Authors: Cresta Asah, Hinuga Sandahl, Lone Baandrup, Jessica Carlsson, Poul Jennum
Categories: Clinical Research Article, Post-Traumatic Stress Disorder, refugees, sleep disturbance, Dream Enactment Behavior, Trauma-Associated Sleep Disorder, TASD, REM Sleep Behaviour Disorder, RBD, Trastorno de estrés postraumático, refugiados, Trastorno del sueño Asociado a Trauma, Trastorno de conducta asociado al sueño REM, Alteración del sueño, TEPT, Sueño, Conducta de Representación Onírica
Source: European Journal of Psychotraumatology
Authors: Cresta Asah, Hinuga Sandahl, Lone Baandrup, Jessica Carlsson, Poul Jennum
Background: Sleep disturbances are widely reported in Post-Traumatic Stress Disorder (PTSD). Although Dream Enactment Behaviour (DEB) has long been associated with PTSD, its high prevalence has only recently been recognized, sparking discussions about the classification of trauma-related sleep disorders. The impact of DEB on treatment outcomes in PTSD remains unexplored.
Objective: To investigate the role of DEB in functional impairment, symptom severity, subjective sleep disturbances, and treatment response in patients with PTSD, and how it relates to Trauma-Associated Sleep Disorder (TASD).
Methods: We analyzed data from a randomized controlled trial carried out in a specialized mental health clinic in Denmark. The trial investigated refugees with PTSD allocated to four groups receiving different combinations of PTSD therapy. Participants completed self-report questionnaires assessing functional impairment, symptom severity, and subjective sleep disturbances, including the REM Sleep Behaviour Disorder Screening Questionnaire (RBDSQ), the Pittsburgh Sleep Quality Index (PSQI), and the Typical Dream Questionnaire (TDQ), at baseline and follow-up. The sample was split into two groups based on the presence of self-reported DEB, and compared at baseline and follow-up. Statistical analyses included chi-square test, Mann–Whitney U test, and regression.
Results: A sample of 176 RBDSQ respondents was studied, of which 71% met the criteria for DEB (N = 125). DEB was significantly associated with a poorer treatment response on sleep quality as assessed by the PSQI (N = 122, p = .035), irrespective of treatment group. No differences were observed in functional impairment or symptom severity. Of the 67 TDQ respondents with DEB, 60% did not have trauma-related nightmares (N = 40).
Conclusions: DEB is a significant feature of sleep in PTSD and seems to limit the efficacy of treatment interventions. TASD does not encompass all cases of DEB in PTSD and the concept needs further development to be clinically useful.
Trail Registration: ClinicalTrials.gov NCT02761161; clinicaltrials.gov/study/NCT02761161.
We are currently witnessing the highest number of forcibly displaced people in the world since the United Nations Refugee Agency (UNHCR) began recording numbers after the Second World War (United Nations, 1951). The most recent UNHCR report estimates that 108.4 million people are forcibly displayed due to war, persecution, disasters, or violence, with 35.3 million of them being registered as refugees (United Nations High Commission on Refugees [UNHCR] 2022). A recent meta-analysis revealed a prevalence of Post-Traumatic Stress Disorder (PTSD) of 31% among refugees and asylum seekers, underscoring the urgent need for improved treatment approaches for this debilitating condition (Blackmore et al., 2020).
Sleep impairment is widely recognized as a hallmark of PTSD, with most patients experiencing varying degrees of sleep disturbances (Germain et al., 2008; Sandahl et al., 2017; Slavish et al., 2022). It was previously thought that sleep disturbances were a consequence of PTSD and would resolve concomitantly with the therapeutic treatment of PTSD. However, they are now being recognized as a distinct and important part of the disorder that requires specific attention and treatment (Spoormaker & Montgomery, 2008), Brownlow et al., 2020).
