Reliability of home-based physiological sleep measurements in snoring and non-snoring 3-year olds

Abstract

To study snoring and non-snoring 3-year olds in their own homes and to establish reference values and night-to-night variability of physiological measurements taken during sleep. One hundred and sixty-six children, aged 3.2-4.0 years, identified as at high (n = 83, M/F = 1.5:1) or low (n = 83, M/F = 1.4:1) risk for persistent snoring, as rated by a parent, wore a sleep screening system (Grey Flash) for up to five consecutive nights, to measure snoring, oxygen saturation, heart rate, movement, sleep efficiency, and sleep timing. The snoring group snored 25% of their total sleep time, while the non-snoring group snored just 7.6%. Reliability estimates were calculated, using intraclass correlations to establish the reliability of single recordings and the Spearman-Brown prophecy formula to estimate reliability over multiple recordings. Snoring (percent), turn index (number of turns per hour), and mean heart rate were adequately assessed in both groups with one recording night (all intraclass correlation coefficients (ICCs) ≥0.70). Furthermore, mean SpO(2) was measured with sufficient reliability with two recordings in non-snorers (ICC = 0.71), while five recording nights were necessary to reach reliability in snorers (ICC = 0.71). SpO(2) nadir did not reach adequate reliability in either group even after seven recording nights. Our results showed that snoring can be measured reliably at home with just one recording night, whereas most other physiological sleep measures require two or more recordings.