Traumatic brain injury and sleep in military and veteran populations: A literature review

2.1 TBI pathophysiology

As brain tissue is acutely disrupted through blast and/or impact injuries, axon membranes are damaged through shearing forces, leading to potassium flooding the extracellular space and depolarization. This “impact depolarization”, causes the release of excitatory amino acids and neurotransmitters (Gruenbaum et al., 2022). Specifically, release of glutamate from impact depolarization leads to an excess concentration of extracellular glutamate (Gruenbaum et al., 2022). Elevated glutamate causes inappropriate activation of ligand-gated ion channels such as the NMDA and AMPA receptors, resulting in the rapid influx of calcium and sodium and efflux of potassium (Jamjoom et al., 2021). Disruptions to ion homeostasis leads to neuronal death leading to extravasation of iron, triggering ferroptosis and protein-rich fluid with free fibrin, thrombin, and plasmin, stimulating inflammation and reactive oxygen species formation (Geng et al., 2021; Sulhan et al., 2020). Cellular hypoxia of non-ruptured but damaged neurons follows, shifting to glycolytic metabolism with lactic acid accumulation. The resulting acidosis damages the blood brain barrier via vasogenic edema and a local inflammatory response (Hay et al., 2015). The dysfunctional ion regulation also causes excess intra-axonal calcium concentration, which activates calpain, a protease, to proteolyze numerous cytoskeletal proteins and lead to permanent axonal damage (Saatman et al., 2008).

The damage caused immediately from a primary injury can often prolong and progress into secondary injuries, which are due to biochemical events that occur hours to days after the initial injury (Mira et al., 2021). The primary culprit of secondary injury is prolonged neuroinflammation, the innate immune response to clear damaged neurons and glial cells from the site of injury (Mira et al., 2021). As astrocytes proliferate and hypertrophy to respond to the injury, an astroglia scar surrounding the damaged neural tissue forms; however, the scar contributes to excitotoxicity, and epileptiform activity may ensue (Shandra et al., 2019). Increasing evidence in animals and humans suggests that TBI also promotes dysfunction of the glymphatic system, possibly related to aquaporin-4 (AQP4) expression mislocalization from the end feet of astrocytes to the cell body (Ferrara et al., 2022). Furthermore, TBI is associated with autonomic dysfunction, resulting in relatively elevated sympathetic activity (Park et al., 2017; Ulmer et al., 2018), another potential contributor to glymphatic dysfunction in addition to autonomic symptoms such as blurred vision, palpitations, diaphoresis, orthostasis and dizziness.

2.2 Mechanisms of sleep disruption

Sleep disturbances result from immediate and delayed damage during and after TBI (Kaleyias & Kothare, 2022). Immediately after injury, neuronal networks are weakened or completely disrupted as neurons are sheared and destroyed (Kaleyias & Kothare, 2022). In severe TBI, as time progresses, additional injury comes from increased intracranial pressure, hypoxemia, hypotension, seizures, hematomas, and the cellular excitotoxicity (Kaleyias & Kothare, 2022). A number of potential sleep master regulator circuits could be compromised, leading to post-TBI sleep symptoms, such as damage to non-rapid eye movement (NREM) sleep coordinating gamma amino butyric acid (GABA)-ergic projections from the ventrolateral preoptic nucleus (VLPO), which could make falling asleep difficult. The hypothalamic-pituitary-adrenal (HPA) axis and the central orexinergic/hypocretinergic are two interconnected pathways that affect sleep and are damaged in TBI (Baumann et al., 2005).

The (HPA) axis’ ability to revert to homeostasis after heightened response from a TBI is impaired (Tapp et al., 2020). The neuroendocrine imbalance, specifically an increase in cortisol, may increase inflammatory cytokines including IL-1, IL-6, and tumor necrosis factor-alpha, and therefore worsening both TBI and sleep disruption outcomes. Whether mild TBI is a risk factor for neuroendocrine dysfunction remains a matter for further research.

