Emotion dysregulation mediates the association between acute sleep disturbance and later posttraumatic stress symptoms in trauma exposed adults

ABSTRACT

Background: Sleep disturbance is common within days to weeks following a traumatic event and has been associated with emotion dysregulation, a strong risk factor for PTSD development. This study aims to examine if emotion dysregulation mediates the relationship between early post-trauma sleep disturbance and subsequent PTSD symptom severity.

Methods: Adult participants (n = 125) completed questionnaires regarding sleep disturbance (via Pittsburgh Sleep Quality Index Addendum; PSQI-A) and emotion dysregulation (via Difficulties in Emotion Regulation Scale; DERS) within 2 weeks after exposure to traumatic events.

Results: PTSD symptom severity was assessed with PTSD Checklist for DSM-5 (PCL-5) at 3-month follow-up. There were strong correlations between PSQI-A, DERS, and PCL-5 (r ranges between .38 and .45). Mediation analysis further revealed significant indirect effects of overall emotion regulation difficulties in the relationship between sleep disturbance at 2 weeks and PTSD symptom severity at 3 months (B = .372, SE = .136, 95% CI: [.128, .655]). Importantly, limited access to emotion regulation strategies emerged as the single, significant indirect effect in this relationship (B = .465, SE = .204, 95% CI [.127, .910]) while modelling DERS subscales as multiple parallel mediators.

Conclusions: Early post-trauma sleep disturbance is associated with PTSD symptoms over months, and acute emotion dysregulation explains part of this association. Those with limited emotion regulation strategies are at particular risk of developing PTSD symptoms. Early interventions focusing on the appropriate strategies for emotion regulation may be crucial for trauma-exposed individuals.

HIGHLIGHTS

• Early post-trauma sleep disturbance and emotion dysregulation associated with subsequent posttraumatic stress symptoms.

• Emotion dysregulation mediates associations between acute sleep disturbance and later posttraumatic stress symptoms.

• Limited access to effective regulatory strategies was the single, significant contributor to this mediation link.

1. Introduction

Posttraumatic stress disorder (PTSD) is a common and debilitating psychiatric condition that may develop after exposure to a traumatic event. With a lifetime prevalence of 6.1% (Goldstein et al., 2016), PTSD is characterized by intrusive thoughts related to the event, avoidance of reminders of the event, negative mood and cognition, and heightened arousal and reactivity (American Psychiatric Association, 2013). In addition, individuals with PTSD often report symptoms of sleep disturbance, most notably insomnia and nightmares, with an estimated prevalence around 50%–90% (Koffel et al., 2016; Maher et al., 2006). Furthermore, it is also common for PTSD patients to exhibit other signs of sleep disturbance, such as periodic leg movements and disruptive nocturnal behaviours (Koffel et al., 2016; Ohayon & Shapiro, 2000).

Mounting evidence suggests that sleep disturbance plays a prominent role in the development and maintenance of PTSD in various ways (Babson & Feldner, 2010; Spoormaker & Montgomery, 2008). First, the presence of sleep disturbance before trauma exposure is associated with increased risk of PTSD development, irrespective of the mechanism of traumatic injury, severity of injury, or history of prior psychiatric illness (Bryant et al., 2010; van Liempt et al., 2013). In addition, multiple studies have indicated that sleep disturbance is strongly correlated to PTSD symptom severity, even after controlling for other psychiatric comorbidity and sociodemographic variables (Fairholme et al., 2013; Germain et al., 2004). Furthermore, numerous studies have shown that sleep disturbance as early as 2 weeks post-trauma can be used to predict PTSD diagnosis months to years later in a variety of trauma-exposed populations, such as motor vehicle accident victims (Klein et al., 2003; Koren et al., 2002; Mellman et al., 2001), foreign refugees (Lies et al., 2020), war veterans (Pigeon et al., 2013; Rosen et al., 2019), and natural disaster survivors (Fan et al., 2017).

