The Role of Emotion Dysregulation in the Relationship between Sleep Disturbances and PTSD Symptom Severity

Abstract

Emotion dysregulation is implicated in the development, maintenance, and treatment of sleep disturbances and posttraumatic stress disorder (PTSD) separately. However, few studies have assessed interactions among these variables. To address this gap, this study examined whether the associations of sleep quality and sleep quantity with PTSD severity were stronger at higher levels of negative and positive emotion dysregulation in a community sample of 199 trauma-exposed individuals. This study found both poorer sleep quality and lower sleep quantity were associated with greater PTSD severity at low to average (but not high) levels of negative emotion dysregulation. Positive emotion dysregulation did not moderate the relationships between sleep quality or quantity and PTSD severity. Exploratory additive multiple moderation analyses showed significant associations between poorer sleep quality and lower sleep quantity with greater PTSD severity at low to average levels of negative emotion dysregulation, coupled with any level of positive emotion dysregulation. Findings inform theoretical perspectives on the sleep-PTSD relationship and clinical applications of targeting emotion dysregulation in the treatment of sleep disturbances and PTSD symptoms for trauma-exposed individuals.

Introduction

The comorbidity between sleep disturbances and posttraumatic stress disorder (PTSD) is well documented (Babson & Feldner, 2010; Spoormaker & Montgomery, 2008). Affective processes, such as emotion dysregulation, play a role in the development, maintenance, and treatment of both sleep disturbances (e.g., Baglioni et al., 2010) and PTSD (e.g., Tull et al., 2020). Critically, no study to our knowledge has examined how negative and positive emotion dysregulation influence the nature and strength of the relationships between sleep disturbances and PTSD severity among trauma-exposed individuals. This was the aim of the current study.

Sleep Disturbances and PTSD Symptoms

Complaints about sleep can manifest in multiple ways, such as poor sleep quality (e.g., difficulty initiating or maintaining sleep, nonrestorative sleep) and lower sleep quantity (i.e., shorter duration of total time asleep; American Psychiatric Association, 2013). About 30% of adults report at least one characteristic of poor sleep quality (Roth, 2007) or chronic sleep deprivation (i.e., sleeping less than seven hours in a 24-hour time period; Krueger & Friedman, 2009). Evidence suggests that sleep disturbances are a risk factor for the development and maintenance of PTSD (DeViva et al., 2021). Pre-trauma sleep problems may lead to impairments in emotional processes, which may make a traumatic event more difficult to process and thus increase risk for PTSD (Short et al., 2020; Spoormaker & Montgomery, 2008). Further, post-trauma sleep disturbances may interact with and maintain PTSD symptoms (Galovski et al., 2016). For instance, sleep disturbances may be a biologically adaptive way to maintain hypervigilance, helping individuals with PTSD symptoms stay awake to manage anticipated threat regardless of sleepiness (Harvey et al., 2003). Further, other trauma-related behaviors, such as the avoidance of sleep-related trauma reminders (e.g., darkness) could also impair sleep. Poor sleep leading to an impairment in daytime functioning may increase physiological responsiveness to and avoidance of trauma-related stimuli (Babson & Feldner, 2010; Spoormaker & Montgomery, 2008), preventing therapeutic exposure to and extinction of learned fear responses to such stimuli (Pace-Schott et al., 2015). Abundant longitudinal empirical evidence shows that sleep disturbances predict PTSD diagnostic status and greater PTSD severity among trauma-exposed samples, such as active duty soldiers (Wang et al., 2019), combat veterans (Rosen et al., 2019), medical patients (Bryant et al., 2010), and natural disaster survivors (Geng et al., 2019). Despite this relationship, clinical research is mixed with regard to the efficacy of sleep disorder treatments in reducing PTSD symptom severity. For instance, Talbot et al. (2014) found in a randomized control trial that both participants in the waitlist control group and the Cognitive Behavioral Therapy for Insomnia (CBT-I) group experienced improvements in PTSD symptom severity, making it difficult to draw conclusions about the effect of CBT-I on PTSD symptoms. Moreover, in their systematic review, Taylor and Pruiksma (2014) noted similar difficulty drawing such conclusions, due to possible confounding effects of trauma-related nightmare treatment in the reviewed studies.

Emotion dysregulation may play an important role in influencing the relationships between sleep disturbances and PTSD symptoms (Kahn et al., 2013; Richards et al., 2020). Emotion dysregulation encompasses (a) a lack of awareness/understanding of emotions; (b) difficulty with behavioral control in the context of emotions; (c) difficulty using suitable strategies to change the duration and/or level of emotional responses to engage in goal-directed behavior; and (d) an unwillingness to experience emotions in the context of pursuing important/meaningful activities (Gratz & Roemer, 2004). The vast majority of literature has focused on negative emotion dysregulation (i.e., emotion dysregulation in response to negative emotions [e.g., fear, anger]) more generally (Gratz & Tull, 2010), in relation to sleep (e.g., Pickett et al., 2016; Şandru & Voinescu, 2014), and in relation to PTSD (Seligowski et al., 2015). Less empirical work has focused on positive emotion dysregulation (Weiss et al., 2020b), which references experiences of emotion dysregulation when experiencing positive emotions, such as love and happiness. As described by Weiss et al. (2015), positive emotion dysregulation encompasses a nonacceptance of positive emotions, difficulty engaging in goal-directed behaviors when experiencing positive emotions, and difficulty managing impulsive behavioral responses when experiencing positive emotions. Ultimately, a better understanding of the role of both negative and positive emotion dysregulation is vital to the improvement and effectiveness of clinical treatments aimed at reducing comorbid sleep disturbances and PTSD symptom severity (Geng et al., 2019).

