Introduction

Recent attention has been given to the role of sleep disturbance in psychopathology, including as a transdiagnostic factor that may contribute to various disorders [1, 2]. Likewise, recent findings implicate sleep disturbance in suicidal ideation and suicide attempts [3]. Among the disorders, considerable research has examined the relationship between sleep disturbance and posttraumatic stress disorder (PTSD). PTSD is a maladaptive response to a traumatic event and is characterized by intrusive thoughts related to the event, avoidance of reminders of the event, negative mood and cognitions, and heightened arousal and reactivity [4]. Although the relationship between sleep and PTSD is complex, results generally indicate a link between these two variables [5]. The links between sleep disturbance and suicidal thoughts and behaviors may be particularly important in the context of PTSD, given high rates of suicide among those with PTSD, particularly veterans [6, 7].

The purpose of the present review is to integrate findings from research examining sleep disturbance in PTSD over the past 5 years. As the role of sleep disturbance in PTSD is a rapidly growing field, there may be value in integrating the most contemporary research on this topic. First, subjective and objective sleep disturbance in PTSD will be examined, followed by discrepancies between these measurement modalities among those with PTSD. Results of treatment outcome research will then be reviewed. Finally, conclusions and future directions for the study of sleep disturbance in PTSD will be discussed, including the potential role of sleep disturbance as a causal factor in the development of PTSD and possible sleep-specific PTSD prevention and treatment targets. Given evidence for the downstream effects of sleep disturbance across physiological, cognitive, and affective function, as well as the proposed role of sleep disturbance as a transdiagnostic factor, it is important to move beyond the typical view of sleep disturbance as a mere symptom of PTSD to examine how the effects of sleep disturbance may culminate to confer risk for the development of PTSD following trauma exposure [2].

Subjective Sleep in PTSD

Considerable research has examined the role of subjective sleep disturbance in relation to PTSD, with the vast majority finding evidence for a link between these two variables. Compared to healthy controls, those with PTSD consistently report increased sleep disturbance, and this finding has been shown across diverse trauma samples, including veterans, sexual assault survivors, female samples, childhood abuse, and mixed trauma samples [8•, 9,10,11,12,13,14,15,16,17,18,19,20]. Similarly, on daily sleep diaries, individuals with PTSD report decreased total sleep time (TST), decreased sleep efficiency, increased wake after sleep onset (WASO), and increased sleep onset latency (SOL) compared to healthy controls [8•, 910, 12, 16, 18].

Those with PTSD also consistently report more sleep disturbance than trauma-exposed controls, and this difference has been seen in veterans, sexual assault survivors, natural disaster survivors, paramedics, and mixed trauma samples [10, 20,21,22,23,24,25,26,27,28,29,30,31,32]. Additionally, trauma-exposed controls report increased sleep disturbance compared to healthy controls [33]. In contrast to the research comparing those with PTSD to healthy controls, few studies have utilized sleep diaries to compare sleep between those with PTSD and trauma-exposed controls. Extant findings are mixed, with one study indicating increased WASO and SOL in those with PTSD and another study finding no differences between groups [10, 27]. Additional research utilizing sleep diaries is necessary to clarify these discrepant findings.

A much larger body of work has utilized correlational designs to assess the relationship between sleep disturbance and symptoms of PTSD in both clinical and nonclinical samples. The majority of this work indicates small to moderate associations between sleep disturbance and increased PTSD symptom severity among veterans, civilians, and assault survivors with PTSD, combat-exposed veterans, emergency responders, medical patients, postpartum women, African-American adults, alcohol-dependent inpatients, refugees, natural disaster survivors, older adults, and young adults [34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65]. In contrast, a small number of studies have found no association between sleep disturbance and symptoms of PTSD [66, 67].

Studies of sleep-disordered samples also indicate a role of sleep disturbance in PTSD. For example, among patients seeking treatment for sleep disturbance, those with elevated PTSD symptoms report increased sleep disturbance [68]. Similarly, patients with comorbid obstructive sleep apnea (OSA) and PTSD report more insomnia than OSA patients without PTSD [69]. Likewise, veterans with insomnia are more likely to have PTSD than veterans without insomnia [70, 71]. Finally, one study examining sleep over the course of a week found similar levels of sleep disturbance between those with PTSD and those with insomnia [8•]. These findings highlight the severity of subjective sleep disturbance in PTSD and suggest a compounding role of clinical sleep disturbance.