On the other hand, emotion dysregulation is a multidimensional construct that refers to difficulty modulating one’s emotional experiences because of inability to effectively recognize, accept, or control emotions, especially in the context of goal-directed behaviours (Gratz & Roemer, 2004). Emotion dysregulation has been implicated in diverse psychopathologies, including PTSD. For example, Tull et al. (2007) reported a strong association between PTSD symptom severity and self-reported difficulties with multiple components of emotion regulation (i.e. lack of emotional acceptance, impulse control difficulties, and lack of access to effective emotion-regulation strategies), above and beyond simply the level of negative affect. Additionally, other prospective studies identified pre-trauma (Bardeen et al., 2013) and peri-trauma (Forbes et al., 2020; Fujisato et al., 2020; Pencea et al., 2020) emotion dysregulation as a predictor of PTSD symptom severity and development.

Emotion dysregulation has been linked to sleep disturbance in PTSD as well. Fairholme et al. (2013) demonstrated that both insomnia severity and emotion dysregulation contribute to PTSD symptom severity independently and together may act as transdiagnostic processes in a highly comorbid sample. Short et al. (2014) further revealed that emotion dysregulation moderates the relationship between sleep disturbance and PTSD symptom severity in a sample of trauma-exposed smokers. Similar interactive effect between sleep quality and emotion regulation on anxiety symptomatology has been reported in veterans with PTSD (Mantua et al., 2018). With an experimental design, Kleim et al. (2016) showed that healthy subjects who were randomly assigned to sleep immediately after laboratory trauma experienced fewer and less distressing intrusive emotional memories, when compared to those who stayed awake. The distress and recurrence of intrusive thoughts and memories, in turn, have been strongly linked to emotion dysregulation (Jungmann et al., 2016). Therefore, it is plausible that early post-trauma sleep deprivation gives rise to intrusive thoughts and, consequently, emotion dysregulation (Harrington & Cairney, 2021).

While the associations among sleep disturbance, emotion dysregulation, and PTSD have been established, to our knowledge, no study has examined the mediating role of emotion dysregulation in the relationship between early post-trauma sleep disturbance and subsequent PTSD symptom severity. Given the theoretical model that trauma-induced sleep loss causes emotion dysregulation (Harrington & Cairney, 2021; Murkar & De Koninck, 2018), we hypothesized that acute sleep disturbance following a traumatic event leads to early emotion dysregulation, which in turn contributes to the development and exacerbation of PTSD symptoms. We then further explored which aspects of emotion dysregulation would play a prominent role in this mediation relationship.

2. Methods

2.1. Participants and procedure

Adult participants were recruited from local hospital emergency departments (EDs) within 48 h following a traumatic event as part of a larger, ongoing longitudinal neuroimaging study. Participants were excluded from the larger study for 1) severe injuries (including traumatic brain injury), 2) requiring surgical interventions, 3) history of severe neuropsychiatric disorders, 4) being under the influence of alcohol or substances at the time of trauma, 5) MRI contraindications (e.g. pregnancy or metal implantation), or 6) inability to read or write in English.

Data from participants who completed questionnaires regarding sleep disturbance and emotion dysregulation within 2 weeks post-trauma, as well as a questionnaire regarding PTSD symptom severity at 3 months post-trauma, were analyzed in the current study. The study was approved by the university Institutional Review Board. All participants provided written informed consent and received monetary compensation for their participation in the study.

2.2. Measures

2.2.1. Sleep disturbance

The Pittsburgh Sleep Quality Index Addendum (PSQI-A; Germain et al., 2005) was used to assess participants’ sleep disturbance at early post-trauma period (i.e. within 2 weeks). The PSQI-A is a brief, 7-item self-report questionnaire designed to assess the frequency of 7 disruptive nocturnal behaviours, including: hot flashes, general nervousness, memories or nightmares of trauma, anxiety or panic not related to traumatic memories, bad dreams not related to traumatic memories, episodes of terror during sleep, and episodes of acting out dreams. Each item is rated on a 0–3 scale referring to frequency of each disturbance, where 0 = not in the past month, 1 = less than once a week, 2 = once or twice a week, and 3 = three or more times a week. The PSQI-A has demonstrated adequate psychometric properties (Germain et al., 2005; Insana et al., 2013). In the current study, the PSQI-A displayed acceptable internal consistency (Cronbach's α = 0.78).