The central orexinergic/hypocretinergic system promotes arousal through the effects of orexins, which are excitatory neuropeptides produced in the lateral hypothalamus. Orexigenic neurons project to the limbic system, thalamus, spinal cord, cortex, brainstem, and hypothalamic nuclei to excite cholinergic and monoaminergic wake-promoting systems, and they can be disrupted by TBI (Kousi et al., 2021). Studies in moderate and severe TBI have shown decreased orexinergic neuron content and orexin levels after TBI, but this may be due to the commonality of damaging the hypothalamus during TBI versus discrete disruption of the orexinergic/hypocretinergic system (Baumann et al., 2007a). This loss of orexinergic function could, in theory, trigger excessive daytime sleepiness and even obstructive sleep apnea (Sithirungson et al., 2023), common after TBI. Orexins, specifically orexin-A, have been found to increase ACTH levels through the activation of both corticotropin releasing factor and arginine vasopressin mRNA expression in the parvocellular cells of the paraventricular nucleus (Al-Barazanji et al., 2001). Such association between the HPA and orexinergic system further suggests the effects of decreased arousal and wakefulness through the dysfunction of both pathways after TBI.

However, when Baumann and colleagues followed TBI patients for six months, they showed that most patients’ orexin levels had returned to normal (Baumann et al., 2007b). Furthermore, orexin deficiency is unlikely to account for manifestations of insomnia, which may be more common than hypersomnia in the mild TBI population, suggesting damage may be related to damage in other pathways of sleep-wake and circadian regulatory control (Werner & Baumann, 2017).

Insomnia, importantly, can be a manifestation of a circadian rhythm disorder, where the circadian drive for alertness is poorly aligned (or perhaps diminished in signal amplitude) with the homeostatic drive to sleep (Borbely, 1982). In one small study of 42 TBI patients with insomnia, Ayalon and colleagues found that 36% of patients with insomnia were likely symptomatic secondary to an unrecognized circadian rhythm disorder (Ayalon et al., 2007). Furthermore, multiple small studies have shown irregularities in the melatonin secretion patterns in TBI patients, both in timing of the dim light melatonin onset (DLMO) and the amplitude of the melatonin peak (Bell et al., 2023; N. A. Grima, Ponsford, St. Hilaire, et al., 2016; Shekleton et al., 2010a), and unlike in other populations, melatonin therapy for insomnia has yielded favorable results in several clinical trials (Cassimatis et al., 2023),

Another potential explanation for post TBI insomnia could include the relative hyperactivity in the sympathetic nervous system as has been observed with baroreflex, arterial stiffness, heart rate variability, and catecholamine studies (Clifton et al., 1981; Hilz et al., 2015; B. D. Johnson et al., 2018; Woolf et al., 1987). Sleep should be associated with periods of reduced noradrenergic activity in the locus coeruleus during non-REM and near deactivation during REM (Léger et al., 2009; Swift et al., 2018a; Verret et al., 2006). Abnormal elevations in noradrenergic activity could prevent several important processes during sleep - contributing to disorders such as insomnia. However, normative data during the sleep cycles is lacking and is a major gap in our understanding of sleep physiology. Elevated noradrenergic activity is also reported in the PTSD population, a common comorbidity with TBI, and parsing the relative contributions from both disorders towards sleep dysfunction can be difficult. Many military members with combat-related PTSD also have suffered comorbid TBI, and at times the blast was the inciting event that also precipitated PTSD. The complicated interrelationship between these conditions can be a challenge for study design often resulting in studies of the two comorbid conditions together.

3.1 In the context of SWD, is the disruption with neural circuitry after TBI similar to the psychological stress associated with PTSD?