Other studies utilizing medical patient samples likewise indicate a relationship between sleep disturbance and PTSD symptoms. Among intensive care unit (ICU) patients, those with insomnia following ICU discharge are more likely to endorse PTSD symptoms compared to those without insomnia [72]. Similarly, chronic pain patients with PTSD report increased sleep disturbance compared to pain patients without PTSD [73]. Interestingly, among injury patients, SOL predicts increased PTSD symptoms among men, but decreased symptoms among women [74]. Additional research is necessary to understand the mechanism behind this gender difference. Taken together, these results suggest the importance of considering the link between sleep disturbance and PTSD in medical samples.

While few studies have examined the effect of specific trauma types, some studies indicate that sexual trauma may be particularly linked to sleep disturbance. For example, sleep disturbance is associated with sexual trauma among African-American young adults, and among veterans, those with sexual trauma are more likely to experience sleep disturbance and report increased sleep disturbance compared to those without sexual trauma [75,76,77]. Further, extant research suggests a relationship between the complexity of the disorder and sleep disturbance. For example, those with complex PTSD report increased sleep disturbance compared to non-complex PTSD, individuals with PTSD and dissociative symptoms report increased sleep disturbance compared to individuals with PTSD without dissociative symptoms, and those with PTSD who have hyperarousal symptoms report more sleep disturbance than those without hyperarousal symptoms [13, 78, 79]. Likewise, the severity of the trauma may lead to increased sleep disturbance—one study found that survivors of fatal accidents report more sleep disturbance than survivors of non-fatal accidents, and another study of earthquake survivors found decreased TST for children who had to evacuate their homes or whose homes were damaged [80, 81].

Extant research also suggests that sleep disturbance predicts PTSD symptoms over time. For example, among ICU patients, pre-hospitalization sleep disturbance is associated with increased PTSD symptoms 6 months after discharge [82, 83]. Likewise, studies examining sleep disturbance in natural disaster survivors indicate that post-disaster sleep disturbance predicts post-disaster PTSD symptoms over periods ranging from 3 to 12 months [84,85,86,87]. Among veterans, pre-deployment sleep disturbance predicts PTSD symptoms 2 years post-deployment over and above the effect of baseline PTSD symptoms, and pre-deployment sleep disturbance predicts the development of post-deployment PTSD [88•, 89•]. Similarly, post-deployment sleep disturbance predicts PTSD symptoms over 6 and 12 months, but initial PTSD symptoms do not predict subsequent sleep disturbance [90, 91]. A similar unidirectional relationship is found among Palestinian adults [92]. Further, one study examining new mothers found that sleep disturbance at 8 weeks postpartum predicts PTSD symptoms 2 years later, controlling for baseline symptoms [93]. Finally, among motor vehicle accident survivors, sleep disturbance at 10 days following the accident predicts increased PTSD symptoms over 6 months [94]. In contrast, one study found that pre-deployment nightmares, but not insomnia, predicted post-deployment PTSD [95]. Taken together, these findings provide evidence for a prospective association between sleep disturbance and symptoms of PTSD across diverse samples. These results highlight two potential roles for sleep disturbance in the development of PTSD. First, sleep disturbance prior to a traumatic event may confer vulnerability to developing PTSD. Second, sleep disturbance following a traumatic event may amplify or prolong typical stress responses and increase the likelihood of the development of PTSD. In either case, although sleep disturbance is typically considered a symptom of PTSD, findings from recent longitudinal studies suggest that the sleep disturbance observed among those with PTSD may actually reflect an underlying causal process.

Given the high rates of suicide associated with PTSD, research examining processes than may contribute to this relationship is critical [6]. Research has shown that sleep disturbance is associated with suicidal ideation among veterans, even when controlling for mood and anxiety symptoms and PTSD diagnosis [36, 39, 48, 96•]. Further, sleep disturbance predicts suicide attempts over 1 month, controlling for mood symptoms [96•]. Similarly, one study found that sleep disturbance mediated the effect of PTSD on suicidal ideation in a community sample [97]. However, other studies suggest that nightmares may be a unique predictor of suicidal behaviors. Specifically, nightmares are independently associated with suicidal behaviors in a trauma-exposed sample and mediate the relationship between PTSD and self-injury in an outpatient sample [37, 98]. In contrast, one study found no associations between sleep disturbance or nightmares and suicidal ideation in a veteran sample [99]. Despite these conflicting results, the majority of the extant research suggests that aspects of sleep disturbance, including nightmares, are an important predictor of suicidal thoughts and behaviors in trauma-exposed and PTSD samples. This relationship may be due to the negative downstream effects of sleep disturbance on cognitive and emotional function. For example, sleep disturbance leads to increased impulsivity, which may then lead to increased suicidal thoughts and behaviors [100]. Alternatively, difficulties with emotional regulation may account for this relationship [101].