Emotion Dysregulation and Sleep Disturbances

Current neurocognitive models theorize that sleep disturbances impact emotion dysregulation via deficits in cognitive functioning (Gruber & Cassoff, 2014). Specifically, sleep maintains higher-level cognitive functions, such as cognitive control (which encompasses executive functions like reasoning, inhibition, and cognitive flexibility); this, in turn, influences adaptive emotion regulation (Mauss et al., 2013; Tucker et al., 2010). Thus, sleep disturbances may result in deficits in these cognitive functioning processes (i.e., “executive dysfunction”), which can have detrimental impacts on an individual’s ability to adaptively regulate intense emotions (Gruber & Cassoff, 2014). In addition, downstream effects of emotion dysregulation may contribute to sleep disturbances. Indeed, emotion dysregulation is related to other maladaptive changes in emotional reactivity (Pickett et al., 2016); these changes in emotional responses, like more negative self-reported ratings of aversive emotional stimuli (Wagner et al., 2002), greater negative affect/worsening of mood (Kahn-Greene et al., 2007; Tsypes et al., 2013), and a decrease in positive affect (Tsypes et al., 2013) all relate to sleep disturbances (e.g., poor sleep quality, sleep deprivation). Further, emotion dysregulation during the daytime can impair sleep at night, which can subsequently impair next-day emotion regulation (Schmidt et al., 2011). Sleep disturbances, such as poor sleep quality and lower sleep quantity, have been associated with negative emotion dysregulation across multiple samples, such as undergraduates (Pickett et al., 2016), community members (Şandru & Voinescu, 2014), outpatients (Palagini et al., 2017), and substance use treatment inpatients (Fairholme et al., 2013). For instance, in an outpatient sample, Palagini et al. (2017) found that individuals with (vs. without) clinically significant insomnia symptoms reported greater negative emotion dysregulation. Şandru and Voinescu (2014) similarly found, in a sample of community members, those with poor (vs. good) sleep quality had significantly greater negative emotion dysregulation overall and in the specific domains of negative emotion dysregulation (except lack of awareness of emotional reactions). In addition, negative emotion dysregulation has been found to mediate the relationship between a generalized anxiety disorder (GAD) diagnosis and various indicators of sleep disturbances (independent of the effects of depression and other anxiety disorder symptom severity; Tsypes et al., 2013), which may be relevant to PTSD considering the overlapping symptomatology of GAD and PTSD (Price & van Stolk-Cooke, 2015). Notably, research on positive emotion dysregulation and sleep is quite limited. Talbot et al. (2009) found that difficulty regulating positive emotions contributed to more disturbed sleep (e.g., longer sleep onset latency) in those with bipolar disorder (vs. healthy controls). Gujar et al. (2011) found that sleep deprivation was related to imbalanced positive emotional reactivity, such that sleep-deprived (vs. non-sleep-deprived) individuals experienced amplified neural reactivity in response to emotionally positive stimuli. Further, Baglioni et al. (2010) found that poor sleep quality was related to increases in negative emotion and decreases in positive emotion in clinical and non-clinical samples. Importantly, no empirical research to our knowledge has examined the relationship between positive emotion dysregulation and both poor sleep quality and lower sleep quantity in trauma-exposed individuals.

Emotion Dysregulation and PTSD Symptoms

Emotion dysregulation is involved in the development and maintenance of PTSD symptomatology (Seligowski et al., 2015; Tull et al., 2007; Weiss et al., 2013). Theoretically, post-trauma, individuals experience various intense emotions and related psychophysiological distress (e.g., fear of trauma reminders; American Psychiatric Association, 2013). These individuals may perceive intense emotions as overpowering or frightening, may experience dampened emotional clarity, and/or may fear physiological hyperarousal experienced in response to negative emotions (which may expand to positive emotions via stimulus generalization; Tull et al., 2020; Weiss et al., 2019b). Such reactions foster emotion dysregulation (Tull et al., 2020). Subsequently, PTSD symptoms can develop as individuals may fear (and eventually avoid) trauma-related stimuli and associated intense emotional reactions via various strategies (e.g., avoiding certain trauma-related situations, attempting to escape emotions through substance use; Tull et al., 2020). In addition, emotion dysregulation can maintain PTSD symptoms. As outlined in the acceptance-based model of emotion regulation (Gratz & Roemer, 2004), emotion regulation strategies that involve the repeated avoidance or control of emotions are maladaptive. In the context of PTSD, such avoidance of trauma-related stimuli associated with upsetting emotional reactions and/or fear of arousal prevents exposure to such stimuli. This pattern could ultimately maintain PTSD symptoms as exposure to corrective information and healthy emotional processing is missed (Tull et al., 2020). Substantial empirical evidence shows PTSD severity relates to negative emotion dysregulation (Tull et al., 2007; Weiss et al., 2013) and to positive emotion dysregulation (Weiss et al., 2019b; Weiss, Dixon-Gordon, et al., 2018).

Current Study

There are notable limitations in the existing literature. Most studies focus on negative emotion dysregulation in relation to sleep disturbances and PTSD (Tsypes et al., 2013; Tull et al., 2007). Although more studies have begun to examine PTSD symptomatology in relation to positive emotion dysregulation (e.g., Weiss et al., 2020a), no empirical research to our knowledge has examined the relationship between sleep disturbances and positive emotion dysregulation in trauma-exposed individuals. Difficulties with positive emotion regulation relate to behavioral disturbances (e.g., impulse control issues, increased substance use; Weiss et al., 2019b), making this construct an important clinical target in the context of both sleep and PTSD. In addition, studies examining sleep disturbances as related to affective processes like emotion dysregulation are commonly limited to undergraduate samples and findings may not generalize to other diverse samples (Richards et al., 2020). Addressing these limitations and drawing from the abovementioned literature, this study examined whether negative and positive emotion dysregulation moderated the relationship between subjective sleep disturbances (quality and quantity) and PTSD symptom severity in a broader community sample of trauma-exposed individuals.

First, based on prior theory and empirical evidence (DeViva et al., 2021; Tull et al., 2020), the current study hypothesized significant main effects, such that greater negative emotion dysregulation, greater positive emotion dysregulation, poorer sleep quality, and lower sleep quantity would significantly relate to greater PTSD symptom severity. Broadly, trauma-exposed individuals with sleep disturbances develop or experience greater emotion dysregulation (Gruber & Cassoff, 2014). These emotion dysregulation processes, such as experiencing intense emotions as distressing (Seligowski et al., 2015; Tull et al., 2007), avoiding emotional arousal (Weiss et al., 2018), and/or engaging in reckless behaviors when experiencing intense emotions (Weiss et al., 2012) may exacerbate PTSD symptoms (Tull et al., 2020). Thus, regarding our primary research questions, this study hypothesized amplifying moderating effects of negative and positive emotion dysregulation on the relationship between sleep disturbances and PTSD symptom severity. Specifically, this study expected stronger significant associations between poorer sleep quality and greater PTSD symptom severity at higher (vs. lower) levels of negative emotion dysregulation (Hypothesis 1) and of positive emotion dysregulation (Hypothesis 2). Similarly, this study expected stronger significant associations between lower sleep quantity and greater PTSD symptom severity at higher (vs. lower) levels of negative emotion dysregulation (Hypothesis 3) and of positive emotion dysregulation (Hypothesis 4). As exploratory analyses, this study examined if negative and positive emotion dysregulation (both moderators in the same model) influenced the nature and strength of the relationships between poorer sleep quality and greater PTSD symptom severity (Exploratory Hypotheses 1), and between lower sleep quantity and greater PTSD symptom severity (Exploratory Hypotheses 2).