Individuals recovering from a TBI or PTSD often experience similar sleep-wake disorders (SWDs) such as insomnia or a REM-based sleep disorder (RBD) manifesting in excessive daytime sleepiness and pleiosomnia. Although the association between PTSD and TBI with clinical sleep disturbances has been well established (Ross et al., 1989; Wickwire et al., 2016; Yehuda et al., 2015), there lacks a consensus whether these conditions are two distinct SWDs and if they should be approached with different therapeutic strategies. Notably, PTSD symptoms strongly correlate with measures of sleep quality, making it difficult to separate their effects in comorbid populations (Gottshall, Agyemang, et al., 2022). In linear models, this situation can lead to issues with multicollinearity, limiting interpretation. Nonetheless, longitudinal studies examining sleep diagnoses that occur after the onset of TBI have successfully distinguished effects from comorbid psychiatric disorders (Haynes et al., 2021; Leng et al., 2021). It is likely that there is overlap between TBI and PTSD mechanisms of sleep disruption, and management of related SWDs requires biomarkers to elucidate and target underlying mechanisms. For example, as discussed earlier, elevated noradrenergic activity is a well-known mechanism of symptomology in the PTSD population (Naegeli et al., 2018; Pan et al., 2018; Wingenfeld et al., 2015). Addressing symptoms with noradrenergic blockade has been helpful in the PTSD population via stellate ganglion block (Rae Olmsted et al., 2020) and prazosin (Raskind et al., 2013), and clinical anecdotal evidence suggests it may have value in the TBI population for insomnia and limited evidence supports that it could help a number of parasomnias that involve arousals, such as the REM behavior disorder subset that some refer to as trauma associated sleep disorder, characterized by dream enactment (loss of REM atonia) and surges in sympathetic activity (Brock et al.,2022).

3.2 Neurotransmitter data

Since 2010, multiple studies have shown elevated norepinephrine (NE) levels in PTSD patients (Hendrickson et al., 2018; Wingenfeld et al., 2015) and hypothesize this could be responsible for their sleep symptoms (Vanderheyden et al., 2014; M. P. Walker & van der Helm, 2009). As the primary source of NE production within the CNS, increased activity in the locus coeruleus during sleep, particularly REM, is thought to be responsible for sleep disturbance within PTSD patients given their noradrenergic influence over wake and sleep promoting nuclei (Germain et al., 2013; Gimbel et al., 2023; Naegeli et al., 2018; Swift et al., 2018b; Wang et al., 2021).

The theory of a dysregulated functional connectivity continues to gain traction, and we and others are investigating whether modulating sleep- and arousal-promoting networks (e.g., basolateral amygdala, locus coeruleus) can correct the driving factor behind PTSD SWD sympathetic hyperactivity. Although there is considerable variability within the reported changes in sleep architecture in patients with PTSD, reports ranged from nonspecific fragmented REM sleep with a higher percentage of REM to increased N2 (C. Lewis et al., 2020; Mellman et al., 2014). Elucidating these findings in patients with TBI or PTSD and subsequent changes following treatment remains a target for future research.

3.3 Electroencephalogram features, polysomnography, and sleep architecture

The sleep electroencephalogram (EEG) data in TBI patients is widely variable and does not appear to support any clearly defined mechanism, however most studies are small, limited in their approach, and enroll varying severity and timing of TBI. Some studies reveal a decreased density of spontaneous K-complexes with fewer evoked K-complexes and decreased delta power during non-rapid eye-movement (NREM) sleep (A. Cote et al., 2015; Rao et al., 2011) while others report no differences in NREM or rapid eye-movement (REM) sleep when compared to healthy, non-TBI patients (Gosselin et al., 2009; Williams et al., 2008).

Several studies have demonstrated that patients with TBI experience increased sleep latency, increased wake after sleep onset (WASO) and decreased sleep efficiency (A Cote et al., 2015; Modarres et al., 2016; Rao et al., 2011; Shekleton et al., 2010b). Although the underlying neuropathological changes contributing to these disturbances remains hypothetical, various case reports have identified structural factors potentially responsible for TBI SWD, including direct damage to arousal-promoting neurons, inflammation in the cortex or subcortical structures and alterations in neural signaling (Chamoun et al., 2010; DelRosso et al., 2014; Valko et al., 2016; Yassin et al., 2015). Polysomnography (PSG) data from patients with PTSD also revealed increased sleep latency, increased WASO, and decreased sleep efficiency (Ansbjerg et al., 2023; Breslau et al., 2004; Mellman et al., 2014; Straus et al., 2015). However, patients without comorbid TBI do not necessarily have an instigating structural injury and cytotoxic event resulting in direct physical damage to the cortex or subcortical structures, rather, their impairment is likely related to stress-induced network remodeling detrimental to sleep regulatory and inhibitory mechanisms (Ansbjerg et al., 2023; Mellman et al., 2014; Straus et al., 2015).