A small body of research has also begun to examine how sleep disturbance may act as a mediating or moderating factor in the relationship between PTSD and other relevant variables. Among those with trauma exposure, sleep disturbance moderates the effect of anxiety sensitivity and emotion regulation difficulties on PTSD symptoms, as well as the effect of PTSD on aggression [102,103,104]. Further, among Israelis, sleep disturbance moderates the effect of the salience of Iranian nuclear threat on PTSD symptoms [105]. Finally, among adults using medical cannabis, sleep disturbance moderates the impact of problematic cannabis use on PTSD symptom severity [106]. These findings suggest that sleep disturbance may amplify maladaptive processes relevant to PTSD. Further, while limited studies have examined sleep disturbance as an intervening variable in PTSD-related processes, one study found that sleep disturbance mediates the relationship between rumination and PTSD in veterans, suggesting that sleep disturbance may play a mechanistic role in the relationship between maladaptive cognitive processes and PTSD [107].

Objective Sleep in PTSD

Recent research utilizing objective measures of sleep has largely found evidence for sleep disturbance in PTSD. In studies comparing veterans with PTSD to healthy controls, the former exhibit decreased TST and sleep efficiency, increased WASO, and increased awakenings [8•, 9, 108]. Likewise, sexual assault survivors with PTSD exhibit increased REM onset latency, as well as decreased sleep efficiency and REM percentage and increased WASO and awakenings in the second half of the night compared to healthy controls [109]. Among mixed trauma samples, those with PTSD exhibit decreased slow wave sleep percentage and time, decreased TST, decreased sleep efficiency, and increased WASO and awakenings [10, 13, 16, 17]. However, a subset of studies comparing veterans with PTSD and mixed trauma samples to healthy controls found no differences on any objective sleep parameters [11, 18, 110]. Although several studies utilizing actigraphy found group differences, the null findings in two of these studies utilizing actigraphy suggest that actigraphy may not be as sensitive as polysomnography in detecting sleep deficits between those with PTSD and healthy controls [8•, 9, 11, 16, 18]. Despite these conflicting findings, the majority of recent research has found evidence for an objective sleep deficit in those with PTSD compared to healthy controls.

In contrast, the recent evidence for increased sleep disturbance in those with PTSD versus trauma-exposed controls is mixed. Although two studies comparing veterans with and without PTSD indicate increased awakenings in the PTSD group, one study found no differences [10, 21, 108]. Similarly, one study comparing trauma-exposed African-Americans with and without PTSD found decreased REM among those with PTSD, but other studies of this population indicate no differences on any objective sleep parameters [111,112,113]. In contrast, recent studies in trauma-exposed women have found that injury patients with PTSD exhibit decreased TST compared to injury patients without PTSD, and sexual assault survivors with PTSD exhibit increased REM onset latency and decreased sleep efficiency and REM percentage and increased WASO and awakenings compared to sexual assault survivors without PTSD [109, 114]. Taken together, these findings suggest that in some trauma-exposed samples, those with and without PTSD are not distinguished by sleep disturbance. However, recent findings suggest that among women, sleep disturbance may be a unique predictor of PTSD and healthy sleep may predict resilience. Recent evidence suggests that women exhibit better sleep than men in healthy samples [115]. Thus, baseline healthy sleep may buffer the impact of trauma on subsequent sleep health in women. Alternatively, men may be more vulnerable to increased sleep disturbance in response to a stressor. Future research should examine gender differences in the relationship between sleep and PTSD symptom severity among those with PTSD and trauma-exposed controls.

Few recent studies have examined the relationship between objective sleep disturbance and PTSD symptom severity, and the results are mixed. One study found that decreased TST was associated with PTSD symptom severity in a veteran sample [67]. In contrast, though one study of African-American adults found a link between PTSD symptoms and REM onset latency and percentage, another study found no association between objective sleep parameters and PTSD symptoms in this population [116, 117]. Additional research is necessary to clarify the link between objective sleep and PTSD symptom severity. Similarly, recent studies comparing those with PTSD to other psychiatric samples have also found mixed results. Though one study found that veterans with PTSD exhibit increased WASO and awakenings compared to a mixed psychiatric sample, another study found no differences between veterans with PTSD and civilians with MDD on any sleep parameters [118, 119].

