Traumatic brain injury and sleep disturbances in combat-exposed service members and veterans: Where to go next?

1.1Sleep disturbances in military populations

Deployed active duty SMs are expected to perform optimally in high-risk operational environments where maintaining a normative sleep schedule is often not possible, such as night missions or forced extended wakefulness during multi-day missions. Further, chronic insufficient sleep has been shown to compromise safety, which suggests sleep restricted combat-exposed soldiers may be at an increased risk for sustaining mission-related injuries that compromise individual and unit-level readiness (e.g., TBI).

Although sleep loss is inherent to operational missions, chronic sleep loss is also prevalent among non-deployed SMs. Military personnel report sleep disturbances (e.g., insufficient or non-restorative sleep, insomnia, nightmares) before, during, and after deployment at alarming rates (Breen, Blankley, & Fine, 2017; Caldwell, Knapik, & Lieberman, 2017; de Dassel, Wittmann, Protic, Hollmer, & Gorzka, 2017; Mysliwiec et al, 2013; Troxel et al, 2015). Numbers have continuously grown over the years and have well surpassed prevalence and incidence rates of the general population. Further, upwards of 75% of deployed SMs rate the quality of their sleep as significantly worse while in theater when compared to before deployment (Peterson, Goodie, Satterfield, & Brim, 2008), many of whom meet for clinically significant insomnia. It is estimated that over 87% of SMs report at least one sleep disturbance, with the most commonly reported being insomnia, sleep fragmentation, obstructive sleep apnea, and hypersomnia (Bramoweth & Germain, 2013; Mysliwiec, Gill, et al., 2013; Schiehser et al., 2017). Approximately 40% of soldiers report more than one sleep disturbance (Gill et al., 2014; Mysliwiec, Gill, et al., 2013), making the selection of targeted treatments exponentially more complex. Sleep disturbances are a critical issue facing the US military due to their impact on maintaining force readiness, health, well-being (Caldwell et al., 2017). Further, the economic burden of providing both acute and long-term behavioral health services related sleep disturbances has become a major concern. It is important to note that studies examining prevalence of sleep disturbances in military samples often ignore, or at the very least downplay, the contribution of pre-military service or pre-deployment sleep factors. Therefore, greater representation of baseline sleep parameters is needed in future research in order to fully examine the trajectory of sleep disturbances in military groups.

1.2Sleep disturbances in combat-related TBI

Sleep disturbances are one of the most prevalent sequelae of TBI (AFHSC, 2013). In a sample of 114 post deployment military SMs and veterans who had sustained blast-related mTBI, 77% endorsed experiencing sleep disturbances (Farrell-Carnahan, Franke, Graham, & McNamee, 2013). Collen and colleagues reported that 97% of soldiers with TBI reported at least one sleep disturbance (Collen, Orr, Lettieri, Carter, & Holley, 2012). These self-reported sleep disturbances were corroborated by polysomnographic sleep/wake studies and included hypersomnia (85%), insomnia (55%), sleep fragmentation (54%), and obstructive sleep apnea syndrome (OSAS; 35%). Sleep disturbances were more severe among SMs with comorbid TBI and posttraumatic stress symptoms (PTSS) or anxiety. There is preliminary evidence that blast- vs. blunt force-related TBI may be accompanied by distinct sleep disturbance profiles. Specifically, Kempft and colleagues reported that blast injuries were associated with insomnia and elevated anxiety symptoms, whereas blunt force injuries were more commonly associated with OSAS. These sleep disturbances are chronic: Over 60% of those with TBI continued to endorse clinically significant sleep disturbances three years post-injury (Kempf, Werth, Kaiser, Bassetti, & Baumann, 2010).