2.2.2. Emotion dysregulation

The Difficulties in Emotion Regulation Scale (DERS; Gratz & Roemer, 2004) was used to assess participants’ emotion dysregulation at early post-trauma period (i.e. within 2 weeks). The DERS is a multidimensional 36-item questionnaire that assesses six domains of emotion dysregulation, including: nonacceptance of emotion (i.e. NONACCEPT), difficulties engaging in goal-directed behaviours (i.e. GOALS), impulse control difficulties (i.e. IMPULSE), limited access to effective regulatory strategies (i.e. STRATEGIES), reduced emotional clarity (i.e. CLARITY), and a lack of emotional awareness (i.e. AWARENESS). Higher scores reflect greater difficulties regulating emotion across each domain. The DERS demonstrated good internal consistency in the current sample (Cronbach's α = 0.83).

2.2.3. PTSD symptom severity

The PTSD Checklist for DSM-5 (PCL-5; Weathers et al., 2013) was used to assess participants’ PTSD symptom severity at 3 months post-trauma. The PCL-5 is a 20-item self-report measure that examines the severity of PTSD symptom clusters as outlined in the fifth edition of the Diagnostic and Statistical Manual of Mental Disorders. The PCL-5 has shown adequate psychometric properties (Bovin et al., 2016; Weathers et al., 2013) and achieved good internal consistency in the current study (Cronbach's α = 0.85).

2.3. Data analysis

First, partial correlation was used to evaluate bivariate linear relationships among sleep disturbance, emotion dysregulation, and PTSD symptom severity while controlling for age and sex. Next, we fit a simple mediation model to test the extent to which emotion dysregulation mediates the association between sleep disturbance and PTSD symptom severity, while controlling for age and sex. We then fit a multiple mediation model where DERS subscores were parallel mediators for the association between sleep disturbance and PTSD symptom severity. Previous studies suggested that age and sex may influence emotion regulation (Tull et al., 2007) and PTSD (Fujisato et al., 2020), so these contextual factors were considered as covariates. All analyses were conducted using SPSS Version 27 (IBM Corp., Armonk, NY) with PROCESS macro developed by Hayes (2017). Statistical significance level was set at α = .05, two-tailed. For mediation analysis, 95% bootstrapped confidence interval (CI) with 5000 samples was used to evaluate the significance of mediation (i.e. indirect) effect. A significant mediation effect was reported if the 95% CI did not include zero.

3. Results

Participants demographic information is presented in Table 1. One hundred and twenty-five participants (mean age = 33.49 ± 10.62 years) completed assessments of sleep disturbance and emotion dysregulation within 2 weeks post-trauma, as well as PTSD symptoms at 3 months post-trauma. The majority of the participants were female (68%), African American (47.2%), and admitted to EDs due to motor vehicle accident (56.8%). None of the participants reported taking medications or receiving psychotherapy for mental health conditions.

Table 1.

Participants’ demographic information.

	Mean (SD) or n (%)
Age	33.49 (10.62)
Sex (Female)	85 (68%)
Race
White / Caucasian	56 (44.8%)
Black / African American	58 (47.2%)
Other	11 (8%)
Trauma Type
Motor Vehicle Accident	71 (56.8%)
Physical Assault	41 (32.8%)
Sexual Assault	9 (7.2%)
Other	4 (3.2%)
Probable PTSD at 3 months post-trauma*	76 (60.8%)

Note. PTSD = posttraumatic stress disorder. * Probable PTSD was defined using a cut-off score of 33 or higher on PTSD Checklist for DSM-5 assessed 3 months post-trauma.

Bivariate partial correlations among variables of interest (i.e. sleep disturbance, emotion dysregulation, and PTSD symptom severity) are presented in Table 2. PTSD symptom severity was significantly correlated with sleep disturbance, emotion dysregulation total score and subscores (all p < .05), except the AWARENESS subscale. Similarly, sleep disturbance was significantly correlated with emotion dysregulation total score and subscores (all p < .05), except AWARENESS subscale. DERS AWARENESS subscale was significantly correlated with DERS total score (p < .05) but not with any other DERS subscales (all p > .05).

Table 2.

Partial correlations among variables of interest, controlling for age and sex.