Recently, dream enactment behaviour (DEB) has emerged as an interesting area of investigation in the context of sleep in PTSD (Baird et al., 2018; Barone, 2020; Barone, 2024; Brock et al., 2019; Elliott et al., 2020; Feemster et al., 2019; Jones et al., 2022; Lee et al., 2020; Mysliwiec et al., 2014; Taylor et al., 2023). DEB is characterized by vocalization (talking, yelling) and motor activity (punching, kicking, thrashing) correlated with the patient’s dream content. Even though DEB can be seen in relation to several conditions, it is considered the cardinal symptom of REM Sleep Behaviour Disorder (RBD). RBD is a parasomnia diagnosed by a history of DEB in conjunction with the finding of lack of muscle paralysis during REM sleep (REM Sleep Without Atonia; RSWA) by polysomnography (PSG). Idiopathic RBD (iRBD), defined as RBD without an identifiable cause, is recognized as an early marker of neurodegenerative disease, primarily occurring in adults above the age of 50 years and generally without a history of preceding trauma (Fernández-Arcos et al., 2016; Galbiati et al., 2019; Iranzo et al., 2016; Williams et al., 2021). The consideration of RBD-like symptoms in younger patient populations with a history of trauma exposure has therefore led to the suggestion that these symptoms may result from a state of hyperarousal rather than from neurodegeneration, prompting the proposal of a new parasomnia called Trauma-Associated Sleep Disorder (TASD) (Husain et al., 2001; Mysliwiec et al., 2018). In addition to DEB, the diagnostic criteria of TASD include a history of trauma, altered dream mentation related to the prior traumatic experience, and symptoms of autonomic hyperarousal, while RSWA is considered a supportive finding (Brock et al., 2022; Mysliwiec et al., 2014; Mysliwiec & Brock, 2020). Under the current diagnostic criteria, patients with PTSD may exhibit DEB without meeting the criteria for either TASD or RBD, although, conversely, some cases could qualify as both (Barone, 2020; Barone, 2024; Jones et al., 2022). Differentiating PTSD, TASD, and RBD remains a challenge due to their distinct yet overlapping pathologies, each of which requires differentiated treatment and prognostic approaches. The impact of DEB on subjective sleep disturbances and treatment response in patients with PTSD has not previously been explored.
We set out to investigate the role of DEB in PTSD, thereby to enhance its relevance in the clinical management of patients with PTSD.
Specifically, we wished to (a)The prevalence of DEB and differences in demographic and clinical characteristics between patients with PTSD only and patients with PTSD and DEB.(b)Baseline differences in subjective sleep disturbances, level of functioning and PTSD symptom severity in patients with PTSD only and patients with PTSD and DEB.(c)The association of DEB at baseline with treatment response.(d)The association of DEB at baseline with nightmare themes.
We hypothesized that (1) DEB is highly prevalent in patients with PTSD, and (2) patients with PTSD and DEB differ significantly from patients with PTSD but without DEB with respect to their demographic and clinical characteristics, subjective sleep disturbances, level of functioning, PTSD symptom severity, nightmare themes, and treatment response.
Data for the RCT were collected between March 2016 and April 2018 at the Competence Centre for Transcultural Psychiatry (CTP), Mental Health Centre, Ballerup, Denmark, which is a tertiary, out-patient mental health clinic for trauma-affected refugees. The study was approved by The Ethics Committee of the Capital Region of Denmark and was registered at ClinicalTrials.gov (identifier: NCT02761161). Patients who fulfilled the inclusion criteria for the trial were offered the opportunity to participate. They were accepted for the study only if their written informed consent had been obtained. The patient population were refugees aged 18 years or above who fulfilled the World Health Organization International Classification of Diseases, 10th edition (ICD-10) research criteria (World Health Organization, 1993) for a PTSD diagnosis. Inclusion criteria prior psychological trauma experienced outside Denmark, sleep disturbances measured as a score of ≥ 8 on The Pittsburgh Sleep Quality Index (Mollayeva et al., 2016), and nightmares measured as a score of ≥ ‘a little’ on the Harvard Trauma Questionnaire Nightmare item (Wind et al., 2017). Exclusion criteria with a severe psychotic disorder (part of an independent psychotic disorder but not a part of severe PTSD), with current alcohol or drug use disorder, with a known neurodegenerative disorder, requiring admission to a psychiatric hospital, and pregnant or breastfeeding women. For more detailed inclusion and exclusion criteria, see the published trial protocol (Sandahl et al., 2017).
