Clinical and functional prediction of moderate to severe obstructive sleep apnoea

INTRODUCTION: Upper airway inflammation and narrowing are characteristics of obstructive sleep apnoea (OSA). Inflammatory markers have been found to be increased in exhaled breath and induced sputum of patients with OSA. OBJECTIVES: The aim of this study was to investigate if the measurement of exhaled nitric oxide (F(ENO) ), as marker of airway inflammation, together with the forced mid-expiratory/mid-inspiratory airflow ratio (FEF(50) /FIF(50) ), as marker of upper airway narrowing, may help to predict OSA. METHODS: Two hundred one consecutive outpatients with suspected OSA were prospectively studied between January 2004 and December 2005. All patients underwent clinical examination, spirometry with measurement of FEF(50) /FIF(50) , maximum inspiratory pressure, arterial blood gas analysis, exhaled nitric oxide (F(ENO) ) and overnight polysomnography. Linear regression models were used to evaluate the effect of measured variables on the apnoea-hypopnoea index (AHI). Models were cross-validated by bootstrapping. RESULTS: Most of the patients were obese and had severe OSA. FEF(50) /FIF(50) , F(ENO) and an interaction term between smoking and F(ENO) contributed significantly to the predictive model for AHI, in addition to age, neck circumference, body mass index and carboxyhaemoglobin saturation. A nomogram to predict AHI was obtained, which converted the effect of each covariate in the model to a 0-100 scale. The nomogram showed a good predictive ability for AHI values between 25 and 64. CONCLUSIONS: The measurement of F(ENO) and of FEF(50) /FIF(50) improves the ability to predict OSA and may be used to identify patients who require a sleep study.