3.4 Challenges in treating TBI SWD and PTSD SWD

Isolating TBI SWD from PTSD SWD when they are both present in an individual’s history remains challenging. A 2019 meta-analysis of 31 low-risk bias studies reporting a pooled prevalence percentage of 15.6% of patients developing PTSD following moderate or severe TBI (Van Praag et al., 2019). Data from this meta-analysis varied widely (2–36%), which further demonstrates the additional challenge of accurately diagnosing a patient with PTSD in the setting of TBI (Carlson et al., 2011b; Van Praag et al., 2019). The lines are further blurred when considering most of the current research on TBI and sleep is centered on combat veterans and active-duty warfighters suffering from concurrent TBI and PTSD or TBI with undiagnosed PTSD (Cooper et al., 2015; Elliott et al., 2020b; Hoge et al., 2008; Karr et al., 2014; Pietrzak et al., 2009; Trudeau et al., 1998; Warden, 2006). Whether the separation of pathophysiological contributions from PTSD and TBI is clinically relevant remains unclear, however, an understanding of underlying pathophysiology may inform the development of novel therapeutic targets.

Therefore, clinicians should have a high suspicion for multiple contributing factors and pursue sleep expertise to examine the pathophysiology carefully with a detailed sleep history, actigraphy and sleep diary monitoring, and polysomnogram. Successful management will likely require the employment of multiple strategies personalized to each patient and their suspected underlying contributing mechanisms (e.g., poor sleep hygiene, untreated OSA, PTSD, sympathetic hyperactivity, circadian dysfunction, motor contributions such as restless legs/PLMD. There are also significant challenges in patients with PTSD that impact response to therapy, such as non-compliance (Haynes et al., 2022; Lettieri et al., 2016). Military members with PTSD face the added challenge of frequent work-related travel and changing duty station to locations that may not have comparable resources and expertise. Comprehensive multidisciplinary care is probably the most effective solution for delivering personalized care in this population. Unfortunately, such an approach is time consuming, widely variable across clinical practices, and challenging to maintain over an extended period.

Recent and ongoing animal and human studies with promising data investigating novel neuromodulation techniques (e.g., acoustic or electrical stimulation, focused ultrasound) could provide the first TBI- or PTSD-specific interventions, but these approaches require additional studies prior to general acceptance in patient care paradigms (Huang et al., 2022; Kim et al., 2022; Tegeler et al., 2020).

3.5 What knowledge gaps need to be addressed with future research?

Although sleep disruption after TBI or PTSD has been well documented, the circumstances surrounding why certain patients develop SWD after varying severity of TBI and the lack of targeted treatment for PTSD SWD vs TBI SWD remain unclear. One avenue to address the complicated relationship between PTSD and TBI includes future studies investigating comparisons in the prevalence, subjective sleep measures, and objective physiological sleep measures in individuals with TBI SWD without evidence of psychological trauma (e.g.., concussions in athletes) often present in patients exposed to combat or motor vehicle collisions (Gosselin et al., 2009). Such data could be invaluable in identifying clinically relevant differences between PTSD SWD and TBI SWDs.