The RCT examined four treatment (1) Treatment As Usual (TAU), consisting of an interdisciplinary treatment approach with medicine that adopted best clinical practice, physiotherapy, psychoeducation, and manual-based CBT; (2) TAU combined with Imagery Rehearsal Therapy (IRT); (3) TAU combined with treatment with mianserin; and (4) TAU in combination with both IRT and mianserin. The efficacy of the treatments was compared through questionnaires, with changes in PSQI being the main outcome measure. The RCT found an improvement in the PSQI, with no significant differences between the treatment groups. The workflow of the RCT is explained in detail in the paper by Sandahl et al. (Sandahl et al., 2021).
We derived a study population from participants who had completed the RCT and filled out the REM Sleep Behaviour Disorder Screening Questionnaire (RBDSQ), allowing us to divide them into two groups based on their history of DEB. We also administered the following the Pittsburgh Sleep Quality Index (PSQI) (Mollayeva et al., 2016), the Nightmare Effects Survey (NES) (Krakow et al., 2000), the WHO Disability Assessment Schedule 2.0 (WHODAS-12) (Ustün et al., 2010), the Sheehan Disability Scale (SDS) (Sheehan & Sheehan, 2008), the Harvard Trauma Questionnaire (HTQ) (Wind et al., 2017), and a modified version of the Typical Dream Questionnaire (TDQ) (Zadra et al., 2003). The questionnaires were all self-administered, apart from the WHODAS and the modified TDQ. The modified TDQ was completed 21 times at 2–3 week intervals over the 12-month treatment period. Data were collected prospectively through daily dream logs, which patients filled out based on free recall. Every two weeks, a psychologist reviewed the diaries with the patients, confirmed the themes, and documented them in the modified TDQ. The RBDSQ was completed at baseline. All other questionnaires were completed at baseline (pre-treatment) and again at follow-up (post-treatment) after approximately 12 months of treatment, as shown in Figure 1. Figure 1.Inclusion of patients based on symptoms of Dream Enactment Behavior. RBDSQ: REM Sleep Behavior Disorder Screening Questionnaire. PSQI: Pittsburgh Sleep Quality Index. NES: Nightmare Effect Survey. WHODAS: WHO Disability Assessment Schedule 2.0. SDS: Sheehan Disability Scale. HTQ: Harvard Trauma Questionnaire. TDQ: Typical Dream Questionnaire. *TDQ was completed 21 times in a subset of 110 patients. Treatment As Usual (TAU), TAU + mianserin, TAU + Imagery Rehearsal Therapy (IRT), TAU + IRT + mianserin. N = number.
The RBDSQ consists of 10 items with a maximum score of 13. It is one of the most widely used screening tools for DEB and is commonly employed in the diagnosis of RBD, in which context it has been validated against PSG findings of nocturnal motor activity (Li et al., 2017). A meta-analysis evaluating the diagnostic accuracy of the RBDSQ in the general population identified a cut-off score of 5, which yielded a pooled sensitivity of 0.91 and a specificity of 0.77 (Li et al., 2017).The PSQI is based on seven components from 18 questions and provides a measure of general sleep quality during the previous month. Total scores may vary between 0 and 21. It is generally accepted that poor sleep is indicated by a score of ≥ 5 in the normal population, but of ≥ 8 in certain patient populations (Mollayeva et al., 2016).The NES consists of 10 questions that may each score between 0 and 4, thus allowing total scores from 0 to 40, with higher values indicating more severe life impairment due to nightmares (Krakow et al., 2000).The WHODAS assesses functioning (Ustün et al., 2010). It consists of 12 questions, each with a maximum score of 4, thereby allowing scores between 0 and 48. There is no specific cut-off, but a score of ≥ 10 is indicative of significant clinical difficulty (Andrews et al., 2009).The SDS assesses functional impairment (Luciano et al., 2010). Scores range from 0 to 30 points, with high scores indicating poor functioning. There is no generally accepted threshold score defining poor functioning, but it has been found to be sensitive to treatment effects (Sheehan & Sheehan, 2008).The HTQ assesses the severity of PTSD symptoms (Wind et al., 2017). It consists of 16 questions, which produce a total score between 0 and 4. A score greater than 2.5 is considered indicative of clinical symptomatic PTSD (Mollica et al., 1992).Nightmare themes were documented using a modified version of the Typical Dream Questionnaire (Zadra et al., 2003), see Supplementary Table 1.
