-related OSA severity probability stratified by hypoxemia severity (mild: >90%; moderate: 85–90%; severe: < 85%). (D) BRI and BMI: Effect of BRI on OSA severity probability across BMI categories (Underweight, Normal, Overweight, Obese). The shaded bands surrounding the prediction curves represent the 95% confidence intervals.

• 

  FIGURE 4. Predictive performance of BRI for different severity levels of OSA. ROC curves show the diagnostic performance of the Body Roundness Index (BRI) in predicting Mild OSA vs. Normal, Moderate OSA vs. Normal, and Severe OSA vs. Normal. AUC, Area Under the ROC Curve; CI, Confidence Interval. Colored lines represent different severity comparison models.

• 

  FIGURE 5. Predictive performance of BMI and Neck circumference for different OSA severity levels. Comparative ROC analysis of Body Mass Index (BMI, left panels) and Neck Circumference (right panels) across different severity thresholds. AUC, Area Under the Curve; CI, Confidence Interval.

• 

  FIGURE 6. Predictive performance of BRI and BMI for severe OSA in gender subgroups. Gender-stratified ROC curves comparing the diagnostic efficacy of BRI (blue line) and BMI (red line) for severe OSA in males (left panel) and females (right panel).

• 

  FIGURE 7. Predictive performance of the combined model (BRI+BMI+neck circumference) for severe OSA. ROC curve illustrating the diagnostic performance of the integrated model combining Body Roundness Index (BRI), Body Mass Index (BMI), and neck circumference.

• 

  FIGURE 8. Data integration visualization. The joint display juxtaposes quantitative results (left panel: statistical association, subgroup disparities) with qualitative insights (right panel: disease perception bias, cognitive disconnect, and coping patterns) to resolve the “high BRI without severe OSA” paradox. The bottom panel represents the derived BRI-OSA Multilevel Regulatory Framework integrating individual cognition, clinical characteristics, and behavioral outcomes.