5.1 TBI and glymphatics

Iliff and colleagues showed in the rodent model that glymphatic activity is directly and indirectly impaired after TBI, causing an accumulation of phosphorylated tau, a biomarker and possible mediator of neuronal injury and neurodegeneration (Iliff et al., 2014). TBI was also shown to disrupt the meningeal lymphatic networks, which can have a role in accumulation of these putative toxins (Bolte et al., 2020). Neuronal injury biomarkers such as GFAP, neurofilament light and neuron-specific enolase (NSE), have been detected in human plasma in the acute and chronic setting after TBI. The levels of these biomarkers appear modulated by sleep deprivation and glymphatic drainage (Holth et al., 2019; Ju et al., 2017; Ooms et al., 2014; Xie et al., 2013). Furthermore, cross-sectional measurements of glymphatic exchange in humans show that plasma levels of several biomarkers of neurodegeneration and neuroinflammation correlate with measured glymphatic clearance, raising the possibility that peripheral blood can inform glymphatic activity, at least in part (Eide et al., 2023). Supporting research has shown that the same plasma biomarkers known to correlate with glymphatic activity also can predict sleep quality in the TBI population (Werner et al., 2021). Whether glymphatic biomarkers will inform the prognostication or management of TBI is not yet clear.

Glymphatic dysfunction after TBI could be secondary to disrupted blood-brain barrier (BBB), alterations in AQP4 localization, cerebral blood flow and neuronal activity, and further complicated by edema and inflammatory responses (Plog et al., 2015). Biomarkers of neuroinflammation and brain injury such as GFAP and UCH-L1 are elevated immediately after a TBI and are predictive of the appearance of blood on head imaging, which was previously attributed to simple BBB disruption. However, sleep deprivation after TBI in rodents has been shown to suppress or eliminate the expected serum increases of GFAP, S100 calcium binding protein B (S100B), and NSE, suggesting that a more complicated, sleep-regulated process is involved (Benveniste et al., 2017; Plog et al., 2015). Furthermore, sleep deprivation after TBI promotes increased AQP4 mislocalization (Tapp et al., 2020), suggesting that the sleep process itself is important for glymphatics maintenance and repair. Thus, after TBI, decreased glymphatic clearance of brain proteins such as GFAP or tau could theoretically result in lower plasma/serum concentrations after TBI, however this may be offset by increased production triggered by TBI and secondary inflammation and nonspecific diffusion across the BBB into the bloodstream.

It is further suggested that the glymphatic system through the AQP4 channel allows for adequate elimination of cellular wastes, which can malfunction in situations of TBI or other neurologic disorders (Benveniste et al., 2017). In AQP4 knockout mouse models, the clearance of cellular waste is decreased, showing accumulation of AB42 (Iliff et al., 2012). After TBI, there is a loss of perivascular AQP4. GFAP, S100B, tau, and NSE levels in these rodents are unexpectedly decreased in the serum after TBI (Piantino et al., 2022; Plog et al., 2015). These findings further contribute to the hypothesis that post-traumatic impairments of the glymphatic system, specifically through changes of the AQP4 channels, contribute to the decreased clearance of metabolites from the brain.

An additional mechanism of glymphatic impairment after TBI worth considering in future study is the inhibitory effect of sympathetic hyperactivity. This may explain data from Iliff et al showing that the glymphatic activity contralateral to the injury became reduced within 3 days of injury. Recent evidence suggests that mitigating sympathetic activity with alpha blockade can restore glymphatic activity after TBI (Hussain et al., 2023).

5.2 Measuring the glymphatic system and future directions

Measurement of glymphatic biomarkers is an ongoing challenge for several reasons. In order to provide clear evidence that glymphatic function is altered after TBI, definitive measures of glymphatic activity are needed. Currently this has only been achieved by invasive measures when Eide and colleagues injected gadolinium contrast to the intrathecal space, using dynamic contrast-enhanced (DCE)-MRI to track the glymphatic CSF-ISF exchange(Eide et al., 2021; Eide & Ringstad, 2015). While this method validated the existence of glymphatic activity and its link to sleep in humans, it is too invasive to routinely employ to study glymphatic activity. Noninvasive measures are under development, but more work is needed to validate them (Richmond et al., 2023).