We analyzed the dataset, comparing patients with PTSD and DEB to patients with PTSD without DEB. The presence of DEB was determined using the RBDSQ. To enhance diagnostic accuracy, we applied a modified set of stringent criteria, based on previous studies’ findings that positive responses to questions 1 and 6 were the most sensitive and specific (Lee et al., 2015; Stiasny-Kolster et al., 2007). DEB was concluded to be present when a positive answer to Question 1, a positive answer to at least one of the four sub-questions in Question 6, and a total score of ≥ 5 were obtained. If a patient answered ‘no’ to Question 1 or ‘no’ to all sub-questions in Question 6, or had a total score of < 5, they were categorized as having no history of DEB. Question 1 of the RBDSQ assesses agreement with the ‘I sometimes have very vivid dreams’. Question 6 of the RBDSQ assesses agreement with the I experience or have experienced the following phenomena during my (6.1) speaking, shouting, swearing, laughing; (6.2) sudden limb movements or ‘fights’; (6.3) gestures, complex movements, that are useless during sleep, e.g. waving, saluting, shooing away mosquitoes, falling off the bed; and (6.4) things that have fallen down around the bed, e.g. bedside lamp, book, glasses. To address concern that vivid dreams (Question 1 of the RBDSQ) might drive classification, we assessed the distribution of answers and within-group treatment responses in the non-DEB group, and examined collinearity between vivid dreams (Question 1) and motor behaviour during sleep (Question 6) (Supplementary Tables 2–4).
Across all questionnaires, participants who completed fewer than 50% of the questions were considered ineligible for inclusion in the analysis. Participants with partially complete questionnaires with a minimum of 50% complete responses were included. Their average score of the answered questions was calculated and used as a substitute for the missing answers to equal a total score. This was done to control for underestimation when the two total scores (baseline and follow-up scores) were compared. Separate analyses were performed for each questionnaire, resulting in varying sample sizes for each questionnaire. This approach was adopted to make use of as much of the available data as possible and to ensure the biggest possible sample size in each analysis (Supplementary Figure 1). For questionnaires that revealed statistically significant differences between groups, we examined the basic characteristics and treatment randomization to check for bias arising from missing values. From the modified TDQ, the theme of ‘reexperiencing traumatic events’ was assessed as one of the diagnostic criteria for TASD (Mysliwiec & Brock, 2020). Three themes were characterized as those often seen in relation to RBD (hereafter referred to as RBD-like nightmares; see Supplementary Table 1) (Blumberg & Plumeau, 2016; Fantini et al., 2005; Oudiette et al., 2009).
Pearson’s chi-squared test was used to test for differences in the proportions of categories of demographic and clinical characteristics between patients with PTSD only and patients with PTSD and DEB.
The Mann–Whitney U-test was used to test for baseline differences in questionnaire scores, assessing subjective sleep disturbances, level of functioning and PTSD symptom severity in patients with PTSD only and patients with PTSD and DEB.
Ordinary least squares (OLS) regression was used to analyze the association of DEB at baseline with treatment response at follow-up by evaluating changes in questionnaire scores (PSQI, NES, WHODAS, SDS, HTQ) between baseline and follow-up. This technique enabled us to quantify the relationship between the presence of DEB (independent variable) and the change in questionnaire scores (dependent variable), while adjusting for other potential confounding age, sex, and allocated treatment groups. The OLS regression model assumptions were tested, and the model adjusted accordingly.