Sleep disturbances can also impede the TBI rehabilitation process (Bell et al., 2018; Sandsmark, Elliott, & Lim, 2017) and are risk factors for the development of anxiety and mood disorders following stress exposure (Babson & Feldner, 2010; Bryant, Creamer, O’Donnell, Silove, & McFarlane, 2010; Gehrman et al., 2013). Realizing the impact TBI has on functional outcomes and quality of life, as well as the role of disturbed sleep as a risk factor for the development and maintenance of death, disease, and disability, the comorbidity between TBI and chronic sleep disturbance may increase the risk for poor health outcomes among SMs and veterans relative to the general population (Barr et al., 2015).

The purpose of this review is to synthesize the literature focusing on sleep disturbances and TBI as it pertains specifically to military SMs and veterans. This review uniquely contributes to the existing science due to the overall scarcity of prior summaries of sleep disturbances and TBI, as well as existing reviews focusing primarily on the general population (see Barshikar & Bell, 2017; Bogdanov, Naismith, & Lah, 2017; Orff, Ayalon, & Drummond, 2009; Ponsford et al., 2012 for reviews), on sleep disturbances occurring post-TBI (Bogdanov et al., 2017; Ponsford et al., 2012), or that isolate only one sleep disturbance (i.e. insomnia; Sandsmark, Elliott, & Lim, 2017). First, we discuss the nature and prevalence of sleep disturbances based on varying TBI assessment methods. Specific sleep profiles that may guide diagnostic and treatment efforts targeting disruptive sleep in TBI relative to other diagnostic groups are then suggested. Second, a summary of findings regarding the role of sleep in the trajectory of TBI is presented. Finally, critical knowledge gaps that, if filled, have a high potential for enhancing sleep health and functional outcomes in SMs and veterans with TBI are discussed.

3.1Time course of sleep disturbances & TBI

Two of the 41 papers selected explicitly sought to evaluate the time course between sleep disturbance and TBI is SMS or veterans. Only one study to date at the time of this publication assessed sleep in the acute phase following TBI. Norris showed that TBI with loss of consciousness (LOC) was associated with greater self-reported difficulty sleeping compared to TBI without LOC in 210 active duty SMs who reported sustaining an mTBI to a US military-operated concussion clinic in Afghanistan. The greatest self-reported sleep difficulty associated with LOC was endorsed at 72 hours post-injury and again 48–72 hours after the initial TBI assessment (Norris, Sams, Lundblad, Frantz, & Harris, 2014). These findings suggest that the onset of sleep disturbance can occur immediately following the point of injury and are strongly associated with LOC.

The Armed Forces Health Surveillance Center ([AFHSC]; 2013) reviewed the electronic medical records of 821,611 SMs with TBI at 3- and 12-months post injury from a 13-year surveillance period examining 14 common TBI-related symptoms or conditions. Severity of sleep disturbance had the second greatest increase from three to 12 months across all levels of TBI severity when compared to a non-TBI control group. Other highly prevalent symptoms included headache, alcohol and substance abuse, and PTSD—all sequelae or comorbidities of TBI with known links to disrupted sleep. (AFHSC, 2013).

3.2Prevalence & characterization of sleep disturbances in TBI

Three studies focused primarily on estimating the prevalence of sleep disturbances in SMs and veterans with combat-related TBI using a wide variety of assessment tools including, self-report subjective and objective questionnaires, medical records review, objective sleep measurement (e.g., actigraphy, polysomnography (PSG)) and clinical interviews conducted in different settings (e.g., sleep clinics, laboratories). In a study by Collen et al (2012), one-hundred sixteen OEF/OIF combat-exposed active duty SMs with documented mild to moderate TBI underwent a comprehensive sleep evaluation, including PSG and subjective assessments of sleep quality and daytime somnolence in a hospital sleep medicine clinic. Almost all (97.4%) SMs experienced sleep disturbances, with the most common being hypersomnia, sleep fragmentation, insomnia, and OSAS (Collen et al., 2012). Using a large self-report battery of health assessments, Gill and colleagues estimated that one-third of 110 combat-exposed SMs with diagnosed sleep disorders and sleep disturbances also met criteria for two or more service-related disorders (Gill et al., 2014). The most commonly reported sleep disturbances were insomnia and OSA. SMs with sleep disturbances were divided into two groups: low-comorbidity (0–1 comorbid conditions) and high-comorbidity (2–3 comorbid conditions). SMs in the high comorbidity group were significantly more likely to be diagnosed with PTSD, depression, and/or TBI symptoms compared to SMs in the low comorbidity group. In combat-exposed active duty SMs, the percentage of soldiers endorsing two or more service-related sleep disturbances has been reported to be as high as 47.3% (Mysliwiec, Gill, et al., 2013; Mysliwiec, McGraw, et al., 2013).

