Neck circumference and other clinical features in the diagnosis of the obstructive sleep apnoea syndrome.

Abstract

BACKGROUND: Neck circumference has been suggested to be more predictive of obstructive sleep apnoea than general obesity, but the statistical validity of this conclusion has been questioned. Combining neck circumference with other signs and symptoms may allow the clinical diagnosis or exclusion of sleep apnoea to be made with reasonable confidence. This study examines these issues. METHODS: One hundred and fifty patients referred to a sleep clinic for investigation of sleep related breathing disorders completed a questionnaire covering daytime sleepiness, snoring, driving, and nasal disease. Body mass index and neck circumference corrected for height were measured and obstructive sleep apnoea severity was quantified as number of dips in arterial oxygen saturation (SaO2) of more than 4% per hour of polysomnography. Multiple linear regression was used retrospectively to identify independent predictors of SaO2 dip rate, and the model derived was then prospectively tested in a further 85 subjects. RESULTS: The retrospective analysis showed that the question "Do you fall asleep during the day, particularly when not busy?" was the best questionnaire predictor of variance in the SaO2 dip rate (r2 = 0.13); no other question improved this correlation. This analysis also showed that neither body mass index nor any of the questionnaire variables improved the amount of variance explained by height corrected neck circumference alone (r2 = 0.35). A statistically similar prospective analysis confirmed this relationship (r2 = 0.38). CONCLUSIONS: Prospective study of these patients referred to a sleep clinic with symptoms suggesting sleep apnoea shows that neck circumference corrected for height is more useful as a predictor of obstructive sleep apnoea than general obesity. None of the questionnaire variables examined add to its predictive power, but alone it is inadequate to avoid the need for sleep studies to diagnose this disease.