Although the direct measurement of glymphatic activity remains challenging, there are several potential alternative approaches to glymphatic system measurement in humans that have been proposed (Piantino et al., 2022). Non-invasive and contrast free measurements of the glymphatic system such as trans-endothelial water exchange, perivascular diffusion, and MRI encephalography have been used to better understand glymphatic biology, but they have not been validated against the gold standard of DCE MRI (Hennig et al., 2021; Piantino et al., 2022). Notably, there is growing interest in measuring a proposed surrogate for glymphatic function non-invasively using DTI along the perivascular space (DTI-ALPS) (Taoka et al., 2017). The DTI-ALPS index assesses the movement of water molecules along the perivascular space through a specialized use of diffusion tensor imaging. Deficits in DTI-ALPS in patients with a history of TBI were shown compared to controls, and, interestingly, the measures correlate with the plasma biomarker, neurofilament light (Dai et al., 2023; J. H. Park et al., 2023; Butler et al., 2023). Further, DTI-ALPS has been found to be reduced in various sleep disorders, including OSA, insomnia, REM sleep disorder, narcolepsy, and restless leg syndrome (H.-J. Lee et al., 2022; Siow et al., 2022; D. A. Lee et al., 2022; Gumeler et al., 2023; K. M. Park et al., 2023). These findings are supportive of the link between DTI-ALPS and glymphatic function. Replication of this work with more investigators and cross-correlation with relevant glymphatic biomarkers will increase confidence in its validity.

7.1 Insomnia

A number of studies indicate that post-TBI patients suffer from increased insomnia compared to the general population. A meta-analysis found that 50% of individuals post-TBI experienced some form of disordered sleep, with 25% to 29% being formally diagnosed with sleep disorders, including insomnia (Paredes et al., 2021). A recent study used latent class models to identify five trajectories of insomnia over 12 months after TBI, using data from the prospective TRACK-TBI study. It found that the majority of participants had ongoing insomnia after 12 months, with greater risk associated with substance use and psychiatric comorbidities, and prior sleep disorders (Wickwire et al., 2022).

First-line therapy for insomnia is anchored in behavioral strategies, including cognitive behavioral therapy for insomnia (CBT-i), when feasible face-to-face (virtual or in person), which has been shown to be effective in TBI related insomnia(Ludwig et al., 2020; Ouellet & Morin, 2007). However, many hospitals do not have access to CBT-i or specific expertise for the application of CBT-i to military and veteran populations. CBT-i requires close follow up, and logistical considerations can be limiting. Moreover, many military veterans with combat related TBI and insomnia suffer from short sleep duration, which may represent a phenotype of insomnia with heightened arousal levels that may be less responsive CBT-i. The efficacy of CBT-i in short sleepers is unclear (Bastien et al., 2017). When providers are not available, self-paced CBT-i can also be employed online with free websites and application made available by the VA such as “Insomnia Coach” (Kuhn et al., 2022).

There are several options for the pharmacologic management of post-TBI insomnia. Comorbid pain in TBI patients has led to the frequent use of nightly gabapentin to also treat insomnia. Gabapentin has been shown to enhance deep sleep (Foldvary-Schaefer et al., 2002), but evidence for insomnia in the TBI population is lacking. Clinicians should use caution, because it also has been shown to cause obstructive sleep apnea by unclear mechanisms (Piovezan et al., 2017). Sedating antidepressants such as trazodone and quetiapine are frequently employed, but evidence is also lacking for these in the TBI population. One prospective observational cohort study evaluated psychotropic medication administration patterns for the inpatient rehabilitation after TBI, citing that hypnotic agents were prescribed for 30% of the patients (Hammond et al., 2015). Among the hypnotic agents, the most frequently prescribed sleep-inducing drugs were nonbenzodiazepine GABA-A agonists, with zolpidem being the top-prescribed drug, followed by eszopiclone and zaleplon. The nonbenzodiazepine GABA-A agonists can be used for both sleep-onset and sleep-maintenance insomnia in the general population, with minimal residual sedation in the following morning.