Measures	1.	2.	3.	4.	5.	6.	7.	8.	9.	Mean (SD)
1. PCL-5	–									37.0 (18.6)
2. PSQI-A	.45**	–								9.6 (5.1)
3. DERS	.41**	.38**	–							102.4 (26.7)
4. DERS – NONACCEPT	.37**	.28*	.74**	–						17.0 (5.8)
5. DERS – GOALS	.37**	.43**	.82**	.58**	–					16.2 (5.1)
6. DERS – IMPULSE	.33**	.30**	.80**	.55**	.61**	–				15.6 (6.8)
7. DERS – AWARENESS	-.01	-.06	.44**	.06	.19	.15	–			17.0 (5.8)
8. DERS – STRATEGIES	.47**	.38**	.87**	.67**	.70**	.65**	.19	–		23.1 (7.8)
9. DERS – CLARITY	.27*	.34**	.77**	.39**	.59**	.53**	.48**	.56**	–	13.6 (4.6)

Note. * = p < .01. ** = p < .001. PCL-5 = PTSD Checklist for DSM-5. PSQI-A = Pittsburgh Sleep Quality Index Addendum. DERS = Difficulties in Emotion Regulation Scale total score. DERS – NONACCEPT = DERS Nonacceptance of Emotion subscale score. DERS – GOALS = DERS Inability to Engage in Goal-Directed Behavior When Distressed subscale score. DERS – IMPULSE = DERS Impulse Control Difficulties subscale score. DERS – AWARENESS = DERS Lack of Emotional Awareness subscale score. DERS – STRATEGIES = DERS Lack of Effective Emotion Regulation Strategies subscale score. DERS – CLARITY = DERS Lack of Emotional Clarity subscale score. PSQI-A and DERS were taken at 2 weeks post-trauma. PCL-5 was taken at 3 months post-trauma.

In the simple mediation model (Figure 1) after controlling for age and sex, the indirect effect of emotion dysregulation on the association between sleep disturbance and PTSD symptom severity was significant (B_ab = .372, SE = .136, 95% CI [.128, .655]). This model accounted for a significant amount of variance in PTSD symptom severity (R² = .348, p < .01). The total effect between sleep disturbance and PTSD symptom severity was significant (B_c = 1.579, SE = .283, 95% CI [1.018, 2.140]), and the direct effect between sleep disturbance and PTSD symptom severity when holding emotion dysregulation level constant was also significant (B_c’ = 1.207, SE = .294, 95% CI [.626, 1.788]), suggesting partial mediation.

Figure 1.

Mediating role of emotion dysregulation on sleep disturbance at 2 weeks and PTSD symptoms at 3 months follow-up, controlling for age and sex. p < .001 for all effects.

In the multiple parallel mediation model with emotion dysregulation subscores as mediators, the total mediation effect was significant (B_Total = .515, SE = .217, 95% CI [.147, 1.013]), but only the indirect effect of STRATEGIES subscale was significant (B_Strategies = .465, SE = .204, 95% CI [.127, .910]). This model accounted for a significant amount of variance in PTSD symptom severity (R² = .330, p < .01). The total effect between sleep disturbance and PTSD symptom severity was significant (B_c = 1.688, SE = .294, 95% CI [1.107, 2.269]), and the direct effect between sleep disturbance and PTSD symptom severity when holding emotion dysregulation level constant was also significant (B_c’ = 1.173, SE = .323, 95% CI [.533, 1.813]), suggesting partial mediation. These findings are presented in Table 3.

Table 3.

Indirect effects of multiple parallel mediation model.

	Indirect Effect	Standard Error	95% CI
Total	.515	.217	[.147, .1.013]
DERS – NONACCEPT	.127	.119	[-.097, .378]
DERS – GOALS	-.072	.185	[-.423, .307]
DERS – IMPULSE	.009	.130	[-.287, .251]
DERS – AWARENESS	.011	.042	[-.060, .117]
DERS – STRATEGIES	.465	.204	[.127, .910]
DERS – CLARITY	-.006	.163	[-.326, .337]

Note. Significant effects (i.e. 95% CI not included 0) are bolded. DERS = Difficulties in Emotion Regulation Scale total score. DERS – NONACCEPT = DERS Nonacceptance of Emotion subscale score. DERS – GOALS = DERS Inability to Engage in Goal-Directed Behavior When Distressed subscale score. DERS – IMPULSE = DERS Impulse Control Difficulties subscale score. DERS – AWARENESS = DERS Lack of Emotional Awareness subscale score. DERS – STRATEGIES = DERS Lack of Effective Emotion Regulation Strategies subscale score. DERS – CLARITY = DERS Lack of Emotional Clarity subscale score.

