Abstract
STUDY OBJECTIVES: Time estimation is a complex cognitive task that is especially challenging when the time period includes sleep. To determine the accuracy of sleep duration perception, we investigated 44 healthy subjects participating in multi-day inpatient sleep protocols during which they had extended nighttime and short daytime sleep opportunities but no time cues or knowledge of time of day. METHODS: The first sleep opportunity was at habitual sleep time and duration. The subsequent 3, 4, or 11 days had 12-h nighttime sleep opportunities and 4-h daytime nap opportunities, potentially creating an experimentally induced "insomnia" with substantial time awake during scheduled sleep. RESULTS: Subjective sleep duration estimates were accurate for the first (habitual) sleep opportunity. The subjective reports following nighttime 12-h sleep opportunities significantly underestimated objective sleep duration, while those following daytime 4-h sleep opportunities significantly overestimated objective sleep duration. Misperception errors were not explained by poor sleep efficiency, which was lower during 4-h (~39%) than 12-h opportunities (~71%). Subjective sleep estimates after 4-h opportunities correlated with the percentage of REM and N3 sleep. Subjective sleep estimates following 12-h opportunities were, unexpectedly, negatively correlated with NREM stage 2 sleep. CONCLUSION: The estimation of sleep duration in the absence of time cues may depend on length of sleep opportunity and/or time of day. The results have implications for understanding sleep state misperception, which is an important consideration in patients with insomnia.