We found that 125 patients (71%) had DEB, while 51 patients (29%) had PTSD without symptoms of DEB. The two groups both had an average age of 44 years, and comprised similar proportions of men and women. The time between the trauma and joining the study ranged from 1 to more than 20 years. Table 1 summarizes the demographic and clinical characteristics. No significant differences were found between the groups with respect to educational level, whether a previous head trauma involved loss of consciousness, or the types of psychotropic and non-psychotropic medications used, for which reason these factors are not included in the table. Table 1.Summary of Basic Characteristics and Clinical Information. + DEB− DEBTotal sample in analysis Basic characteristicN%N%N**p 12571.05129.0 Sex 176.7521Male6753.62651.0 Female5846.42549.0 Weekly nightmares 110.0087Yes8568.02549.0** ** No118.81121.6 Age, years 176.6841Mean ± SD44 ± 8 44 ± 4 30 or younger1310.4815.7 31–507560.02854.9 51 or older3729.61529.4 Time since trauma, years 140.68611–52019.2719.4 6–1065.8411.1 11–201312.538.3 > 206562.52261.1 Smoker 148.9921Yes3931.21733.3 No6451.22854.9 Highest level of education 159.3071Primary5644.82447.1 Secondary2923.2713.7 Tertiary2822.41529.4 Head trauma 136.8532Yes3729.61529.4 No6148.82345.1 Previous depression 131.1074Yes9072.03466.7 No75.600 Medication - Psychotropic drugs5544.02243.1176.8828- Non-psychotropic drugs7963.23262.7168.8931Somatic comorbidities 148.6494At least one comorbidity9979.23772.5 None86.447.8 Treatment group 176.54231. TAU3024.01325.5 2. TAU + mianserin2822.41631.4 3. TAU + IRT3729.61121.6 4. TAU + mianserin + IRT3024.01121.6 1: Treatment As Usual (TAU), TAU + mianserin, TAU + Imagery Rehearsal Therapy (IRT), TAU + IRT + mianserin. Total sample in entire population analyzed, excluding missing data. Probabilities are those associated with Pearson’s chi-squared test.
There was a significant difference in the frequency of nightmares between the groups, with 68% of patients with DEB reporting weekly nightmares compared with 49% of patients without DEB (χ²(1 d.f., N = 176), p = .0087), as shown in Table 1. Table 2 presents the baseline scores of the questionnaires. Patients in the two groups had comparable scores on all questionnaires, including the PSQI, which reflects a similar sleep quality at baseline. Table 2.Baseline Questionnaire Scores. + DEB− DEB QuestionnaireNMedianIQRSDNMedianIQRSDpPSQI8617.05.02.83616.72.92.7.7735NES8031.08.05.72628.510.07.2.1233WHODAS5726.010.07.62127.015.08.3.4636SDS7923.08.05.33021.07.05.6.1169HTQ903.130.630.40373.060.780.44.2053PSQI: Pittsburgh Sleep Quality Index, NES: Nightmare Effects Survey, WHODAS: WHO Disability Assessment Schedule 2.0, SDS: Sheehan Disability Scale, HTQ: Harvard Trauma Questionnaire. N = number. IQR = interquartile range. SD = standard deviation. Probabilities are those associated with the Mann–Whitney U test.
Table 3 shows the changes in questionnaire scores from baseline to follow-up. After adjusting for sex, age, and allocated treatment group, we found that the presence of DEB was significantly associated with a poorer treatment response in overall sleep quality, as measured by the PSQI (for patients with DEB; estimate = 1.76, SE = 0.82, p = .0305). To check for bias due to missing values in the PSQI, we analyzed the three groups – patients with DEB (N = 86), patients without DEB (N = 36) and patients with missing data (N = 43) – by age, sex, and treatment randomization, but found no significant differences between the groups (Supplementary Table 5). Neither were any significant associations found between DEB and changes in scores in the other questionnaires. Table 3.Changes in Questionnaire Scores for Patients with DEB Compared with Patients without DEB from Baseline to Follow-up.QuestionnaireNEstimate for + DEBSEMpPSQI1221.760.82.0350NES106−1.642.16.4508WHODAS78−2.463.37.4672SDS109−0.911.75.6033HTQ1270.030.11.7958Data were analyzed by OLS regression adjusted for age, sex, and allocated treatment group. SEM = standard error of the mean.
Table 4 presents the distribution of nightmare themes between groups, in a subset of 89 patients asked to complete the modified TDQ. Among patients with DEB, 40.3% (N = 27) reported having nightmares featuring a replay of traumatic events, whereas 59.7% (N = 40) did not. RBD-like dreams were also reported by 40.3% (N = 27) of patients with DEB. In the patient group without DEB, 13.6% (N = 3) had trauma-related nightmares and 18.2% (N = 4) reported RBD-like dreams. Table 4.Reported Dream Content During the 12 Months of Treatment. + DEB(N = 67)− DEB(N = 22)Dream ContentYesN (%)NoN (%)YesN (%)NoN (%)Replay of Traumatic Events27 (40.3)40 (59.7)3 (13.6)19 (86.4)RBD-like Dreams27 (40.3)40 (59.7)4 (18.2)18 (81.8)Both Themes19 (28.4)48 (71.6)1 (4.5)21 (95.5)RBD-like dreams were defined as ‘Being chased’, ‘Physical aggression’ or ‘Insects, vermin’ (see Supplementary Material for further specification).
