Objective To evaluate diagnostic value of body mass index (BMI) in screening and diagnosing obstructive sleep apnoea-hypopnoea syndrome and to determine the reference standard of body mass index (BMI) in both male and female population by receiver operating characteristic (ROC) curve.
Method A total of 1348 subjects (1063 male and 285 female) who experienced polysomnography (PSG) during 2004–2010 in The Third Hospital of Peking University ENT department were included. The correlativity between apnoea-hypopnoea index (AHI) and BMI was conducted by pearson correlation test. The diagnostic value and reference standard in both male and female were analysed by ROC curve.
Result 1) A positive relationship between AHI and BMI was detected, both in male and female subjects, with pearson coefficients 0.423 and 0.419 (p<0.001). 2) Results of ROC analysis show a favourable value of BMI in diagnosing OSHAS, especially in male subjects. In male population, area under curve (AUC) was 0.759 (95% CI 0.724 to 0.795) with a best cutoff value 25.3 kg/m2. The sensitivity and specificity in man were 74% and 61%, respectively; in women, area under curve (AUC) was 0.691 (95% CI 0.632 to 0.751) with a best cutoff value 24.7 kg/m2. The sensitivity and specificity were 68% and 61%. The screening cutoff values were 22.4 kg/m2 and 20.7 kg/m2 in male and female, with a sensitivity of 95%. The diagnostic cutoff values were 30.0 kg/m2 and 36.0 kg/m2, with a specificity of 95%.
Conclusion BMI seemed a better diagnostic value of OSAHS in diagnosing than screening, especially in male population. Sexual-separated cutoff values should be use in clinical practice to enhance diagnostic accuracy.