Capaldi and colleagues isolated sleep features present in 69 active duty combat-exposed veterans with TBI and sleep disturbances through an electronic medical record review (including PSG, OSAS, and daytime sleepiness data) and found that frequency of trouble sleeping, number of nighttime awakenings, and daytime sleepiness were elevated across all TBI injury severity groups (Capaldi, Guerrero, & Killgore, 2011). Compared to SMs without TBI, SMs with TBI are three times more likely to report sleep disturbances immediately following deployment (Macera, Aralis, Rauh, & MacGregor, 2013). Focusing on more chronic, post-deployment sleep disturbances, Tanev and colleagues reported that veterans with TBI history had higher rates of OSAS, hyposomnia, REM sleep behavior disorder, narcolepsy, periodic limbic movements, sleep bruxism and increased slow wave sleep relative to veterans without TBI (Tanev, Pentel, Kredlow, & Charney, 2014).

The variable prevalence estimates reported across studies of SMs and Veterans with TBI are likely a reflection of different assessment methods (e.g., PSG vs. self-report questionnaires) and a bias that may be introduced by studying patients in different settings (e.g., sleep clinics vs. general survey).

3.3Influence of TBI-related factors on sleep disturbances

Eight studies suggested that injury-related factors may influence the nature and trajectory of sleep disturbances in SMs and veterans, as well as heath factors and functional impairment.

The mechanism of injury leading to TBI is a primary factor that may influence sleep disturbances in military samples. In the sample previously referenced by Collen and colleagues, participants with blast TBI reported significantly worse insomnia, whereas those with blunt TBI had greater levels of OSAS (Collen et al., 2012). Petrie et al. showed that when compared to veterans with TBI history without blast exposure, veterans with TBI history and blast exposure endorsed higher levels of sleep disturbances overall (Petrie et al., 2014). Furthermore, blast exposure followed by concussive symptoms has been associated with sleep disturbances, but this relationship was not found in those exposed to blast without current symptoms. These findings suggest that both blast- and blunt-TBI may increase risk for specific sleep disturbances, and that blast exposure may be especially important to the development of sleep disturbances in military samples, especially when concussive symptoms follow the initial injury.

TBI severity may also impact the severity of sleep disturbances and injury prognosis. In one of the few studies measuring sleep disturbances longitudinally after TBI, the AFHSC used electronic medical records to track changes in sleep disturbances change from three to 12 months post injury in SMs and veterans. Results showed that severe TBI was associated with the greatest increase in occurrence of new sleep disorders at 12 months relative to mild and moderate TBI, those with prior TBI, and SMs and veterans with no current symptoms (AFHSC, 2013).

LOC has also shown to be significantly associated with more sleep disturbances, in particular nightmares, in those with TBI compared to those with TBI without LOC (Farrell-Carnahan et al., 2013). As detailed above, Norris et al. also found that LOC at time of injury was associated with more severe sleep disturbances (Norris et al., 2014).

The number of TBIs may also be of particular importance when estimating the prevalence of sleep disturbances in military samples with TBI. Bryan assessed the prevalence of clinical insomnia across TBI frequency and found compared to those with zero TBIs, four times as many participants with one TBI and 10 times as many participants with multiple TBIs met for clinical insomnia (Bryan, 2013). These results remained robust when controlling for PTSD, depressive symptoms, and post-concussive symptoms.

