Sleep apnea in morbidly obese patients: prevalence and clinical predictivity.

Background: Little is known about the prevalence of obstructive sleep apnea-hypopnea syndrome (OSAHS) in morbidly obese patients and whether such patients show peculiar clinical findings that may make it easier to suspect and diagnose OSAHS.

Objectives: To investigate prevalence of OSAHS in patients with morbid obesity and find a simple structured model for predicting the results of polysomnography.

Methods: The study enrolled a group of 101 consecutive inpatients (33 males, age range 20-80 years) with a body mass index > or =40, whose symptoms of OSAHS were not known, and a validation group of 45 patients.

Results: Habitual snoring, nocturnal apneas or awakening as well as diurnal sleepiness were frequent findings (90.1, 40.6, 50.5 and 61.4%, respectively). Chronic obstructive pulmonary disease, hypertension, diabetes and myocardial ischemia were also frequently associated (22.8, 56.4, 30.7 and 6.9%, respectively). OSAHS was found in 61 (60.4%) patients, in 33.7% it was of severe degree. A multivariate logistic regression model allowed to select the independent predictors of OSAHS: age, male sex, diurnal sleepiness and the value of minimum nocturnal saturation. Sensitivity of 97%, specificity of 77% as well as positive and negative predictive values of 87% and 95%, respectively, were obtained; similar results were found in the validation group. When the best obtainable cutoff on the receiver operating characteristic curve is below 40%, the instrumental diagnosis might be excluded in as many as 33% of cases, since they are not affected by OSAHS or have OSAHS of mild degree.

Conclusions: OSAHS is present in almost two thirds of morbidly obese patients. By applying the prediction model we propose, one may calculate the probability of a morbidly obese patient of being affected by OSAHS.