Methods

Procedure

The data utilized in the current study were collected by the first author as part of a parent study examining positive autobiographical memory processes and emotional distress among trauma-exposed individuals. Participants were recruited via Amazon’s Mechanical Turk (MTurk), which produces reliable data of high quality (Shapiro et al., 2013) and is a time- and cost-effective way to conduct research on difficult-to-reach trauma-exposed individuals with PTSD symptoms (van Stolk-Cooke et al., 2018). To be included in the study, participants were asked to confirm that they (1) were ≥ 18 years of age, (2) were fluent in English, (3) experienced a traumatic event (Primary Care PTSD Screen for DSM-5; Prins et al., 2016), and (4) had never received a sleep apnea diagnosis. Four validity check items were included to confirm comprehension of and attention to the survey (e.g., Thomas & Clifford, 2017). An example of a validity check item included: “Please ignore the response options one through ten below, and instead please select the small circle at the bottom of the screen.” Eligible participants who consented to participate, met all four inclusion criteria, and passed all validity check items received $1.50 in compensation.

Measures

Life Events Checklist for DSM-5 (LEC-5; Weathers, Blake, et al., 2013).

The LEC-5 is a 17-item self-report measure of exposure to lifetime traumatic events. Participants rated their degree of exposure using these response options: happened to me, witnessed it, learned about it, part of my job, not sure, or doesn’t apply. An additional 18^th item asked participants to indicate their most distressing traumatic event. Endorsement of a traumatic event was characterized by a participant’s choosing of any of the first four response options on items 1–16.

PTSD Checklist for DSM-5 (PCL-5; Weathers, Litz, et al., 2013).

The PCL-5 is a 20-item self-report measure of PTSD symptom severity over the past month. Participants referenced the most distressing traumatic event endorsed on the LEC-5 and responded to items using a 0 (not at all) to 4 (extremely) scale. In this study, the two items assessing sleep-related PTSD symptoms (i.e., item 2 [“Repeated, disturbing dreams of the stressful experience?”] and item 20 [“Trouble falling or staying asleep?”]) were removed from the PCL-5 total score to reduce the potential of multicollinearity effects (Taylor et al., 2020). The PCL-5 demonstrates excellent psychometric properties (Forkus et al., 2022). Cronbach’s α for the PCL-5 total score was .95 in this study.

Brief Version of the Difficulties in Emotion Regulation Scale (DERS-16; Bjureberg et al., 2016).

The DERS-16 is a 16-item self-report measure of dysregulation of negative emotions with good psychometric properties (Hallion et al., 2018). While referring to the past month, participants responded to items using a 1 (almost never; 0–10% of the time) to 5 (almost always; 91–100% of the time) scale. Cronbach’s α for the DERS-16 total score was .97 in this study.

Difficulties in Emotion Regulation Scale – Positive (DERS-P; Weiss, Gratz, et al., 2015).

The DERS-P is a 13-item self-report measure of dysregulation of positive emotions with good psychometric properties (Weiss et al., 2019a). While referring to the past month, participants responded to items using a 1 (almost never; 0–10% of the time) to 5 (almost always; 91–100% of the time) scale. Cronbach’s α for the DERS-P total score was .97 in this study.

Pittsburgh Sleep Quality Index (PSQI; Buysse et al., 1989).

The PSQI is a 21-item self-report measure of past month sleep disturbances with good psychometric properties (Beck et al., 2004). In the current study, the subjective sleep quality component score was used as a measure of sleep quality. Notably, this variable was recoded for ease of interpretation, where lower scores reflected poorer sleep quality (i.e., sleep quality was recoded using a 0 [very bad] to 3 [very good] scale). Further, this study used participants’ reported sleep duration as a measure of sleep quantity (continuous variable); participants reported the number of hours of sleep they obtained per night on average during the past month (Buysse et al., 1989). Empirical evidence supports examining the PSQI component scores individually (Grandner et al., 2006).

Exclusions and Missing Data

The original sample consisted of 695 participants. From this sample, this study excluded 229 participants who attempted to answer the survey more than once, 89 participants who did not meet all inclusion criteria, and 138 participants who failed one or more validity checks. This study further excluded one participant who did not endorse at least one of the first four response options on any LEC-5 items 1–16, and 30 participants who selected “none of these events happened to me” when asked to indicate their worst/most distressing traumatic event on the LEC-5. This study additionally excluded 8 participants who provided invalid responses on the measure of sleep quantity (i.e., values which greatly exceeded the possible maximum score; e.g., 400 hours of sleep obtained per night on average during the past month). Finally, this study excluded one participant missing > 30% of item-level data on study measures of interest. There was no missing data for any measures of interest in the final sample of 199 participants. In the final sample, the average age was 35.48 (SD = 11.43), and 119 participants (59.80%) identified as female. See Table 1 for additional information on demographic variables.

Table 1.

Demographic Information

Variables	Mean (SD)
Age	35.48 (11.43)
Years of education	15.44 (2.41)
Variables	n (%)
Gender Identity
Male	78 (39.20%)
Female	119 (59.80%)
Transgender	1 (.50%)
Gender Queer	1 (.50%)
Agender	0 (0%)
Other	0 (0%)
Current Employment Status
Part-time	32 (16.10%)
Full-time	140 (70.40%)
Retired	4 (2.00%)
Unemployed	17 (8.50%)
Unemployed Student	6 (3.00%)
Income Level
< $15,000	17 (8.50%)
$15,000 – $24,999	27 (13.60%)
$25,000 – $34,999	18 (9.00%)
$35,000 – $49,999	36 (18.10%)
$50,000 – $64,999	42 (21.10%)
$65,000 – $79,999	21 (10.60%)
≥ $80,000	38 (19.10%)
Current Relationship Status
Single	62 (31.20%)
Living with significant other	32 (16.10%)
Married	87 (43.70%)
Divorced	17 (8.50%)
Separated	1 (.50%)
Widowed	0 (0%)
Ethnicity
Hispanic or Latino	17 (8.50%)
Not Hispanic or Latino	179 (89.90%)
Unknown	3 (1.50%)
Racial Identity (could select more than one response option)
Caucasian or White	165 (82.90%)
African American or Black	26 (13.10%)
Asian	9 (4.50%)
American Indian or Alaskan Native	5 (2.50%)
Native Hawaiian/other Pacific Islander	1 (.50%)
Unknown	0 (0%)
Mental Health Treatment (could select more than one response option)
Currently in therapy	29 (14.60%)
Been in therapy in the past	83 (41.70%)
Currently taking medications for mental health/emotional problem	40 (20.10%)
Have taken medications in the past for mental health/emotional problem	57 (28.60%)
Never received treatment (therapy and/or medications) for a mental health/emotional problem	76 (38.20%)
Worst/Most Distressing Traumatic Event Endorsed on the LEC-5
Natural disaster	30 (15.10%)
Fire or explosion	8 (4.00%)
Transportation accident	34 (17.10%)
Serious accident at work/home/during recreational activity	9 (4.50%)
Exposure to a toxic substance	0 (0%)
Physical assault	20 (10.10%)
Assault with a weapon	7 (3.50%)
Sexual assault	36 (18.10%)
Other unwanted/uncomfortable sexual experience	8 (4.00%)
Combat or exposure to war	1 (.50%)
Forced captivity	0 (0%)
Life-threatening illness	12 (6.00%)
Severe human suffering	3 (1.50%)
Sudden, violent death	14 (7.00%)
Sudden, accidental death	13 (6.50%)
Serious injury/harm/death you caused to someone else	4 (2.00%)

Note. n = 199; SD = standard deviation; LEC-5 = Life Events Checklist for DSM-5.