Discrepancies between Sleep Measurements in PTSD

Previous research suggests that objective and subjective measures of sleep are often discrepant in both healthy and clinical samples [120, 121]. Recent studies utilizing both objective and subjective sleep methods in PTSD populations has likewise yielded discrepancies between assessment modalities. For example, a study of women with PTSD found no association between subjective and objective measures, and only global sleep questionnaires were linked to PTSD symptom severity [122]. Likewise, one study found that veterans with PTSD underestimate TST and sleep efficiency and overestimate SOL compared to actigraphy [123]. Interestingly, this study also found that veterans with PTSD underestimate number of awakenings, which suggests that those with PTSD do not globally overestimate sleep disturbance. One recent study also indicates that the degree of discrepancy between objective and subjective sleep may depend on the measures utilized. In a sample of African-American adults with PTSD, participants underestimated WASO compared to actigraphy, but not polysomnography, and sleep parameters measured by polysomnography and actigraphy were correlated with sleep parameters reports from a sleep diary, but not a questionnaire of sleep over the past month [124].

Taken together, these findings suggest that subjective and objective measures of nightly sleep yield similar results, while global sleep reports are more discrepant. Likewise, it is unclear whether the relationship between global sleep reports and PTSD symptom severity represents an effect of subjective perception or increased time. Additional research utilizing objective measures over longer periods of time is necessary to clarify this discrepancy. Finally, given that the vast majority of recent research on sleep disturbance and PTSD has utilized subjective measures, these discrepancies between methods highlight the importance of utilizing a multimethod approach to fully understand the role of sleep disturbance in PTSD.

Sleep Outcomes in the Treatment of PTSD

Given the considerable evidence for links between sleep disturbance and PTSD, intervention efforts targeting sleep may prove useful in the treatment of PTSD. Extant findings from such treatment studies suggest promising results. A recent review of pharmacological interventions found that prazosin, an adrenergic inhibitor, improves sleep disturbance, nightmares, and PTSD symptoms [125]. In contrast, standard psychopharmacological interventions, such as antidepressants and benzodiazepines, have yielded inconsistent results for the treatment of sleep-related symptoms in PTSD, and some studies indicate an adverse impact of antidepressants on sleep health among those with PTSD [125]. Several studies have found that diverse non-pharmacological intervention efforts, including virtual reality exposure combined with d-cycloserine, cognitive behavioral social rhythm therapy, cognitive therapy, yoga, eye movement desensitization and reprocessing, and mantram repetition, improve both subjective sleep disturbance and PTSD symptoms among individuals with PTSD [126,127,128,129,130,131]. Likewise, cognitive behavioral therapy for insomnia (CBT-I) improves PTSD symptoms and both objective and subjective sleep [132, 133]. Further, one study of cognitive processing therapy found that improvements in subjective sleep disturbance predicted improvements in PTSD symptoms, and another study found that those whose sleep improved following CBT-I reported decreased PTSD symptoms [134•, 135]. Studies examining exercise interventions for PTSD indicate reductions in sleep disturbance and hyperarousal symptoms among those with poor baseline sleep [136, 137]. In contrast, one study of imagery rehearsal therapy found that treatment improved PTSD symptoms and nightmares, but not sleep disturbance, in a mixed psychiatric sample, and another study found that yoga improved sleep but not PTSD symptoms in a trauma-exposed sample [138, 139]. These findings may be due to the discrepancies in the samples utilized. Overall, extant research suggests that reductions in sleep and PTSD symptoms co-occur across multiple treatment modalities.

While some studies indicate sleep-related interventions improve sleep disturbance in those with PTSD and trauma-exposed veterans, other studies indicate that sleep problems persist even after treatment [140, 141]. For example, one study found that both cognitive processing therapy and prolonged exposure reduced but did not eliminate sleep disturbance in sexual assault survivors [142]. Similarly, a study comparing cognitive processing therapy and present-centered therapy found that sleep disturbance was one of the most persistent symptoms following treatment in both conditions [143•]. Thus, while several studies indicate a reduction in PTSD symptoms and sleep disturbance across multiple intervention methods, residual sleep disturbance may prove resistant to current treatments. These findings indicate that current PTSD treatments that do not target sleep disturbance are insufficient, and treatment approaches should include sleep interventions for optimal treatment outcomes. Further, these findings also suggest that sleep may not be a mere epiphenomenon of PTSD. That is, if there was a unidirectional relationship between PTSD and sleep disturbance, it would follow that remission of PTSD should result in improved sleep. In contrast, 57% of those whose PTSD remitted following non-sleep targeting treatment continued to report sleep disturbance [143•]. These findings support the hypothesis that sleep disturbance may confer vulnerability for the development of PTSD. Future research should examine whether reductions in sleep disturbance account for reductions in PTSD symptoms, as well as the impact of residual sleep disturbance on treatment outcome.