We found a prevalence of DEB of 71% in a population of 176 trauma-affected refugees diagnosed with PTSD, a proportion that is consistent with the findings of two recent studies reporting high prevalences (55% and 64%) of DEB in PTSD patients (Elliott et al., 2020; Lee et al., 2020). In our study, the presence of DEB was associated with diminished improvements in sleep quality post-treatment, as measured by the PSQI, irrespective of treatment group (1.76 points lower, p = .035). In a subsample of 67 patients with DEB, we found that 59.7% did not report nightmares featuring a replay of traumatic events. These findings provide insight into the relationship between traumatic experiences and alterations in normal sleep physiology.
DEB is believed to originate from the involvement of the brainstem regions responsible for motor control during sleep, regardless of whether the underlying pathology is neurodegeneration or hyperarousability. These areas overlap with regions involved in emotional processing, memory consolidation, and dreaming that are important in the development and maintenance of PTSD symptoms (Germain et al., 2008; Husain et al., 2001; Spoormaker & Montgomery, 2008). This overlap could explain the association of DEB with PTSD and trauma (Husain et al., 2001). The co-occurrence of trauma-related nightmares and DEB in individuals without daytime PTSD symptoms has led to the proposal of a condition known as Trauma-Associated Sleep Disorder (TASD) (Feemster et al., 2019; Mysliwiec et al., 2014). While TASD and PTSD share features and are often comorbid, their management differs (Jones et al., 2022). Case reports suggest that Prazosin can be effective for treating TASD, whereas IRT and CBT-I are considered first-line treatments addressing nightmares and insomnia in PTSD (Brock et al., 2019; Jones et al., 2022; Koffel et al., 2016). However, these therapeutic approaches may have limited efficacy in addressing DEB in this population, which may account for our observation of reduced improvements in sleep quality in patients with DEB following treatment (Mysliwiec et al., 2018; Talbot et al., 2014; Ulmer et al., 2011).
Our cohort, composed of younger, trauma-exposed patients, supports the notion that DEB in our population is more closely aligned with a trauma-induced sleep disorder than with a neurodegenerative condition. However, a substantial portion of patients with DEB did not meet the suggested TASD criteria because many of them did not have dreams linked to previous traumatic experiences – a finding commonly observed in PTSD patient populations (Phelps et al., 2011). Also, most of the patients in our population had their first traumatic experience many years before entering the study. Current TASD diagnostic criteria suggest a close temporal proximity between trauma and symptoms, with nightmares and disruptive behaviours subsiding over time (Feemster et al., 2019; Mysliwiec et al., 2014; Mysliwiec et al., 2018), even though a definition of temporal proximity is not included in current diagnostic criteria (Brock et al., 2019). This finding suggests that DEB in relation to PTSD is not fully captured by current sleep disorder classifications (Mysliwiec & Brock, 2020), thereby highlighting a need for further refinement of TASD diagnostic criteria, as other authors have suggested (Barone, 2020).
Baseline comparisons showed no significant differences in PTSD symptom severity or sleep quality between groups, suggesting that the occurrence of DEB was not associated with greater PTSD severity or sleep impairment prior to treatment. This contrasts with a recent study that found DEB to be associated with poorer overall sleep quality and higher PTSD symptom severity (Jones et al., 2020), although the latter association was not evident once sleep impairment had been controlled for. Another study examined the development of intrusive daytime memories across non-trauma-exposed, trauma-exposed, and PTSD patients (Ney et al., 2021). It found that DEB was associated with an increased likelihood of intrusive memories only in PTSD patients, an effect not fully explained by poorer sleep quality or increased nightmare frequency in the PTSD group (Ney et al., 2021). However, nightmares in PTSD have previously been associated with increased daytime and nighttime distress (Bertolazi et al., 2024; Campbell & Germain, 2016; Pigeon et al., 2013). In our study, we found that a greater proportion of patients exhibiting DEB reported weekly nightmares than did those without DEB (p = .0087), as expected from our classification criteria requiring vivid dreams. To explore whether the increased nightmare frequency in the DEB group could explain the difference in treatment response independently of DEB, we controlled for the occurrence of weekly nightmares in our regression analysis. However, our results were largely unchanged, indicating that the frequency of nightmares was not the primary factor driving the differential treatment response.