Finally, the presence or absence of prior TBI may also impact sleep disturbances in TBI. In a longitudinal study of the relationship between TBI status and sleep disturbances, Holster et al. found that Air Force veterans with a pre-deployment history of TBI endorsed more insomnia symptoms at baseline when compared to those without prior TBI history (Holster, Bryan, Heron, & Seegmiller, 2017). In turn, baseline insomnia symptoms were associated with increased risk for sustaining a TBI during deployment. Chronic insomnia at baseline was also associated with both pre-deployment TBI history and TBI during deployment. It is important to note that differences have been observed in neurological changes between military SMs with blast exposure occurring both within and greater than 10 meters that were not influenced by sleep quality (Robinson et al., 2015).

Of note, none of the articles reviewed accounted for the timing of injury occurrences in those with more than one TBI. This is likely a critically important factor to consider when assessing the impact of TBI frequency and prior TBI on sleep disturbances. More specifically, it is important to consider the timing between the injuries themselves. When full recovery from one injury does not preceded the onset of a second injury, symptoms are likely to exacerbate and become pervasive. Thus, summative TBIs may result in unique profiles of sleep disturbances that influence treatment effectiveness and outcomes.

Together, the findings to date unequivocally establish a strong relationship between TBI and sleep disturbances in SMs and veterans. Furthermore, a number of factors related to injury etiology may modulate or mediate this relationship but remain underexplored. Larger samples and more comprehensive study designs are needed to further characterize this relationship and determine whether assessment and treatments strategies should be informed by these factors. Recognizing the challenges of conducting longitudinal studies, access to medical records would provide a more efficient and cost-effective study design for better determining estimates of time between injury and symptom onset.

3.4Interactions between sleep disturbances, TBI, and comorbid conditions

Sleep disorders and TBI are complex and heterogeneous conditions with shared symptomology that and are both independently and jointly associated with psychiatric and medical comorbidities. Twenty-four of the studies included in this review reported on comorbid psychiatric and medical conditions of TBI. For instance, Gill et al. showed severity of PTSD, sleep disturbances, and depressive symptoms increase as the number of comorbid conditions increase in SMs with TBI (Gill et al., 2014). The combination of these conditions complicates assessment and management strategies and uniquely impacts functional outcomes. TBI resulting from combat exposure has been shown to most commonly co-present with psychological trauma-related disorders (i.e. PTSD), chronic pain, and insomnia in over 51% of veterans (Lang, Veazey-Morris, & Andrasik, 2014). The tight relationship between TBI, sleep disturbance and psychiatric conditions has risen the possibility that sleep may serve as a common factor linking TBI with its comorbidities, which in tune may suggest a summative effect of co-occurring conditions and TBI on sleep disturbances relative to TBI alone. Lew and colleagues observed that sleep problems were more severe when PTSD was comorbid with TBI compared to those with TBI alone (Lew et al., 2010).

Furthermore, Macera et al. observed that veterans who screened positive for TBI and PTSD were 3-times more likely to report sleep problems compared to TBI alone. This suggests that the relationship between TBI and its associated comorbidities may even be moderated or mediated by sleep disturbances (Macera et al., 2013). This finding was supported in two studies by Verfaellie et al. and Waldron-Perrine that each reported sleep disturbances mediated the relationships between emotional distress, cognitive functioning, and self-reported post concussive symptoms more so than TBI alone (Verfaellie, Lee, Lafleche, & Spiro, 2016; Waldron-Perrine, Hennrick, Spencer, Pangilinan, & Bieliauskas, 2014). In one of the few studies on pre-TBI sleep factors, Elliott et al. showed that the resiliency personality phenotype predicted less sleep troubles in military samples who go on to develop TBI when compared to other phenotypes (Elliott et al., 2017). This finding suggests that pre-injury factors may play a role in the course of TBI and provides support for future research examining pre-injury sleep factors in military SMs and veterans prior to deployment. A greater understanding of the role of sleep in the trajectory of TBI may yield novel, modifiable sleep-related targets for preventive efforts or engagement through sleep-focused, individualized interventions aimed at mitigating symptoms and restoring function through improved sleep.