Statistical Analyses

An a prior power analysis was conducted with G*Power 3.1.9.6 (Faul et al., 2009). This study used a medium effect size convention (f² = .15) and standard parameters for alpha (α = .05) and power (.80; Cohen, 1992). It was determined that a sample size of 150 participants was optimal to ensure this study sustained sufficient power for the analyses. Data were analyzed using SPSS Version 25. Assumptions of normality (i.e., skewness values < 2 and kurtosis values < 7; Curran et al., 1996), linearity, homoscedasticity, and multicollinearity (variance inflation factor [VIF] values < 10 and tolerance values > 0.1; Hair et al., 2009) were examined. Moderation analyses conducted using PROCESS Macro Version 3.5 (Hayes, 2020) were used to examine all research questions. Ninety-five percent bootstrap confidence intervals (CIs) were estimated based on 5,000 random bootstrapping samples, which is the default in PROCESS Macro and sufficient to detect significant effects (Hayes, 2017). This study also used PROCESS Macro to mean-center variables. To examine strength and significance of main and interaction effects, this study examined the unstandardized regression coefficients (b) and whether associated p values were < .05 (Hayes, 2017).

Empirical evidence shows older age and female gender are both significantly related to sleep disturbances (Crowley, 2011; Madrid-Valero et al., 2017) and PTSD symptom severity (Olff, 2017; Parker et al., 2016). Thus, age and gender were added as covariates to all models. To examine Hypotheses 1–4, this study used PROCESS Model 1 (simple moderation model; Hayes, 2020). In each model, sleep disturbance (sleep quality or quantity) was the independent variable (IV), age and gender (female vs. not female) were the covariates, emotion dysregulation (negative or positive) was the moderator, and PTSD symptom severity was the dependent variable (DV; see Figure 1). To examine Exploratory Hypotheses 1 and 2, this study used PROCESS Model 2 (i.e., additive multiple moderation model; Hayes, 2020). In each model, sleep disturbance (sleep quality or quantity) was the IV, age and gender (female vs. not female) were the covariates, negative and positive emotion dysregulation were the moderators, and PTSD symptom severity was the DV (see Figure 2).

Figure 1.

Simple Moderation Models for Hypotheses 1–4

Note. Model depicts negative or positive emotion dysregulation moderating the association between sleep disturbances (i.e., sleep quality or sleep quantity) and posttraumatic stress disorder (PTSD) symptom severity, with age and gender as covariates. Effects were examined in four separate models, where either sleep quality or sleep quantity was the independent variable (IV) and either negative or positive emotion dysregulation was the moderator. DV = dependent variable.

Figure 2.

Additive Multiple Moderation Models for Exploratory Hypotheses 1 and 2

Note. Model depicts negative emotion dysregulation and positive emotion dysregulation moderating the association between sleep disturbances (i.e., sleep quality or sleep quantity) and posttraumatic stress disorder (PTSD) symptom severity, with age and gender as covariates. Effects were examined in two separate models, where sleep quality was the independent variable (IV) in one model and sleep quantity was the IV in the other model. DV = dependent variable.

Simple slopes analysis was utilized to probe any significant interaction effects (i.e., signifigant moderation; Hayes, 2020). This involves estimating the conditional effect of the IV on the DV at one standard deviation (SD) below, at, or above the mean value (Hayes, 2017). In addition, the Johnson-Neyman (J-N) technique (Johnson & Neyman, 1936) was utilized to probe any significant interaction effects in the simple moderation models; this technique cannot be applied to additive multiple moderation models in PROCESS Macro (Hayes, 2017). The J-N technique avoids the potential arbitrariness of simple slopes analysis by mathematically identifying the point(s) in the range of the specified moderator where the effect of the IV on the DV changes from statistically significant to not statistically significant (Hayes & Matthes, 2009).

Results

The average total score on the PCL-5 was 25.23 (SD = 16.90). The average reported sleep quantity was 6.63 hours (SD = 1.24), and the average reported sleep quality component score was 1.85 (SD = .82). See Table 2 for additional descriptive information on variables of interest. See Table 3 for correlations between study variables. Assumptions of normality, linearity, homoscedasticity, and multicollinearity were not violated for any analyses (Hair et al., 2009).

Table 2.

Information on Study Variables of Interest

Variable	Mean (SD)	Cronbach’s Alpha	Skewness (SE)	Kurtosis (SE)
PCL-5 total score	25.23 (16.90)	.95	.19 (.17)	−.97 (.34)
DERS-16 total score	36.90 (16.50)	.97	.53 (.17)	−.49 (.34)
DERS-P total score	21.74 (11.83)	.97	1.32 (.17)	.41 (.34)
PSQI sleep quantity	6.63 (1.24)	-	−.43 (.17)	−.28 (.34)
PSQI sleep quality	1.85 (.82)	-	−.28 (.17)	−.50 (.34)

Note. n = 199; PCL-5 = PTSD Checklist for DSM-5; DERS-16 = The Difficulties in Emotional Regulation Scale – Brief 16-Item Version; DERS-P = Difficulties in Emotional Regulation Scale – Positive Emotions; PSQI = Pittsburgh Sleep Quality Index; SD = standard deviation; SE = standard error.

Table 3.

Pearson Correlations between Study Variables of Interest

Variable	1	2	3	4	5
1. PTSD Severity	-
2. Sleep Quality	−.18 p=.010	-
3. Sleep Quantity	−.26 p<.001	.65 p<.001	-
4. Negative ED	.72 p<.001	−.24 p=.001	−.18 p=.010	-
5. Positive ED	.52 p<.001	.07 p=.365	.01 p=.860	.57 p<.001	-

Note. PTSD = posttraumatic stress disorder; ED = emotion dysregulation; bold values = significant results.