4. Discussion

The current study aimed to fill a knowledge gap in the literature by investigating the extent to which emotion dysregulation, as well as its specific features, mediates the association between sleep disturbance and PTSD symptoms. Previously, associations among sleep disturbance, emotion regulation, and PTSD have been established (Bardeen et al., 2013; Forbes et al., 2020; Koren et al., 2002; Pencea et al., 2020; Wright et al., 2011). Furthermore, emotion dysregulation has been shown as a transdiagnostic and moderating factor in the relationship between sleep disturbance and PTSD development (Fairholme et al., 2013; Short et al., 2014). To our knowledge, the current study is the first to demonstrate emotional dysregulation as a mediator of acute post-trauma sleep disturbance and subsequent development of PTSD symptoms. A similar mediating role of emotion dysregulation has been reported in other psychiatric conditions, including psychosis (Akram et al., 2020), depression (Hom et al., 2016), and borderline personality disorder (Grove et al., 2017). Finally, when examining facets of emotion dysregulation, we found that limited access to effective emotion regulation strategies is likely an underlying mechanism of the association between sleep disturbance and PTSD symptom severity.

Our findings provide support for a theoretical model proposed by Harrington and Cairney (2021), who speculate that acute post-trauma insomnia leads to intrusive thoughts and eventually to emotion dysregulation. However, it is difficult to empirically assess the direction of causality. Perhaps, trauma exposure disrupts sleep and emotion regulation simultaneously, rather than sequentially. Nonetheless, current literature indicates that poor sleep precedes PTSD but not the reverse (Babson & Feldner, 2010; Wright et al., 2011), and that acute sleep deprivation negatively impacts emotion regulation within days in both laboratory (Gordon & Chen, 2014) and real world setttings (Zohar et al., 2005). As early symptoms progress, the self-perpetuating cycle of disturbed sleep, intrusive thoughts, and emotion dysregulation creates a vicious circle leading to development and maintenance of chronic PTSD (Harvey et al., 2011). Interestingly, emotional awareness is not correlated with sleep disturbance or PTSD symptom severity in the current study, which aligns with previous results (Pickett et al., 2016; Şandru & Voinescu, 2014). There were no significant correlations between emotional awareness and other facets of emotion dysregulation, in contrast to the observed correlations among other DERS subscales and the total score. This suggests that awareness may not be an essential element of emotion regulation, as proposed in a revised DERS model by Bardeen et al. (2012). However, other studies that measure emotional awareness with DERS (Short et al., 2014) or its related construct (Nagy et al., 2020; Smith et al., 2020) have reported significant association between awareness, sleep, and PTSD, suggesting the need for further research to elucidate the mixed findings.

When examining which components of emotion dysregulation drive this mediational link, we showed that limited access to effective regulatory strategies emerged as the single, significant contributor to the relationship. Similar results were reported in depression (Hom et al., 2016), psychosis (Akram et al., 2020), and suicidality (Serrano et al., 2021). Furthermore, individuals with limited strategies specifically endorse a high level of subjective sleep dysfunction (Şandru & Voinescu, 2014). Together, these findings support the notion that poor sleep quality decreases one’s cognitive reappraisal and ability to regulate negative emotions (Mauss et al., 2013). Regarding PTSD, studies have linked several domains of emotion dysregulation, including lack of strategies, difficulty with emotional clarity, and problems with impulse control, to increased PTSD symptom severity (Short et al., 2014; Tull et al., 2007). Further investigation into sleep disturbance by Short et al. (2014) showed that difficulties with emotional clarity, awareness, and impulse control were specifically associated with greater effect of insomnia on PTSD symptom severity, which disagrees with our findings. These inconsistent findings between Short et al. (2014) and the current study may have occurred because the former focused on insomnia whereas the PSQI-A used in the current study measures disruptive nocturnal behaviours and is specifically designed for a trauma-exposed population. Further, Fujisato et al. (2020) argue that merely identifying undesired emotions and their causes (i.e. emotional clarity and awareness) worsens PTSD symptoms, but knowing how to identify and deal with negative emotion (i.e. adaptive responses and strategies toward emotions) improves PTSD symptoms over time. Therefore, we speculate that adaptive and maladaptive strategies would appear as the most important explanatory factor between sleep disturbance and PTSD symptom severity in the current study.

