Perception versus Reality: The Relationship between Subjective and Objective Measures of Sleep When On-call under Simulated Laboratory Conditions

ABSTRACT

Background

On-call working arrangements have been shown to negatively impact sleep. However, workers may perceive their sleep to be worse than it actually is. The aim of this study was to compare participants’ pre- and post-sleep estimates of sleep duration and sleep quality with objectively measured sleep when on-call under laboratory conditions.

Participants

72 healthy, adult males.

Methods

Analyses were performed on three interrelated studies, all of which consisted of four nights in a sleep laboratory. Following adaptation and baseline nights were two on-call nights (sleep opportunity 23:00 h – 07:00 h). Before and after each sleep opportunity, participants provided subjective estimates of sleep. Sleep was objectively measured using polysomnography.

Results

Estimated sleep duration (6.74 ± 1.13 h) and sleep onset latency (20.55 ± 14.85 min) were significantly poorer than objectively measured sleep outcomes (sleep duration 7.21 ± 1.25 h; sleep latency 13.20 ± 10.06 min). Of the variance in post-sleep estimated sleep duration, 14% was associated with objectively measured minutes of N3 (R²Δ = 0.55) and REM (R²Δ = 0.75). Additionally, 14% of post-sleep sleep quality estimation variance was associated with minutes of N2 (R²Δ = 0.60) and N3 (R²Δ = 0.79), measured by polysomnography.

Conclusions

Some objective measures of sleep were associated with subjective estimates of sleep duration and sleep quality. However, individuals may overestimate sleep onset latency and underestimate sleep duration during on-call periods. It may be beneficial for on-call workers to actively reflect on feelings of fatigue/alertness for workplace fatigue management, rather than relying solely on estimates of sleep.

Acknowledgments

The authors would like to thank Dr Xuan Zhou, the project staff, and the participants for volunteering their time.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

This study was funded by an Australian Research Council Discovery grant (DP 150104497).