An ecological approach to clinically assess nightmares in military service members with severe PTSD

Emeric Saguin,Dorone Feingold,Jean-Baptiste Roseau,Michael Quiquempoix,Marie Boussaud,Clotilde Izabelle,Arnaud Metlaine,Mathias Guillard,Pascal Van Beers,Charles Gheorghiev,Bertrand Lahutte,Damien Leger,Danielle Gomez-Merino,Mounir Chennaoui

DOI: https://doi.org/10.1016/j.sleep.2023.01.024

Abstract:Background: Trauma-related nightmares (TRNs) are distressing events which contribute to insomnia severity, chronicity and treatment resistance of PTSD. Therefore, recording TRNs is a crucial technical challenge in order to understand their physiopathological patterns and their impact on sleep. However, TRNs are difficult to record during a single night in a sleep laboratory, which, moreover, is likely to be considered by patients as a protective sleep environment that is therefore not representative of home sleep conditions. Method: In the present study, we investigate if objective sleep measures acquired at-home using two ambulatory devices is of clinical value by correlating with PTSD patients' complaints about sleep and nightmares. A secondary objective is to relate awakenings associated with TRNs to sleep stages and to provide new insights into the use of electrodermal activity (EDA) as a potential physiological marker of TRNs. Sixty veterans and active-duty service members were assessed by questionnaires and recorded for 5 consecutive nights in their homes. Results: Our approach firstly identified positive correlations between subjective and objective sleep parameters (total sleep time, sleep-onset latency and TRNs frequency). We also developed a method of synchronization between the two ambulatory devices that allowed us to match 200 TRNs (reported by event marker push button) with sleep stages corresponding to 91 nights and 37 patients. Most awakenings associated with TRNs occurred during NREM sleep (65.5% versus 34.5% during REM sleep). Our results also reveal significant differences in the frequency of EDA peaks 10 min before the reported events, with a lower frequency in REM (13.7 peaks) than in NREM (24.8 peaks) awakenings associated with TRNs. This EDA peaks frequency in REM sleep is not statistically different from that in REM sleep preceding awakenings that are not associated with TRNs. Conclusion: The development of wearable devices to collect physiological parameters is of interest in clinical practice to improve our knowledge of sleep and trauma-related nightmares in patients with PTSD.