The current study highlights the importance of early identification and treatment of emotion dysregulation in trauma-exposed individuals. It is reasonable to develop and implement a short, effective emotion dysregulation screening questionnaire that focuses on maladaptive strategies for trauma-exposed individuals. Such a questionnaire would allow for timely risk stratification and brief intervention for vulnerable individuals during their early medical encounters. Several studies (Bryant et al., 2013; Cloitre et al., 2002; Ford et al., 2018) support the incorporation of emotion regulation training into cognitive behavioural therapies in PTSD patients, but evidence is mixed (Van Toorenburg et al., 2020) and requires further validation in the early (i.e. days-to-weeks) post-trauma period. Concurrently or alternatively, interventions for insomnia and nightmares are also beneficial (Galovski et al., 2016; Koffel et al., 2016; Maher et al., 2006).

Notably, the current study chose the framework of mediation analysis based on the theoretical model proposed by Harrington and Cairney (2021). Yet, other alternative models on the associations among early sleep disturbance, emotion dysregulation, and subsequent PTSD symptoms may also exist. For example, emotion dysregulation may moderate relationships between early sleep disturbance and subsequent PTSD symptom severity. In addition, due to the difficulty determining the sequence of symptom development, it is plausible that sleep disturbance mediates early emotion dysregulation and subsequent PTSD symptom severity. We conducted exploratory analyses to address this issue. First, we fit the data with moderation model and showed that emotion dysregulation did not moderate the relationship between early sleep disturbance and subsequent PTSD symptom severity in the current sample (Supplemental Table 1). Second, we tested an alternative mediation model and the results showed that sleep disturbance partially mediates early emotion dysregulation and subsequent PTSD symptom severity (Supplemental Table 2). Yet, it should be noted that the PSQI-A assesses disruptive nocturnal behaviours during the previous month whereas DERS assesses domains of emotion dysregulation during the previous two weeks. Therefore, the original mediation model tested in the primary analyses is more likely than the alternative mediation model tested in the exploratory analyses.

The current study should be considered in the context of its limitations. First, sleep disturbance was measured subjectively without clinical assessment. Although self-reported sleep problems with validated questionnaires, such as PSQI-A, are well-established in PTSD research, the objective evidence of sleep disturbance is mixed (Babson & Feldner, 2010; Koffel et al., 2016; Lewis et al., 2020; Richards et al., 2020). Therefore, future research should consider multimethod designs to ensure objective measures of sleep disturbance, such as polysomnography and actigraphy. Second, the design of the current observation study does not allow for causal conclusions. Future research may focus on exploring the early post-trauma stages of emotion dysregulation and sleep disturbance, perhaps by developing an experimental design for empirical support. Third, it is unclear to what extent if any the pre-trauma emotion dysregulation affected the development of PTSD in our sample, as there was no assessment of baseline emotion dysregulation prior to traumatic exposure (Bardeen et al., 2013). Lastly, the current study only elucidates parts of the whole picture. For example, a recent meta-analytic review revealed a moderate relationship between dissociation and emotion regulation, specifically to the maladaptive domains but not to the adaptive domains of emotion regulation (Cavicchioli et al., 2021). Therefore, future studies integrating pre-existing and peritraumatic emotion dysregulation and sleep quality, dissociative symptoms, and other relevant factors are warranted.

Despite these limitations, in the current study we extend the literature on the relationship between early sleep disturbance and emotional dysregulation and development of PTSD symptoms, and we demonstrate the mediating role of emotion dysregulation and maladaptive strategies in these relationships. These findings indicate the importance of early post-trauma screening and identification of those with sleep disturbance and emotion dysregulation. Continued research into effective strategies and early interventions targeting these risk factors is crucial for better short- and long-term health outcomes.

Funding Statement

This work was supported by National Institute of Mental Health [Grant Number R01MH110483].

Disclosure statement

No potential conflict of interest was reported by the author(s).