Comparison of snoring measured at home and during polysomnographic studies.

Snoring characteristics depend on several factors (sleep position, sleep architecture, breathing route) that can be influenced by changes in sleep habits and by the presence of the different probes and electrodes during polysomnographic studies. Our objective in this study was to compare the characteristics of snoring in the home environment with those of the sleep laboratory where most conventional studies are carried out. Fourteen nonapneic snorers were subjected to three night recording sessions within a two-week period, two at home and one in the sleep laboratory. To eliminate any sleep interference by the apparatus, breathing sounds were recorded with two microphones symmetrically placed on either side of the bed, the signal being preamplified and stored on a VHS hi-fi video recorder. The recorded signal was analyzed by using a spectrum analyzer (real time analyzer) and an equalizer to correct for acoustic resonances of the bedrooms. A snoring event was defined as a breathing sound with a sound pressure level (SPL) greater than 60 dB SPL. The snoring index (number/sleep hour) and the sound intensity of each event were automatically determined. The total sleep time (TST) was similar for the two home recordings (6.8 +/- 0.2 and 7.0 +/- 0.2 h, respectively, mean +/- SEM), but it was significantly shorter during the hospital study (6.0 +/- 0.3 h). The snoring indices obtained at home were 141.4 +/- 33.3 and 144.1 +/- 41.2/h and not statistically different from those obtained during the hospital recording (209.1 +/- 41.5/h). The percentage of TST spent above 60 dB SPL was significantly greater during the polysomnographic study (4.3 +/- 1.2 percent) than during the home recordings (2.5 +/- 0.7 and 2.9 +/- 1.0 percent, respectively). We conclude that the severity of snoring may be overestimated during polysomnographic recordings.