Persistent sleep disturbances in PTSD are well-documented and are associated with significant health risks, including suicidal ideation and behaviour (Bertolazi et al., 2024; Campbell & Germain, 2016; McCarthy et al., 2019; Pigeon et al., 2012; Pigeon et al., 2013; Pruiksma et al., 2016). Patients with PTSD and DEB may represent a treatment-resistant subgroup requiring tailored interventions specifically targeting DEB. Our findings support the explanation of a trauma-induced sleep disorder, highlighting the need for further research to clarify the role of DEB in PTSD and to develop effective management strategies for this population.
This study is one of the biggest investigations into symptoms of DEB within a population of PTSD patients, and is the first to explore the influence of these symptoms on treatment response. Our findings suggest that DEB may be associated with diminished improvements in sleep quality post-treatment, regardless of treatment choice. This association highlights the potential clinical importance of identifying DEB in PTSD patients because it could indicate the existence of a subgroup of patients who are more challenging to treat. The relatively large sample size (N = 176) strengthens the validity of this observation and highlights the need for tailored interventions in clinical practice.
However, our study focuses on a highly selected population of PTSD patients, pre-screened for significant sleep disturbances (PSQI > 8) and nightmares (responses of at least ‘a little’ on the Harvard Trauma Questionnaire Nightmare item), with an additional requirement to have completed the REM Sleep Behaviour Disorder Questionnaire (RBDSQ). This pre-selection introduces a potential selection bias, as it may have enriched our sample with patients particularly inclined to engage with sleep-related questionnaires, potentially inflating the observed prevalence of DEB. Consequently, it is difficult to generalize our findings to the broader PTSD population, especially in the case of patients with milder or no sleep difficulties.
Additionally, although the RBDSQ is one of the most widely used screening instruments for DEB, its validation has been limited to general and iRBD populations. There has been no formal validation in psychiatric populations, which represents a limitation on its reliability in this context. To address the concern that vivid dreams might drive classification, we evaluated the distribution of answers to Question 1 of the RBDSQ in the non-DEB group, examined collinearity between vivid dreams (Question 1) and motor behaviour during sleep (Question 6) and analyzed treatment responses in the non-DEB group based on patients having given a positive answer to Question 1 (Supplementary Table 2–4). We found a significant but weak correlation between the questions (Φ = 0.16, p = .03), and no difference in treatment response in the non-DEB group when based on vivid dreams alone. These findings suggest that the classification criteria capture distinct aspects of DEB: vivid dreams (enabling dream recall) and motor behaviour during sleep (enabling enactment behaviour). Studies exploring the intersection of PTSD, TASD, and RBD have employed various assessment methods for DEB, which makes direct comparison of results more challenging (Jones et al., 2020; Jones et al., 2022; Lee et al., 2020; Ney et al., 2021; Xie et al., 2024). This highlights the need for large-scale validation of tools for diagnosing DEB in PTSD populations, ideally incorporating multi-night PSGs for the objective characterization of DEB.
Nightmare themes were assessed through biweekly conversations with a psychologist, supplemented by prospective dream logs maintained by patients over the course of a year. This comprehensive and longitudinal approach is a significant strength underpinning these findings. However, most sessions were conducted with an interpreter rather than in the patients’ native language, presenting a limitation in accurately categorizing nightmare themes.
Lastly, although we recorded the time between the first traumatic experience and study entry, we did not have data about the time between the trauma and DEB onset. This information could have allowed additional insight into the progression of symptoms and their alignment with current sleep disorder classifications.
Despite these limitations, our study provides valuable insights into the relationship between traumatic experiences and alterations in sleep physiology, contributing to the discussion about the classification of sleep disorders in PTSD. Further research is needed to determine whether these findings hold in more general PTSD populations, and to refine our understanding of how DEB fits within the broader spectrum of PTSD-related sleep disturbances.