See Table 4 for detailed results of simple moderation analyses for Hypotheses 1–4. Regarding significant main effects, greater negative emotion dysregulation, greater positive emotion dysregulation, poorer sleep quality, and lower sleep quantity were associated with greater PTSD symptom severity. There was a significant moderation effect of negative emotion dysregulation on the relationship between sleep quality and PTSD symptom severity (b = 0.17, p = .002). Simple slopes analysis (see Figure 3) revealed that poorer sleep quality was associated with greater PTSD symptom severity when levels of negative emotion dysregulation were low (−1 SD; b = −3.33, SE = 1.49, p = .027), but not when levels were average (b = −0.47, SE = 1.05, p = .655) or high (+1 SD; b = 2.39, SE = 1.30, p = .068). J-N analysis showed that when negative emotion dysregulation (centered) values were less than −12.55, the slope between poorer sleep quality and greater PTSD symptom severity was statistically significant at p < .05. When negative emotion dysregulation (centered) values were greater than 18.21, the slope between better sleep quality and greater PTSD symptom severity was statistically significant at p < .05 (range of observed [centered] values of negative emotion dysregulation was [−20.90, 43.10]). There was also a significant moderation effect of negative emotion dysregulation on the relationship between sleep quantity and PTSD symptom severity (b = 0.09, p = .043). Simple slopes analysis (see Figure 4) revealed that lower sleep quantity was associated with greater PTSD symptom severity when levels of negative emotion dysregulation were low (−1 SD; b = −3.21, SE = 1.01, p = .002) or average (b = −1.75, SE = .68, p = .011), but not high (+1 SD; b = −0.29, SE = .97, p = .761). J-N analysis showed that when negative emotion dysregulation (centered) values were less than 4.27, the slope between poorer sleep quantity and greater PTSD symptom severity was statistically significant at p < .05. Our results also showed that positive emotion dysregulation did not moderate the relationships between sleep disturbances (i.e., sleep quality and quantity) and PTSD symptom severity.

Table 4.

Hypotheses 1–4: Moderated Regression Results of Sleep Quality or Quantity and Negative or Positive Emotion Dysregulation on PTSD Symptom Severity

Predictors	b	95% CI [LL, UL]	SE	t	R 2	F(df)
Dependent Variable = PTSD Severity
					.55	47.93 (5, 193) p<.001
Age	−0.02 p=.820	[−0.17, 0.13]	.08	−.23
Gender	−4.00 p=.022	[−7.42, −0.59]	1.73	−2.31
Sleep Quality	−0.47 p=.655	[−2.55, 1.60]	1.05	−.45
Negative ED	0.75 p<.001	[0.64, 0.85]	.05	14.05
Sleep Quality X Negative ED	0.17 p=.002	[0.06, 0.28]	.06	3.09
Dependent Variable = PTSD Severity
					.36	21.54 (5, 193) p<.001
Age	−0.18 p=.042	[−0.36, −0.01]	.09	−2.05
Gender	−5.72 p=.007	[−9.84, −1.59]	2.09	−2.73
Sleep Quality	−4.20 p=.001	[−6.59, −1.82]	1.21	−3.48
Positive ED	0.75 p<.001	[0.58, 0.92]	.09	8.63
Sleep Quality X Positive ED	0.11 p=.282	[−0.09, 0.32]	.10	1.08
Dependent Variable = PTSD Severity
					.56	48.28 (5, 193) p<.001
Age	−0.03 p=.719	[−0.18, 0.12]	.08	−0.36
Gender	−3.06 p=.080	[−6.49, 0.37]	1.74	−1.76
Sleep Quantity	−1.75 p=.011	[−3.10, −0.41]	.68	−2.58
Negative ED	0.72 p<.001	[0.62, 0.82]	.05	13.72
Sleep Quantity X Negative ED	0.09 p=.043	[0.003, 0.17]	.04	2.04
Dependent Variable = PTSD Severity
					.38	23.53 (5, 193) p<.001
Age	−0.19 p=.032	[−0.36, −0.02]	.09	−2.16
Gender	−5.09 p=.014	[−9.14, −1.04]	2.06	−2.48
Sleep Quantity	−3.52 p<.001	[−5.09, −1.96]	.79	−4.44
Positive ED	0.75 p<.001	[0.58, 0.91]	.08	8.89
Sleep Quantity X Positive ED	0.07 p=.311	[−0.06, 0.20]	.07	1.02

Note. n = 199; ED = emotion dysregulation; PTSD = posttraumatic stress disorder; b = unstandardized regression weights; bold values = significant results; CI = confidence interval; LL and UL indicate the lower and upper limits of a confidence interval, respectively; SE = standard error; For gender: 0 = female [reference group]; 1 = not female.

Figure 3.

Moderating Impact of Negative Emotion Dysregulation between Sleep Quality and PTSD Symptom Severity

Note. PSQI = Pittsburgh Sleep Quality Index; PCL-5 = Posttraumatic Stress Disorder Checklist for DSM-5; DERS-16 = Brief Version of the Difficulties in Emotion Regulation Scale; Simple slopes are shown at the mean and at +/− 1 standard deviation (SD) above and below the mean of negative emotion dysregulation; For sleep quality, higher values indicate better sleep; Both PSQI sleep quality and DERS-16 scores were centered at the mean.

Figure 4.

Moderating Impact of Negative Emotion Dysregulation between Sleep Quantity and PTSD Symptom Severity

Note. PSQI = Pittsburgh Sleep Quality Index; PCL-5 = Posttraumatic Stress Disorder Checklist for DSM-5; DERS-16 = Brief Version of the Difficulties in Emotion Regulation Scale; Simple slopes are shown at the mean and at +/− 1 standard deviation (SD) above and below the mean of negative emotion dysregulation; Both PSQI sleep quantity and DERS-16 scores were centered at the mean.

See Tables 5, 6, and 7 for detailed results of additive multiple moderation analyses for Exploratory Hypotheses 1 and 2. There was a significant moderation effect of negative emotion dysregulation on the relationship between sleep quality and PTSD symptom severity (b = 0.18, p = .008). Simple slopes analysis revealed that at low levels of negative emotion dysregulation (−1 SD) coupled with low (−1 SD; b = −2.96, SE = 1.50, p = .050), average (b = −3.96, SE = 1.57, p = .012), or high (+1 SD; b = −5.32, SE = 2.27, p = .020) levels of positive emotion dysregulation, there was a significant association between poorer sleep quality and greater PTSD symptom severity. The effect became stronger as levels of positive emotion dysregulation increased. However, at average or high (+1 SD) levels of negative emotion dysregulation coupled with any level of positive emotion dysregulation, there was no significant association between sleep quality and PTSD symptom severity. In addition, there was a significant moderation effect of negative emotion dysregulation on the relationship between sleep quantity and PTSD symptom severity (b = 0.11, p = .031). Simple slopes analysis revealed that at low levels of negative emotion dysregulation (−1 SD) coupled with low (−1 SD; b = −2.62, SE = 1.00, p = .010), average (b = −3.66, SE = 1.05, p = .001), or high (+1 SD; b = −5.06, SE = 1.53, p = .001) levels of positive emotion dysregulation, there was a significant association between lower sleep quantity and greater PTSD symptom severity. The effect became stronger as levels of positive emotion dysregulation increased. At average levels of negative emotion dysregulation coupled with average (b = −1.77, SE = .67, p = .009) or high (+1 SD; b = −3.17, SE = .95, p = .001) levels of positive emotion dysregulation, there was a significant association between lower sleep quantity and greater PTSD symptom severity. However, at high (+1 SD) levels of negative emotion dysregulation coupled with any level of positive emotion dysregulation, there was no significant association between sleep quantity and PTSD symptom severity.

