Discriminating military and civilian traumatic brain injuries

Highlights

• Traumatic brain injury occurs at higher rates among service members than civilians.

• Explosions are the leading cause of TBI in the military.

• Postcuncussion, PTSD and chronic pain symptoms are very common in military TBI.

• PTSD develops after mild TBI more frequently among service members than civilians.

• Development of CTE after multiple blasts is not yet supported by empirical evidence.

Abstract

Traumatic brain injury (TBI) occurs at higher rates among service members than civilians. Explosions from improvised explosive devices and mines are the leading cause of TBI in the military. As such, TBI is frequently accompanied by other injuries, which makes its diagnosis and treatment difficult. In addition to postconcussion symptoms, those who sustain a TBI commonly report chronic pain and posttraumatic stress symptoms. This combination of symptoms is so typical they have been referred to as the “polytrauma clinical triad” among injured service members. We explore whether these symptoms discriminate civilian occurrences of TBI from those of service members, as well as the possibility that repeated blast exposure contributes to the development of chronic traumatic encephalopathy (CTE). This article is part of a Special Issue entitled ‘Traumatic Brain Injury’.

Introduction

Advances in technology have resulted in more effective vehicle and body armor, allowing the modern combat environment to afford current service members more protection than any other time in history. Similarly, advancements in acute trauma treatment in theater have increased survivability from otherwise fatal wounds, meaning more soldiers with a TBI are surviving other injuries (also responsible for their TBI) that would have been fatal in the past. This has resulted in a dramatic increase in the occurrence of reported traumatic brain injuries (TBI), so much so that mild TBI (mTBI) has been referred to as the “signature injury” (Snell and Halter, 2010) of Operations Enduring Freedom (OEF) and Iraqi Freedom (OIF). The Department of Defense and Veteran's Affairs Offices classify TBI into one of three severities (mild, moderate or severe) depending on the presence and duration of three factors, i.e. loss of consciousness (LOC), posttraumatic amnesia (PTA, not remembering the injury event or a period of time before or after the event) and alteration of consciousness (AOC, e.g. feeling dizzy, dazed or confused). Mild TBI is diagnosed when any of these factors are experienced due to some external event. If LOC exceeds 30 min, PTA exceeds 24 h or AOC lasts more than 24 h the severity of TBI increases to moderate. A moderate TBI is also diagnosed when abnormalities are present in any neuroimaging modality (computed tomography, magnetic resonance, etc.). A severe TBI is diagnosed when LOC exceeds 24 h or PTA exceeds 6 days, with or without neuroimaging abnormalities (Casscells, 2007).

According to the Department of Defense, in 2013 there were 27,324 new diagnoses of TBI among service members and over 300,000 since the year 2000, with most (82.4%) being of mild severity (DVBIC, 2014). While approximately 80–85% of those who sustain mTBI fully recover within 1 month (Belanger et al., 2005, Belanger and Vanderploeg, 2005, Schretlen and Shapiro, 2003), the remainder show persistent symptoms that may last several weeks, several years or may not resolve within their lifetime. According to these values over 8000 service members currently suffer from persistent symptoms of mTBI or moderate/severe TBI that occurred in 2013 alone. Effective treatment for, and diagnosis of, TBI is of paramount importance for patients and society due to the detrimental impact on physical and psychological quality of life and enormous impact on economic, social and medical resources.

Explosions from improvised explosive devices and mines cause 50–79% (Hoge et al., 2008, Terrio et al., 2009) of deployment related TBIs in the military. The frequency with which service members experience blasts makes it difficult to accurately gauge the occurrence of TBIs, especially mTBIs. Many soldiers do not seek treatment after being briefly dazed or knocked unconscious, which by current definitions would constitute a mild TBI. Therefore, it is likely that the prevalence of TBI in the military is higher than reported and may be as high as 59% (Hoge et al., 2008, Risdall and Menon, 2011).

In most instances, explosions distinguish deployment-related military TBI from civilian TBI occurrences. Injuries due to blast exposure may result from the initial overpressurization (shock) wave caused by an explosion (i.e. primary blast), from secondary (e.g. shrapnel/debris), tertiary (e.g. person thrown into nearby object), quaternary (e.g. burns, smoke/chemical inhalation) or even quiniary (e.g. hyperpyrexia, Moore and Jaffee, 2010) blast mechanisms. Primary blast is ubiquitous in these instances regardless of whether more critical or complex wounds result from secondary–quiniary blast mechanisms. Though primary blast exposure is common among service members little is known about how this force affects the brain. Modeling of primary blast effects in humans has necessarily been based on the results of animal research. Whether animal findings are representative of actual blast exposure effects in humans, though, remains debatable. The most obvious confounding factor is that the human body, head and brain are not similar in size or shape to that of research animals (i.e. rats, pigs, etc.…). As a result, different methods of scaling the primary blast effects seen in animal research must be applied to computational or physical models of the human body. Unfortunately, there is no consensus as to which model or scale is best. Jean et al. (2014) recently developed a physics-based scaling method of primary blast exposure, which takes into account the physical properties of the human skull and soft tissue surrounding it. Their findings suggest that humans are more susceptible to blast brain injuries than smaller mammals, which is counter to most related research findings (Bass et al., 2012, Rafaels et al., 2011, Zhu et al., 2013). That a different modeling approach can result in findings the opposite of well supported past findings underscores the uncertainty as to which method is appropriate for translating the effects of primary blast from animals to humans. As a consequence there is no empirical evidence with which to determine minimum thresholds of blast intensity required to produce damage to the human brain.

