According to a recent theoretical model, snoring is related to instability of the upper airway (UA). Factors promoting UA instability include increased gas density. The aim of this study was to test the influence of gas density on simulated snoring production and supraglottic resistance.Supraglottic pressure and flow rate (V ') were measured in 10 healthy seated subjects during simulated snoring. Subjects breathed three different gas mixtures: Helium±oxygen, He 79%±O 2 21% (He±O 2 ); air; and sulphur hexafluoride±oxygen, F 6 S 79%±O 2 21% (F 6 S±O 2 ) administered in a random order. Supraglottic resistance (Rsg) was measured on its linear range during quiet breathing and V ' was measured at the onset and middle of snoring.Linear Rsg increased and V ' conversely decreased with gas density. These data are in agreement with predictions of a mathematical model of the upper airway showing that snoring occurs at lower flow rates when gas density is increased. Eur Respir J 1999; 13: 679±681. Snoring is a cardinal symptom of obstructive sleep apnoea (OSA) syndrome and a frequent finding in otherwise healthy subjects [1]. Previous studies from the authors' laboratory have shown that simulated and spontaneous snoring are characterized by high-frequency oscillations of the soft palate associated with a decrease in oropharyngeal calibre. Reduction of the pharyngeal cross-sectional area increases supraglottic resistance (Rsg) and is associated with flow limitation [2,3].The influence of gas density and viscosity on the upper airway (UA) pressure±flow relationship has not been studied before, except in animals [4].Recently, GAVRIELY and JENSEN [5], based on a theoretical model of UA, found that among other factors, increases in UA resistance and gas density, as well as decrease in UA size promote UA instability and snoring.In this study, the influence of gas density on supraglottic resistance and flow rate (V ') was tested during quiet breathing and during simulated snoring.
Materials and methodsTen healthy (five females) nonsmoker subjects aged 22± 30 yrs with normal spirometric data [6] and all nonobese (meanSD body mass index, 32.42.9 kg . m -2 ), volunteered for this study. They were nonsnorers and free of any UA pathology. Written, informed consent was obtained and the experiments were approved by the Ethics Committee of the hospital.V ' was measured with a Fleish No. 2 pneumotachograph (Fleicsh, Lausanne, Switzerland), attached to a tightly fitting silicone face mask. Pressure drop across the pneumotachograph was measured by a 5 cmH 2 O Validyne pressure transducer (Northridge, CA, USA). The airtightness of the mask was verified by measuring the helium level around the border of the mask when the subjects were breathing a 79% He±21% O 2 mixture. The pneumotachograph was calibrated for the three gas mixtures: helium±oxygen, He 79%±O 2 21% (He±O 2 ); air; and sulphur hexafluoride±oxy-gen, F 6 S 79%±O 2 21% (F 6 S±O 2 ) by passing these gases through the pneumotachograph in series with a dry gasometer (Parkinson ...