Stocker and colleagues examined positron emission tomography (PET) scans of OEF/OIF veterans with and without comorbid blast TBI and PTSD and found that there was a detectable decrease of regional cerebral metabolic rate of glucose during REM sleep and wakefulness among veterans with prior blast TBI than those without blast TBI history when adjusting for any effects of PTSD (Stocker et al., 2014). It was suggested that the areas of glucose hypometabolism, including the right basal ganglia, amygdala, hippocampus and parahippocampal gyrus, culmen, associative visual cortices and medial frontal cortices may be susceptible to long-term effects of blast TBI, including chronic concussive symptoms and/or maladaptive responses to stress (Stocker et al., 2014). A separate study examining objective sleep features in a comorbid TBI/PTSD sample from Stocker and colleagues revealed that REM sleep fragmentation may be a potentially sensitive index to the synergistic effects of TBI and PTSD, as the only significant interaction effect between TBI and PTSD was observed in the amount of REM sleep fragmentation (Stocker et al., 2016). A summative effect of co-morbid TBI on sleep problems has also been observed in clinical settings, where individuals at a polytrauma clinic had the highest rates of sleep disturbance, anxiety, and depression (Pugh et al., 2014).

Wallace and colleagues showed that insomnia severity was not significantly different between those with PTSD/TBI and PTSD alone (Robinson et al., 2015; Wallace et al., 2011). These results do contrast the literature; however, it is likely that the difference in findings is due to the heterogeneity of the sample when using sleep quality as an outcome or other factors unique to neurological changes.

Finally, those with comorbid TBI and PTSD report more daytime sleepiness despite objective measures indicating they have greater total sleep time relative to those with TBI alone (Tanev et al., 2014). This finding was replicated in a study by Wallace et al (Wallace et al., 2011) and highlights the importance of using multimodal data collection methods to best understand the complex relationships between TBI and its common comorbidities and sleep disturbances. Self-reported sleep disturbances provide rich subjective data, that when taken in combination with objective sleep measures is likely to better support optimal intervention selection compared to intervention selection based solely from objective data sources.

An enhanced understanding of how sleep disturbances present in TBI when comorbidities are present and how the sleep profiles of comorbid groups compare to those with TBI alone is needed to improve assessment and guide optimal courses for intervention (Lang et al., 2014). This line of research is especially important considering around 50% of SMs and veterans experience three or more deployment-related conditions, and that those with three or more deployment-related conditions are more likely to present with substantial disability relative to those with TBI alone (Lippa et al., 2015). Nevertheless, the literature to date highlights that more severe sleep disturbances occur when TBI presents with co-morbid conditions.

3.5Sleep disturbances: Distinct domain or secondary condition?

Findings from two studies using factor analysis suggest a TBI diagnosis presents with both overlapping and distinct symptom clusters with its most common comorbid conditions (e.g., PTSD), and even highlights sleep disturbances as a distinct cluster that transcends diagnostic categories (Jaramillo et al., 2015; Maguen, Lau, Madden, & Seal, 2012).