Melatonin has been another intervention of interest. In 2018, Grima et al. presented a randomized double-blind, placebo-controlled crossover study that compared treatment with melatonin and placebo in outpatients with mild to severe TBI reporting sleep disturbances post-injury. The study found that melatonin improved subjective sleep quality, actigraphic sleep efficiency, reduced self-reported anxiety symptomatology and fatigue, and enhanced self-perceived vitality and mental functioning in patients with TBI and insomnia (N. A. Grima et al., 2018). Although not recommended by the American Association of Sleep Medicine for the general insomnia population, melatonin was favored in a recent meta-analysis across 9 clinical trials in the TBI population (Cassimatis et al., 2023).

Ramelteon is a melatonin receptor agonist, and a pilot study performed in 2015 found that compared to a placebo there was a significant increase in total sleep time with a small increase in sleep latency when it was taken nightly for three weeks (Kemp et al., 2004).

7.2 Hypersomnias

CSF orexin levels were low in 95% of patients within the first 4 days of moderate to severe TBI (Baumann et al., 2005), and a mouse model indicated these deficits was associated with an inability to maintain wakefulness (Lim et al., 2013). Leng and colleagues followed diagnosis codes after TBI in 200,000 veterans, identifying an increased risk for narcolepsy in the TBI population (Leng et al., 2021). In managing narcolepsy following a TBI, the approach is similar to that for non-TBI patients. It involves three key elements that are often employed prior to using stimulants: improving sleep habits, using caffeine strategically, and scheduling naps. When dealing with narcolepsy that does not respond well to these measures, “wake-promoting” and stimulant medications can be considered, which are often monoamine reuptake inhibitors. The primary choice for wake-promoting agents is modafinil as it has fewer cardiovascular side effects and less abuse potential compared to stimulants. Treatment of modafinil in TBI-related sleep disorders has shown mixed results. Patients with TBI treated with Modafinil of 100 mg to 200 mg every morning reported improvement in excessive sleepiness scale (ESS) and maintenance of wakefulness test scores compared to placebo, although it did not improve posttraumatic fatigue (Kaiser et al., 2010). However, another trial comparing patients taking 400 mg Modafinil to placebo found an improvement of the ESS at week 4 but not at week 10 (Jha et al., 2008). Similarly, treatment with armodafinil showed significant improvement in sleep latency (Menn et al., 2014).

Clinical practice guidelines from the American Academy of Sleep Medicine recommend use of pitolisant and solriamfetol to maintain wakefulness, as well as sodium oxybate to improve sleep consolidation, for patients with narcolepsy. However, it is uncertain how this may apply to TBI induced hypersomnia and important to realize that many of these medications cannot be prescribed unless the patient meets formal diagnostic criteria for narcolepsy (due to FDA indications and/or insurance coverage limitations). Focused trials in the TBI population for hypersomnia management are needed.

7.3 Post traumatic stress disorder

Particularly amongst patients with PTSD, those with TBI were more likely to have lower verbal memory function, and this portended a weaker therapeutic response to cognitive behavioral therapy as a sleep intervention (Scott et al., 2017). Patients with mTBI and PTSD reported immediate improvement in the frequency and intensity of headaches, nightmares, and sleep disturbances with the use of a removable mandibular neuroprosthesis/splint, presumably addressing bruxism, and effects did not diminish over time (Moeller et al., 2014). These results were also found in treatment of patients of PTSD without TBI. The management of PTSD involves a combination of pharmacotherapy and psychotherapies. Trauma-focused psychotherapy is the preferred initial treatment. In the realm of pharmacotherapies for the general population, options include SSRI/SNRIs, with sertraline and paroxetine having an FDA indication for PTSD. Extensive clinical experience and limited randomized clinical trial data show that clonidine and prazosin can be employed to reduce trauma-related nightmares in the TBI and PTSD population (Mysliwiec et al., 2022; Raskind et al., 2007). These studies and our current clinical practice lack the use of reliable objective measures of the target, noradrenergic activity. In the absence of useful biomarkers of noradrenergic/sympathetic activity to select participants and monitor responses to therapy, clinical trials will be at the risk of failing due to a poor signal-to-noise ratio – rather than failed efficacy. One possible example is the failure of the largest prazosin trial to date, where the population characteristics (e.g., average blood pressure, alcohol or substance use) suggested that it was a different population than that of previously successful smaller trials (Raskind et al., 2018). Most clinicians who use prazosin would argue it is an effective intervention, but the right patient must be chosen. Wearable technology that uses haptic vibrational feedback when heartrates spike during sleep were shown to have some promising results in nightmare management, but more data is needed – particularly in the TBI population (Davenport & Werner, 2023).