Table 5.

Exploratory Hypotheses 1 and 2: Moderated Regression Results of Sleep Quality or Quantity and Both Negative and Positive Emotion Dysregulation on PTSD Symptom Severity

Predictors	b	95% CI [LL, UL]	SE	t	R 2	F(df)
Dependent Variable = PTSD Severity
					.57	36.72 (7, 191) p<.001
Age	−0.02 p=.826	[−0.16, 0.13]	.07	−.22
Gender	−4.33 p=.013	[−7.73, −0.92]	1.72	−2.51
Sleep Quality	−1.04 p=.326	[−3.12, 1.04]	1.05	−.98
Negative ED	0.66 p<.001	[0.53, 0.79]	.07	9.83
Positive ED	0.24 p=.007	[0.06, 0.41]	.09	2.71
Sleep Quality X Negative ED	0.18 p=.008	[0.05, 0.31]	.07	2.69
Sleep Quality X Positive ED	−0.12 p=.249	[−0.31, 0.08]	.10	−1.16
Dependent Variable = PTSD Severity
					.59	38.64 (7, 191) p<.001
Age	−0.02 p=.817	[−0.16, 0.13]	.07	−.23
Gender	−3.61 p=.036	[−6.98, −0.25]	1.71	−2.12
Sleep Quantity	−1.77 p=.009	[−3.10, −0.44]	.67	−2.63
Negative ED	0.64 p<.001	[0.51, 0.77]	.07	9.74
Positive ED	0.24 p=.006	[0.07, 0.42]	.09	2.81
Sleep Quantity X Negative ED	0.11 p=.031	[0.01, 0.22]	.05	2.17
Sleep Quantity X Positive ED	−0.12 p=.083	[−0.25, 0.02]	.07	−1.74

Note. ED = emotion dysregulation; PTSD = posttraumatic stress disorder; b = unstandardized regression weights; bold values = significant results; CI = confidence interval; LL and UL indicate the lower and upper limits of a confidence interval, respectively; SE = standard error; For gender: 0 = female [reference group]; 1 = not female.

Table 6.

Exploratory Hypothesis 1: Simple Slopes Analysis of Moderated Regression Results of Sleep Quality and Negative Emotion Dysregulation on PTSD Symptom Severity

DERS-16 Total Score	DERS-P Total Score	b	95% CI [LL, UL]	SE	t
−16.50	−8.74	−2.96 p=.050	[−5.92, 0.004]	1.50	−1.97
−16.50	0.00	−3.96 p=.012	[−7.05, −0.87]	1.57	−2.53
−16.50	11.83	−5.32 p=.020	[−9.80, −0.85]	2.27	−2.35
0.00	−8.74	−0.03 p=.982	[−2.73, 2.67]	1.37	−.02
0.00	0.00	−1.04 p=.326	[−3.12, 1.04]	1.05	−.98
0.00	11.83	−2.40 p=.130	[−5.51, 0.71]	1.58	−1.52
16.50	−8.74	2.89 p=.144	[−0.99, 6.78]	1.97	1.47
16.50	0.00	1.89 p=.199	[−1.00, 4.77]	1.46	1.29
16.50	11.83	0.52 p=.724	[−2.40, 3.44]	1.48	.35

Note. DERS-16 = Brief Version of the Difficulties in Emotion Regulation Scale; DERS-P = Difficulties in Emotion Regulation Scale – Positive; PTSD = posttraumatic stress disorder; b = unstandardized regression weights; bold values = significant results; CI = confidence interval; LL and UL indicate the lower and upper limits of a confidence interval, respectively; SE = standard error; Both the DERS-16 and DERS-P scores were centered at the mean. Both the DERS-16 and DERS-P scores are shown at the mean and at +/− 1 standard deviation (SD) above and below the mean.

Table 7.

Exploratory Hypothesis 2: Simple Slopes Analysis of Moderated Regression Results of Sleep Quantity and Negative Emotion Dysregulation on PTSD Symptom Severity

DERS-16 Total Score	DERS-P Total Score	b	95% CI [LL, UL]	SE	t
−16.50	−8.74	−2.62 p=.010	[−4.60, −0.65]	1.00	−2.62
−16.50	0.00	−3.66 p=.001	[−5.73, −1.59]	1.05	−3.48
−16.50	11.83	−5.06 p=.001	[−8.08, −2.03]	1.53	−3.30
0.00	−8.74	−0.74 p=.450	[−2.66, 1.19]	.97	−.76
0.00	0.00	−1.77 p=.009	[−3.10, −0.44]	.67	−2.63
0.00	11.83	−3.17 p=.001	[−5.05, −1.30]	.95	−3.34
16.50	−8.74	1.14 p=.460	[−1.91, 4.20]	1.55	.74
16.50	0.00	0.11 p=.921	[−2.15, 2.37]	1.15	.10
16.50	11.83	−1.28 p=.190	[−3.21, 0.64]	.98	−1.31

Note. DERS-16 = Brief Version of the Difficulties in Emotion Regulation Scale; DERS-P = Difficulties in Emotion Regulation Scale – Positive; PTSD = posttraumatic stress disorder; b = unstandardized regression weights; bold values = significant results; CI = confidence interval; LL and UL indicate the lower and upper limits of a confidence interval, respectively; SE = standard error; Both the DERS-16 and DERS-P scores were centered at the mean. Both the DERS-16 and DERS-P scores are shown at the mean and at +/− 1 standard deviation (SD) above and below the mean.

Discussion

This study examined the potential moderating roles of negative and positive emotion dysregulation in the relationships between subjective sleep disturbances and PTSD symptom severity in a community sample of 199 trauma-exposed individuals. First, significant main effects provided empirical support for the conclusion that greater (negative and positive) emotion dysregulation and sleep disturbances both relate to greater PTSD symptom severity among trauma-exposed individuals. Specifically, our results support theoretical perspectives (Tull et al., 2020) and empirical evidence (e.g., Weiss et al., 2020b) to suggest that fear or avoidance of intense negative and positive emotions could lead to the development and/or maintenance of PTSD symptoms. Further, our findings are in line with theory (Spoormaker & Montgomery, 2008) and empirical evidence (e.g., DeViva et al., 2021) that indicate poorer sleep quality and lower sleep quantity could relate to deficits in emotional coping, which, in turn, may contribute to the onset and perpetuation of PTSD symptoms.