These studies identified distinct symptom clusters (i.e. low/high distress) across larger diagnostic groups that may improve the ability to differentiate SMs experiencing comorbid conditions with shared symptomology. While these studies chose to cluster on symptoms and not diagnoses, results did show some clusters were more likely than others to contain a particular diagnosis. For example, a ‘moderate/high distress’ cluster represented by headache and memory impairment capturing a high rate of those with TBI (Jaramillo et al., 2015) or a high rate of those with PTSD being represented in a ‘high distress’ or ‘pain and behavioral health’ cluster represented by anxiety and substance abuse (Jaramillo et al., 2015). Maguen and colleagues used factor analysis to differentiate distinct from overlapping screening-based symptoms in 1549 veterans with TBI, PTSD, and depression from postdeployment screening data (Maguen et al., 2012). Four constructs merged, three of which were represented by the diagnoses themselves (i.e. PTSD, TBI, depression) and an intriguing 4^th construct: “hypervigilance and sleep problems” (Maguen et al., 2012). The findings from Maguen’s study show symptoms unique to TBI that may help improve assessment and targeted treatments when attempting to separate TBI from its comorbid conditions. Arguably, the most notable finding is the evidence for a distinct ‘hypervigilance/sleep disturbance’ cluster that highlights the needs to further examine hyperviligence and disrupted sleep symptoms as a unique, combined factor influencing post-deployment health in military groups with TBI. Overall, the findings from studies using factor analysis suggest that while the co-occurring diagnoses themselves may be difficult to differentiate, distinctive phenotypes may exist that can aid more optimal assessment and selection of treatment targets.

Pugh et al., and Lippa et al., both showed that although disrupted sleep was distinct from both TBI and PTSD, TBI and PTSD failed to be different from one another across sleep variables (Lippa et al., 2015; Pugh et al., 2014). Capaldi and colleagues also examined the specificity of sleep disturbances related to TBI and PTSD in 69 active duty combat veterans and found that PSG failed for the most part to significantly differentiate TBI from PTSD (Capaldi et al., 2011). However, those with PTSD did have a higher frequency of nighttime arousals and those with TBI had more slow wave sleep. This finding may be reflective of arousal brain states linked to sleep in those with PTSD (Germain, Buysse, & Nofzinger, 2008) theories supporting slow wave sleep as a critical component for recovery after TBI (Shekleton et al., 2010).

3.6Sleep-focused interventions may improve TBI symptoms

TBI interventions rarely include sleep-specific assessments and strategies, and few studies have used sleep-focused interventions to assess TBI symptom change and improvements in functional outcomes in military SMs and veterans (Gilbert et al., 2015).

As previously highlighted, OSA is a prevalent sleep disorder in SMs and veterans with TBI. Barr et al reported over 50% of recently deployed military SMs with TBI endorsed more restorative sleep and improved health-related quality of life (HRQoL) of life at 12 months following provision of standard care, which included continuous positive airway pressure (C-PAP) therapy and/or cognitive behavioral therapy (Barr et al., 2015). Additionally, those with TBI who experienced improved sleep post-treatment also saw a reduction of PTSD and depression symptoms at 3-months. Mysliwiec et al also showed reduced daytime sleepiness and improved sleep quality in 58 male military personnel with and without TBI following autotitrating positive airway pressure (APAP) (Mysliwiec et al., 2013). Participants who improved used APAP for four hours or more a night for at least 70 percent of all nights during a consecutive one-month period. Participants in the improved sleep group also experienced functional improvement related to depressive symptoms, energy/fatigue, and emotional well-being. Although these two studies used a treatment specific to OSAS, findings support the modifiable nature of sleep in TBI and show the potential for using sleep-focused interventions for improving not only sleep, but also health and functioning.

Ruff et al. used a 9-week combination treatment consisting of a behavioral intervention (i.e. sleep hygiene counseling) and prazosin in 126 veterans with blast mTBI to assess change in disrupted sleep and daytime sleepiness (Ruff, Ruff, & Wang, 2009). Disrupted sleep was assessed through two questions on whether or not participants experienced impaired sleep and what they believed to be the nature of their sleep impairment and daytime sleepiness was measured by the Epworth Sleepiness Scale (ESS). At post-treatment, those in the combination treatment group reported, on average, improved self-report sleep quality and decreased or complete remission of nightmares. Furthermore, 100 percent of participants saw a significant reduction in daytime sleepiness. A second study by Ruff and colleagues using a 9-week combination behavioral-pharmacological treatment of sleep counseling and prazosin in 63 veterans with combat mTBI with LOC saw a greater than 50 percent post-treatment reduction in daytime sleepiness in over 78% of participants, as measured by the ESS (Ruff, Riechers, Wang, Piero, & Ruff, 2012). This greater than 50% reduction in daytime sleepiness was maintained and improved upon (78 % to 89%) at a 6 month follow up. It is important to note that this study also found improved outcomes with behavioral and pharmacological interventions alone, which is evidence of the potential relative utility of these approaches.

