Night-to-night variation in snoring sound severity: one night studies are not reliable

Objectives: To quantify the night-to-night variation insnoring severity; to compare this with inter-subject varia-tion in snoring intensity: to compare multinight meansnoring scores with self-reported subjective scores.Design: Prospective observational study.Setting: Subjects were recorded during sleep at theirown homes.Participants: Twenty patients with socially disruptivesnoring awaiting surgery.Main outcome measures: Over four consecutive nightsusing a solid-state sound recording device, the mean,standard deviation and intra-class correlation coefficientwere calculated for (a) the loudest 1% of sound,(b) snore frequency and (c) total snore duration.Results were correlated with Snoring SymptomInventory scores assessed immediately prior to theserecordings.Results: Overall mean and intrasubject standard devia-tion for the loudest 1% of sound was 65.0 (±4.1) dB, forsnore frequency was 245 (±104) per hour and for totalsnore duration was 4.3% (±2.1). Intraclass correlationcoefficients were 0.78, 0.74 and 0.67, respectively, suggest-ing only moderate reliability of these outcome measures.No significant correlation was found between objectiveand subjective scores for either endpoint.Conclusion: Natural night-to-night variation in snoringseverity represents a significant proportion of overallsnoring variance, thus one night studies of snoring arenot reliable. The random error associated with one-nightstudies exceeds the expected effect size of snoring inter-ventions and so multi-night studies of at least four nightsare recommended to reduce the error. However, evenmulti-night objective measurements correlate poorly withsubjective scores of snoring.Snoring is a ubiquitous problem that can have personaland social ramifications. In a telephone survey of UK res-idents, 40% of respondents reported some degree of snor-ing. The prevalence increased with age up to 55 years andwas also higher in men than women.