Primary results revealed that poorer sleep quality was associated with greater PTSD symptom severity when levels of negative emotion dysregulation were low but not when levels were average or high. In addition, lower sleep quantity was associated with greater PTSD symptom severity when levels of negative emotion dysregulation were low or average but not when levels were high. These findings were not as expected, and a few explanations of these results are outlined.

First, trauma-exposed individuals may experience less emotional awareness (i.e., difficulty identifying and describing emotions; Tull et al., 2020), making it difficult to recognize different types and intensity of emotions. Such processes may influence and lessen the utilization of emotion regulation strategies broadly (Feldman-Barrett et al., 2001). Second, a strong inability to regulate negative emotions may lead to emotional flooding (Rizzo et al., 2009); such an overwhelming emotional experience may adversely impact an individual’s ability to accurately self-report levels of emotion dysregulation (Tull et al., 2020). These processes may explain why high levels of negative emotion dysregulation did not influence the sleep-PTSD relationship in the expected direction in this study and require empirical investigation. Third, research has shown that only certain facets of negative emotion dysregulation (i.e., difficulty accessing effective emotion regulation skills and lack of emotional clarity) significantly mediate associations between sleep disturbances and distress (e.g., depression; Hom et al., 2016). Thus, examining specific facets of emotion dysregulation may be an important consideration for future research. Lastly, perhaps concurrent levels of positive emotion dysregulation may influence the moderating role of negative emotion dysregulation in the sleep-PTSD relationship. In this regard, the current study’s exploratory findings suggested that at low levels of negative emotion dysregulation coupled with low, average, or high levels of positive emotion dysregulation, there was a significant association between poorer sleep quality and greater PTSD symptom severity. This study’s exploratory findings also showed that at low levels of negative emotion dysregulation coupled with low, average, or high levels of positive emotion dysregulation, there was a significant association between lower sleep quantity and greater PTSD symptom severity. Relatedly, at average levels of negative emotion dysregulation coupled with average or high levels of positive emotion dysregulation, there was a significant association between lower sleep quantity and greater PTSD symptom severity. Perhaps the inability to regulate positive emotions may have a stronger detrimental impact on sleep and PTSD symptoms even when individuals experience minimal dysregulation of negative emotions. Indeed, positive emotions are characterized by moderate to high levels of physiological arousal (e.g., greater cardiac vagal tone, heightened startle response; Carl et al., 2013), and individuals with PTSD symptoms may experience this heightened physiological arousal as distressing due to recurrent hyperarousal symptoms (Tull et al., 2020). Relatedly, greater PTSD symptom severity is associated with fear of positive emotions (Contractor et al., 2021) and more difficulty accepting positive emotions (Weiss et al. 2019a). Moreover, research suggests positive emotions may contribute to a lack of behavioral control and greater distractibility (Weiss et al., 2019b). Thus, heightened physiological arousal, greater fear or lack of acceptance, and more difficulties with behavioral control associated with positive emotions, coupled with a subsequent inability to regulate these positive emotions, may contribute to poorer sleep and make it increasingly difficult to manage post-trauma distress, such as PTSD symptoms (Spoormaker & Montgomery, 2008).

Of note, results indicated that when negative emotion dysregulation was high, there was a significant relationship between better sleep quality and greater PTSD symptom severity. Relatedly, our exploratory findings indicated that at high levels of negative emotion dysregulation, the additive experience of positive emotion dysregulation (at any level) did not influence the relationships between sleep disturbances and PTSD symptoms. Possibly, individuals managing a greater cognitive load of PTSD symptoms and negative emotion dysregulation have a unique subjective experience of sleep. To elaborate, cognitive models of PTSD suggest that individuals experiencing PTSD symptoms utilize many resources in a way that is cognitively taxing (Hayes et al., 2012), such as increased information processing to detect potential threats and a narrowed focus of attention on neutral stimuli perceived as frightening. Moreover, research suggests that regulating emotions is also a cognitively taxing experience as it requires substantial cognitive effort and control over emotional experiences (e.g., Gross & John, 2003; Werner-Seidler et al., 2013). Perhaps, for those experiencing greater PTSD symptom severity and greater negative emotion dysregulation, sleep is used as a coping strategy to manage this cognitive load. Indeed, improved sleep may enhance sleep-dependent neuroplasticity, thereby promoting cognitive recovery (Gorgoni et al., 2013; van Liempt, 2012). Ultimately, further research is needed to understand if and how the cognitive load of emotion dysregulation and PTSD symptoms may influence current study questions.

Contrary to study hypotheses, primary results also showed that positive emotion dysregulation did not moderate the relationships between sleep disturbances (i.e., sleep quality and quantity) and PTSD symptom severity. Research has shown that the experience of positive affect is related to numerous health benefits such as improved physical health and healthy behaviors (Lyubomirsky et al., 2005; Steptoe et al., 2009), less emotion dysregulation (McLaughlin et al., 2019), less PTSD symptom severity (Etter et al., 2013), and improved sleep (Ong et al., 2017). Thus, perhaps the beneficial health effects of any experienced positive emotions among trauma-exposed individuals (regardless of their ability to appropriately regulate such emotions) may confound the sleep-PTSD relationship. Moreover, positive affect encompasses short-term emotional experiences and enduring mood states (i.e., affective trait); individuals with “high trait” positive affect may have an improved ability to cope with stressors and associated emotional distress because the experience of positive affect may reduce negative appraisals (Ong et al., 2017). Thus, these individuals may not experience detrimental health effects, like sleep disturbances or PTSD symptoms. Future research should examine state vs. trait positive affect to determine if and how this may influence the relationship between sleep disturbances, positive emotion dysregulation, and PTSD symptom severity.