Finally, in the only randomized clinical trial on sleep disturbances and TBI known at date, Vuletic et al tested the effectiveness of a 6-month telephone-based problem-solving treatment (PST) intervention for improving sleep quality in 356 active duty SMs with TBI and poor baseline sleep quality (Vuletic et al., 2016). The PST intervention involved 12 bi-weekly phone calls addressing participated-selected issues related to their sleep. Results showed that when compared to an education only (i.e., educational brochures on sleep) control group, the PST group saw improved sleep quality at 6 months, but not at 12 months, as measured by the Pittsburgh Sleep Quality Index (PSQI). This finding suggests PST may be an effective treatment to improve sleep acutely in military cohorts, but more work needs to be done to determine the long-term effectiveness of this approach.

In a community-based, single-blind longitudinal study, Wiseman-Hakes and colleagues examined objective and self-report cognitive and emotional rehabilitation outcomes in adults with TBI and co-morbid sleep/wake disturbances (2013). This study used individualized sleep/wake disorder treatment, which included sleep hygiene, pharmacological intervention, and sleep apnea interventions. Statistically significant post-intervention changes in mood, performance, and cognitive symptoms were observed—suggesting sleep-focused treatments may help to improve rehabilitation outcomes in those with TBI across all levels of severity.

When examining the perspectives of 19 male veterans with post-acute TBI and comorbid sleep-wake disorder enrolled at the VA at least 1-year post-injury using semi-structured sleep-related interviews, Matthews and colleagues qualitatively extrapolated two perspective dimensions based on qualitative analysis of responses: “messed up sleep” and surviving/management of sleep disturbances (Matthews, Signoracci, Stearns-Yoder, & Brenner, 2016). This sample of veterans also reported that their sleep troubles had a significant negative impact on daily living. Specifically, difficulty falling asleep, nighttime awakenings, and poor sleep quality were cited as having the greatest adverse impact on overall daytime functioning. This study suggests veterans recognize the challenging nature of managing sleep disturbances and that sleep problems can be difficult to manage.

Findings from these studies provide preliminary support for the effectiveness of sleep interventions to improve not only sleep symptoms, but also health and functioning in military samples with TBI. These studies suggest that both sleep-focused pharmacological and behavioral interventions may be efficacious for improving outcomes for SMs and veterans with TBI either individually or in combination. Albeit promising, these studies warrant confirmatory controlled effectiveness trials. It is important to note than only a scarce body of literature to date has looked at how sleep interventions may improve TBI symptoms in military SMs and veterans (Ruff et al., 2012)—another underexplored area to guide future research efforts.

4.1Clinical and research implications

Historically, disturbed sleep has been viewed as a symptom of TBI. This implies a temporal relationship where TBI onset precedes the development of sleep symptoms. Much of the literature in this review examined the co-occurring nature of sleep disturbances and TBI or disturbed sleep as a mediator or moderator in the relationship between TBI and poor outcomes in combat-exposed military samples. Furthermore, it is known that sleep disturbances are associated with reduced overall functioning in military samples with TBI (Ettenhofer, Melrose, Delawalla, Castellon, & Okonek, 2012) and that improving sleep can lead to TBI symptom reduction and increased HRQoL (Barr et al., 2015). Thus, disrupted sleep likely plays a much greater role in the course of TBI—one where sleep disturbances and TBI share bidirectional and even cyclical relationship (Gilbert et al., 2015). Future research efforts are needed to examine the full trajectory of sleep disturbances and TBI in military groups.