Conclusions and Future Directions

Recent research indicates a link between sleep disturbance and PTSD. Overall, those with PTSD exhibit and report increased sleep disturbance compared to healthy controls. Thus, while studies comparing objective and subjective measures of sleep indicate that those with PTSD tend to overestimate their sleep disturbance, these individuals do exhibit an objective sleep deficit. These findings have important treatment implications. That is, treatments of PTSD should target both sleep health and the perception of sleep quality.

In contrast to studies comparing those with PTSD to healthy controls, recent comparisons of those with PTSD and trauma-exposed controls have yielded inconsistent results. Compared to trauma-exposed controls, those with PTSD consistently report increased sleep disturbance, but objective comparisons of these groups is mixed. Thus, the relationship between objective sleep disturbance and trauma exposure remains unclear. Interestingly, women with PTSD may be particularly likely to exhibit increased sleep disturbance compared to trauma-exposed women. Future research assessing gender differences in objective sleep prior to and following trauma exposure is necessary to clarify the relationship between gender, sleep disturbance, and PTSD.

Consistent with previous findings, recent evidence highlights the discrepancies between various methods of assessing sleep among those with PTSD. Recent findings are mixed, but suggest that those with PTSD report increased sleep disturbance compared to objective findings. However, it may be important to consider the method of sleep assessment. That is, night to night methods, such as polysomnography, actigraphy, and sleep diaries, may be more congruent, while global sleep reports may disagree with these measures. Given the link between PTSD symptom severity and global sleep reports, measures of nightly sleep may not span sufficient time to capture the association between sleep and PTSD symptom severity. Taken together, these findings highlight the importance of utilizing both objective and subjective measures, as well as consideration of time in the assessment of sleep and PTSD.

Finally, recent evidence regarding the treatment of sleep disturbance in PTSD suggests promising results. Recent studies indicate that a diverse range of treatments results in reduction in both sleep disturbance and PTSD symptoms and improvements in sleep predict improvements in PTSD. Further, CBT-I improves both objective and subjective sleep among those with PTSD. However, recent findings also point to evidence for residual sleep disturbance following treatment. Taken together, these results suggest that targeting sleep disturbance may be a useful treatment modality for PTSD, but additional research is necessary to develop treatments that can fully address sleep disturbance in PTSD.

An important question in the study of the relationship between sleep disturbance and PTSD is that of temporal causality. That is, is sleep disturbance simply a symptom of PTSD, or does sleep disturbance precede the development of PTSD? Evidence from recent longitudinal studies indicates that pre-trauma sleep disturbance predicts post-trauma PTSD symptoms and that post-trauma sleep disturbance predicts subsequent PTSD symptom severity when controlling for baseline symptom severity. These findings offer preliminary evidence for a causal role of sleep disturbance in the development of PTSD. Recent evidence that sleep disturbance persists after successful PTSD treatment also refutes the idea that sleep disturbance is an epiphenomenon. Likewise, evidence that treatment of sleep disturbance predicts improvements in PTSD symptoms also suggests that sleep disturbance is part of the etiology of PTSD. Pre-trauma exposure sleep disturbance may represent an underlying vulnerability to the development of PTSD following trauma exposure. The combination of the negative effects of sleep loss for physiological, cognitive, and emotional function may create a multi-system diathesis for PTSD that is expressed upon exposure to a traumatic event [144,145,146]. Importantly, these results suggest that intervening on sleep disturbance following trauma exposure may reduce the likelihood of developing PTSD and/or may buffer PTSD symptom severity. Similarly, targeting sleep disturbance prior to likely trauma exposure, as in the case of military personnel may, prevent the development of PTSD.

Although considerable research has examined the association between sleep and PTSD, few studies have moved beyond measuring sleep in PTSD to attempt to identify specific mechanisms that may account for this relationship. Findings from sleep deprivation research offer preliminary candidate mechanisms, including dysregulated cortisol, diminished cognitive control, decreased functional connectivity between medial prefrontal cortex and amygdala, and increased stress reactivity [144,145,146,147,148]. Future research should test specific mechanisms such as these to better understand why sleep disturbance may confer vulnerability for PTSD.