Limitations and Future Directions

The findings of the current study should be interpreted with certain limitations in mind. First, the self-report instruments utilized in this study may have facilitated social desirability effects and biased responding (Van de Mortel, 2008). This limitation is partly mitigated by the current study’s utilization of self-assessment measures with good psychometric properties. However, future studies should utilize multi-method assessment, such as clinician-administered measures of emotion dysregulation (e.g., Semi-Structured Emotion Regulation Interview; Lee et al., 2017), of PTSD symptoms (e.g., Clinician-Administered PTSD Scale for DSM-5; Weathers et al., 2018), and of sleep disturbances (e.g., Structured Clinical Interview for Sleep Disorders-Revised; Taylor et al., 2019). Second, participants subjectively reported their sleep quality and sleep quantity over the past month. Retrospective reporting of sleep over such a time frame could introduce error as individuals may have difficulty grasping the concept of calculating an average rating (needed for making retrospective reports; Harvey et al., 2008), and retrospective ratings are susceptible to recency and/or saliency effects (Harvey et al., 2008). Moreover, this study only used one-item measures of sleep quality and sleep quantity from the PSQI referencing the past month. Though this is not uncommon (Biggs et al., 2020; Short et al., 2017), future research should examine subjective reports of sleep quality and quantity using a multi-item measure of sleep disturbances and should additionally use behavioral or objective measures of sleep disturbances (e.g., actigraphy, polysomnography) to obtain a better understanding of the complex relationships between these constructs. Third, the majority of this study’s sample consisted of individuals who identified as white, non-Hispanic/Latino, and female. Relatedly, only a minority of the sample reported clinically significant levels of symptom severity; 39.7% of the sample reported clinically significant levels of PTSD symptom severity (i.e., PCL-5 total scores ≥ 31), and ≤ 5% of participants reported significant sleep disturbances (i.e., “very bad” sleep quality and/or total sleep quantity scores of < 5 hours). These characteristics of our sample limit the generalizability of our results to more demographically diverse samples of trauma-exposed individuals with varying levels of clinical symptom severity. Fourth, this study utilized a cross-sectional design and therefore could not make inferences about causality between the variables of interest. Relatedly, the current study utilized moderation analyses, and underlying theories of moderation models indicate directional implications that are fundamentally causal (Wu & Zumbo, 2008). Thus, future research examining these constructs of interest using a moderation analytic approach should implement a prospective, longitudinal study design to clarify directionality of these relationships.

Finally, as evident from all study findings, the interactions between negative and positive emotion dysregulation (Carl et al., 2013) and their effects on the relationships between sleep disturbances and PTSD symptom severity are quite complex. In this regard, this study outlines a few additional considerations specific to the construct of emotion dysregulation. First, emotion dysregulation could involve temporary, dynamic, and context-dependent responses or stable and trait-like factors (Carl et al., 2013); participants in the current study reported levels of emotion dysregulation referencing the past month, which precluded our ability to examine the state- vs. trait-like elements involved in emotion dysregulation. Further, emotion dysregulation was measured via self-report rather than with a clinician-administered measure, and our results could have been influenced by participants’ ability to accurately report this type of information. Specifically, individuals with PTSD symptoms experience deficits in emotional clarity (e.g., difficulty understanding emotions; Tull et al., 2007), and may find it difficult to understand certain components involved in emotion dysregulation (e.g., accepting emotions; Lee et al., 2017). Consequently, this population may have difficulty understanding and reporting the complex constructs involved in emotion dysregulation. Perhaps, individuals may even need at least an average ability to identify and regulate emotions to respond to measures of emotion dysregulation accurately (Tull et al., 2020). Therefore, our results may not effectively capture the role of severe levels of negative and positive emotion dysregulation as related to PTSD symptom severity and sleep disturbances. Lastly, as previously mentioned, emotion dysregulation is a multi-faceted construct encompassing multiple maladaptive ways of responding to emotions (e.g., lack of awareness of emotions, inability to manage behaviors when emotionally distressed; Weiss et al., 2018). Thus, perhaps examining the moderating role of facets of negative and positive emotion dysregulation would have revealed more nuanced findings; this is an important future research endeavor.

Implications and Conclusions

Our results have important theoretical and clinical implications. Study findings advance the theoretical literature connecting emotion dysregulation to sleep disturbances (e.g., Baglioni et al., 2010) and PTSD symptoms (e.g., Tull et al., 2020). Specifically, results support that positive and negative emotion dysregulation are conceptually distinct constructs (Carl et al., 2013) as well as show that the valence of emotion dysregulation differentially affects the sleep-PTSD relationship. In this regard, while lower levels of negative emotion dysregulation influenced relationships between sleep disturbances and greater PTSD severity, positive emotion dysregulation was not a significant moderating variable. Further, exploratory findings indicate sleep disturbances may coalesce with negative and positive emotion dysregulation (Gruber & Cassoff, 2014), compared to either negative or positive emotion dysregulation alone, to maintain or exacerbate PTSD symptoms (Spoormaker & Montgomery, 2008). In this regard, the presence of positive emotion dysregulation—irrespective of its level—may interact with low levels of negative emotion dysregulation to influence the relationships between sleep disturbances and greater PTSD symptom severity. Our results suggest that greater difficulty regulating positive (vs. negative) emotions may be more psychologically distressing for trauma-exposed individuals (Tull et al., 2020). As indicated earlier, prospective, longitudinal investigations are needed to ascertain how the interplay of negative and positive emotion dysregulation influences current study findings.

Study results advance clinical knowledge of addressing emotion dysregulation in the treatment of sleep disturbances and PTSD. In particular, our findings suggest negative emotion dysregulation may be an important clinical focus in sleep disorder treatments. Study results indicate that clinicians may benefit from assessing and targeting levels of negative and positive emotion dysregulation (Weiss et al., 2020b) among patients reporting sleep disturbances, as the interaction of these factors with sleep problems may exacerbate PTSD symptoms. Researchers have attempted to modify existing PTSD treatments by incorporating specific adaptive emotion regulation strategies to improve PTSD symptoms (Cloitre et al., 2002); however, this type of modification to sleep disorder treatments has not been conducted (Cerolini et al., 2015; Pickett et al., 2016) and may be warranted pending additional research. It is possible that only targeting sleep disturbances (e.g., poor sleep quality, lower sleep quantity) without targeting emotion regulation difficulties more directly may not effectively address the mechanisms or sequelae involved in the maintenance of sleep disturbances in those with co-occurring PTSD symptoms (Pickett et al., 2016). Relatedly, some research suggests mixed results regarding the effects of sleep disorder treatments (e.g., CBT-I) on PTSD symptom severity (Talbot et al., 2014; Taylor & Pruiksma, 2014). Indeed, Cerolini et al. (2015) argue that the addition of specific emotion regulation techniques could enhance the effectiveness of sleep therapies, such as CBT-I. On the other hand, meta-analytic research has shown CBT for sleep disturbances is efficacious in the reduction of co-occurring PTSD symptom severity (Ho et al., 2016), which suggests that some aspects of emotion dysregulation that occur in the context of PTSD may be adequately addressed without any sleep treatment modifications that target emotion dysregulation more directly.

In summary, the current study addresses many limitations in the existing literature by examining how negative and positive emotion dysregulation influence the associations between sleep disturbances and PTSD symptom severity among those who have experienced trauma. Importantly and relatedly, this study addressed a considerable gap in the literature by examining insufficiently studied relationships between subjective sleep disturbances and positive emotion dysregulation among trauma-exposed individuals. Further strengths of the current study include the utilization of validated measures with good psychometric properties to examine variables of interest and the use of data from a large community sample to conduct well-powered statistical analyses. Our findings add to the existing theoretical and empirical literature and inform future research on how emotion dysregulation influences the sleep-PTSD relationship.