7.4 Obstructive sleep apnea

Sleep disordered breathing is significantly more common following TBI compared to survivors of other types of non-TBI combat trauma, and this risk persists long term (Haynes et al., 2022). OSA is associated with more severe PTSD following TBI, and these patients may struggle with worse CPAP adherence, have more severe sleep related symptoms, and lower response rates to treatment, all of which demand preventative and long-term management strategies (Lettieri et al., 2016). OSA and PTSD are common in the TBI population; one study found over 75% of participants with acute TBI had OSA, and the presence of OSA was associated with slower cognition, memory, and executive functioning (Steward et al., 2022). Longer duration of untreated OSA in participants with a history of hospitalization for TBI had an adverse impact on verbal memory (Silva et al., 2022). The management of OSA following TBI closely mirrors the approach used for sleep apnea in the general population, including the use of positive airway pressure. Other options include mandibular advancement devices, removal of offending medications (e.g., gabapentin), surgical interventions, hypoglossal nerve stimulation devices, and, in rare occasions, positional therapy can be effective.

Because patients with OSA due to TBI and PTSD represent a vulnerable population and may suffer from poor CPAP adherence and comorbid insomnia, we recommend multidisciplinary interventions targeting educational, behavioral, and supportive interventions to improve PAP adherence. Recent changes in the literature include more specificity for weight loss guidance and advances in surgical therapies. For most patients with obesity, it is important to counsel patients that weight loss is a valuable adjunct treatment for OSA, but that significant weight loss, between 10–25% of body weight, is required for efficacy. Expert consultation with obesity medicine specialists to consider bariatric surgery, medically supervised diets, and weight loss medications using a personalized approach may be warranted (Chang et al., 2023; St-Onge & Tasali, 2021).

Surgical intervention for OSA is an alternative to CPAP for some patients. In the past, uvulopalatopharyngoplasty was the preferred approach, but advances in surgery have led to a more individualized approach, based on recognition that the mechanism of OSA and level of airway collapse differs between patients. Surgery is divided into anatomic treatments and neurostimulatory. Anatomic surgeries include maxillomandibular advancement, combinations of surgeries involving uvulopalatopharyngoplasty and palatal interventions, sometimes as staged surgeries. Neurostimulatory treatment for OSA is performed using the hypoglossal nerve stimulator, which has shown to demonstrate improvement in severity of obstructive sleep apnea and daytime sleepiness (Kent et al.,2019).

7.5 Circadian rhythm disorder

It has been demonstrated that TBI induces abnormalities in gene expression and melatonin production, affecting circadian rhythm, however the mechanisms and indications of these changes are uncertain (Zhanfeng et al., 2019). Large-scale TBI studies of circadian function are lacking. Melatonin remains the most studied treatment for CRD in the TBI population, and there is strong evidence for subjective and objective sleep quality improvement (Cassimatis et al., 2023), but the specific mechanisms of this appear to be lacking. A recent metanalysis demonstrated that melatonin had a positive effect on multiple neuropsychiatric outcomes, but this was not limited to only circadian rhythm disorders (Barlow et al., 2019). Daily morning blue light therapy was also associated with improved daytime sleepiness, as well as increased gray matter volume and functional connectivity within networks related to sleep regulation and daytime cognition (Raikes et al., 2021). Although melatonin and light therapy demonstrate improvement in circadian disruption after TBI, studies elucidating the specific mechanism of improvement or isolating improvement specifically in circadian rhythm disorders appear to be lacking and can serve as a focus of further clinical and benchtop research.