Abstract:
This study used archival data to examine variability in sleep diary measures across a two-week period. Data were analyzed for 592 normal sleepers (defined as absence of sleep disorder) ranging in age from 20 to 96 years. Variability was examined in four sleep diary parameters: total sleep time (TST), number of nighttime awakenings (NWAK), sleep-onset latency (SOL), and wake time after sleep onset (WASO) for the overall sample and by age, sex (male, n = 294, female, n = 298) and race (White, n = 415; Black, n = 177). Intraclass correlation coefficients were calculated from multilevel models to describe the extent to which self-reported sleep varied within- and between-individuals across the two-week period. Night-to-night differences in sleep within the same individual generally exceeded sleep differences between individuals for TST, SOL, and WASO. Multiple regression analyses were conducted to test for age, sex, and race differences in night-to-night, intra-individual variability (measured by the intra-individual standard deviation across nightly values for each individual). Results showed that the amount of intra-individual variability in TST and NWAK decreased with older age. Further, the degree of reduction in variability in TST associated with age was dependent on sex and race. Although effect sizes were small, females tended to have more intra-individual variability in SOL and NWAK than males. Results of this study provide empirical support for considerable night-to-night variability in subjective sleep in the general population and should be taken into account in future research linking sleep with various psychological and physical health outcomes. Methodological implications regarding measurement/research design are also discussed.
