PTSD Symptom Severity Associated With Sleep Disturbances in Military Personnel: Evidence From a Prospective Controlled Study With Ecological Recordings

Abstract

Sleep disturbances, including insomnia and trauma-related nightmares (TRNs), are the core symptoms of post-traumatic stress disorder (PTSD) in military personnel. Furthermore these nocturnal manifestations are directly related to the persistence of daytime PTSD symptoms and are known to exacerbate comorbid conditions such as depression, suicidality, and daytime impairments. This prospective study examined the variability of PTSD-related sleep disruptions and its relationship to symptom severity using ecological recordings over several nights. One hundred thirty PTSD-diagnosed service members and 65 healthy military controls recorded sleep data at home for five nights using a polysomnographic headband to measure total sleep time (TST), sleep onset latency (SOL), wake after sleep onset (WASO), sleep efficiency index (SEI), and sleep stages. PTSD severity and comorbid symptoms were assessed by clinical evaluations. Compared to controls, PTSD participants had higher SOL and WASO (+14.1 min and +9.1 min, p < 0.001, respectively), reduced SEI (-6.6%, p < 0.001), and lower N3 and rapid eye movement (REM) sleep durations. In addition, night-to-night variability (NNV) in SOL and WASO was higher in the PTSD group. The sleep fragmentation index (FI)-and more specifically non-REM (NREM) sleep fragmentation-was significantly correlated with PTSD severity, particularly the intrusive and avoidance symptoms clusters in the PCL-5 score. The results highlight the need for customized multinight assessments to study sleep variability in military patients with combat-related PTSD, in order to advance therapeutic strategies for military populations. Trial Registration: ClinicalTrials.gov Identifier: NCT04581850.

Keywords: PTSD; military personnel; sleep; trauma-related nightmares; wearable devices.

Figure 1

Study timeline and selection process for recorded nights.

Figure 2

Boxplots displaying the distribution of the different sleep parameters for within-group variability (WGV) between PTSD-G (n_Subjects = 119, n_Nights = 455) and HM-G (n_Subjects = 59, n_Nights = 275). ⁣^∗∗∗p  < 0.001, statistical differences between the two groups.

Figure 3

Night-to-night variability (NNV) in sleep parameters between PTSD-G (n = 112) and HM-G (n = 59). root mean squared successive differences (RMSSDs) is used as an index. The histograms and accompanying box plots highlight the differences in variability between both PTSD-G and HM-G for each sleep parameter. ⁣^∗∗p  < 0.01 and ⁣^∗∗∗p  < 0.001, statistical differences between the two groups.

Figure 4

(A) Network analysis. ⁣^∗Line's width represents significant Spearman's rank correlation coefficient. s-parameters correspond to subjective parameters from the Pittsburgh sleep quality index (PSQI). The o-parameters correspond to objective parameters obtained with the DH. (B) Spearman correlation matrix between objective sleep parameters and PCL5 total score and symptom clusters (Cluster B reflects intrusive symptoms; Cluster C, avoidance symptoms; Cluster D, negative alterations in cognition and mood; and Cluster E, alterations in arousal and reactivity). Only statistically significant correlations are displayed. FI = fragmentation index (events per hour); %N3 = Stage 3 sleep, percentage of total sleep; PCL5 = post-traumatic stress disorder checklist for DSM-5, total score; %REM, percentage of total sleep time spent in rapid eye movement sleep; oTST, objective TST; SOL, sleep onset latency; TST, total sleep time; WASO, wake after sleep onset.