Although the current state of the field is keenly attuned to sleep disturbances profiles following TBI, research examining TBIs sustained among individuals with preexisting sleep disturbances remains scarce. There is even less research related to the examination of whether sleep/wake difficulties can be conceptualized as presenting independently of TBI rather than as a symptom of TBI. Although disturbance in normative sleep may be a manifestation of the neurobiological changes of TBI in civilians and veterans alike, it is possible for that for military SMs, the nature of military service, specifically while deployed, may give rise to disrupted sleep patterns prior to sustaining a TBI—an area yet to be studied. This is an intriguing line of research to explore not only due to the ability to demonstrate sleep disturbances as a risk factor for injury to the head, but given the existing literature linking inadequate sleep to increased risk for physical injuries it also warrants examining the effect of disrupted sleep on other physical injuries while in the field. Further, it could be argued that just as sleep disturbances can develop during deployment in those who do not sustain TBI, sleep disturbances could develop subsequent to a TBI and be better accounted for by the same factors as those without TBI than the TBI itself.

As our previous studies (Germain et al., 2014; Germain, Shear, Hall, & Buysse, 2007) have demonstrated the effectiveness of brief behavioral sleep treatments in SMs and veterans with deployment-related comorbidities, behavioral sleep treatments offer a promising treatment approach. There is also a need to improve treatment approaches for OSA (to improve adherence), better extend trials on hypersomnia to military populations, and identify and test novel pharmacological agents for improving sleep in SMs that do not violate military drug policies. Research has shown that both military SMs and veterans are willing to try novel approaches to improving their sleep, which establishes this cohort as an engaged population (Matthews et al., 2016).

4.2Limitations

Although this broad overview of the literature on TBI and sleep disturbances in military samples may lack the rigor that comes with a more exhaustive systematic review or even meta-analysis, this review was intended to gather a diverse representation of the research that has been conducted to date. This method allowed for visual inspection of any skewness in the current literature, which this review did demonstrate. This approach also allowed for the identification of knowledge gaps that will be critical to fill moving forward to advance the state of the science on combat-related TBI and sleep disturbances. Scoping reviews are commonly implemented when the goal is to lay the foundation for a guide for when working with specific cohorts and when the extant literature would not be appropriate to summarize through systematic review (Peters et al., 2015). With that in mind, the nature of this review leads to the possibility that articles examining the topics of interest were missed. As such, the current review may not fully characterize the literature to date.

A challenge often presents when consolidating evidence from current bodies of literature is heterogeneity among the articles examined. This review of the literature resulted in diverse patterns across studies on operational definitions and inclusion/exclusion criteria (e.g., how TBI was defined or assessed), variables used to represent constructs of interest (e.g., the use of varying subjective or objective sleep parameters to measure the construct of ‘poor sleep’), and the diversity of assessment tools used for measurement (e.g., semi-structured interviews, PSG, self-report scales). This heterogeneity becomes especially difficult to interpret when research lacks universally accepted terminology, operational definitions, and measurement tools for the constructs being examined. For example, differences in criteria used to identify the presence or absence of a TBI may create unintentional baseline differences between groups across studies that have a direct influence on outcomes independent of the variables measured. For example, two independent studies that established neuropsychological differences between military veterans with and with mTBI, but used different assessment batteries, makes it challenging for researchers to feel confident about the true effect. It is important to note that the issue of research methodological heterogeneity is not a challenge specific to military research, but in fact a widespread challenge the greater field of research is currently facing; and therefore, something researchers should be cognizant of when conducting their own studies. We advise that future research move towards implementing a more standardized methodology, which should include multi-modal assessment of sleep disturbances. Improved and comprehensive assessment methods will lay a strong and much needed foundation for more optimal selection of interventions and their potential for success. Finally, this review represents a bias towards articles that were published only in English speaking, peer-reviewed journals.