Empirical effects of primary blast on the brain are available, however, for animals. This evidence reveals many immediate cellular, molecular and biochemical neuropathological effects after blast exposure. Some of the more pronounced effects are neuronal and glial cell damage (Ahmed et al., 2012), compromised blood brain barrier (Abdul-Muneer et al., 2013), inflammation and elevated cytokines and chemokines (Zou et al., 2013) to mention a few. For an extensive summary of this literature see Kobeissy et al. (2013). Given the similarity of brain tissue across mammals these cellular and molecular findings arguably apply to humans to a large degree and most likely contribute to the heterogeneous symptomatology associated with blast-related TBI, though, as mentioned above, direct translation of these studies' results to humans is controversial.

Numerous studies have shown an association between blast-related (and non-blast-related) TBI and deficits in physiological and cognitive abilities (Norris et al., 2014, Scheibel et al., 2012). See Bogdanova and Verfaellie (2012) for review. Whether blast-related TBI differs from non-blast-related TBI is currently a debatable topic. This question has been tested recently with mixed results. According to several studies, differences do not exist between blast and non-blast-related TBI on measures of postconcussion symptom endorsement and various neuropsychological and cognitive measures (Cooper et al., 2012, Luethcke et al., 2011, Neipert et al., 2014, Norris et al., 2014). Differences between blast and non-blast-related TBI have been demonstrated, however, on measures of white matter integrity (Taber et al., 2014) and neural activation during response inhibition (Fischer et al., 2014) using magnetic resonance imaging. Neuroimaging measures may be more sensitive than behavioral measures for revealing potential group differences due to their ability to detect small alterations in brain tissue. Only additional research will determine whether these differences are consistent and replicable or confer diagnostic, prognostic or therapeutic utility.

Determining whether or not blast effects are dissociable from more traditional mechanical causes of TBI is challenging due to the low frequency with which primary blast mechanisms alone result in TBI. Secondary–quiniary blast injuries occur in most blast-related TBIs, confounding the effects of primary blast. Even so, blast exposure is not a trivial piece of information. Recent evidence suggests that repeated blast exposure might have a cumulative effect on post-concussion symptom endorsement in service members who sustain TBI (Kontos et al., 2013, Reid et al., 2014), making the relative number of blasts experienced by a service member a potentially useful clinical, if not prognostic, piece of information.

The debilitating physical and psychological symptoms of combat exposure have worn several monikers throughout American history; “soldier's heart” after the Civil War, “shell shock” after World War I, “battle fatigue” after WWII, “post-Viet Nam syndrome” and “posttraumatic stress disorder” (PTSD) today. The pervasiveness of these terms and the myriad symptoms associated with them underscores the fact that deployment related military TBI does not occur in isolation. Service members are exposed to the physical and psychological hazards associated with combat in addition to those associated with military training and normal recreational activities. This susceptibility to multiple modes of injury is apparent when considering the spectrum of symptoms endorsed by injured service members. Three such symptoms (postconcussion, posttraumatic stress and chronic pain) are so frequently observed by clinicians their combination has come to be known as the ‘polytrauma clinical triad’ (Lew et al., 2009). A Veteran's Affairs polytrauma study found that 97% of their sample (n = 62) had postconcussion symptoms, 97% complained of chronic pain and 71% qualified for a diagnosis of PTSD (Lew et al., 2007). A similar study found, in a sample of 50 OEF/OIF veterans, that 80% sustained a TBI, 96% reported a problem with pain and 44% experienced symptoms of PTSD (Clark et al., 2007). Lew et al. reported the presence of these three symptoms in 42.1% of 340 OIF/OEF veterans in 2009 (Fig. 1). Diagnosis and treatment of TBI in the military is challenging due to the frequent comorbid presence of PTSD and chronic pain syndrome. How to best treat an injured service member is further complicated by the fact that these three maladies share overlapping symptoms.

It is unknown whether the “polytrauma clinical triad” is as pervasive in instances of civilian TBI as this term is generally applied in military settings. Therefore, a comparison of PTSD and chronic pain symptoms between service members and civilians is presented in the next three sections. Although the prevalence of this triad is large, it is important to mention that other common TBI comorbidities exist, such as depression, anxiety and sleep and substance use disorders. In fact the heterogeneity of symptoms revolving around the occurrence of TBI is quite possibly the most challenging aspect of TBI research and its treatment. Hence, many clinicians are starting to adopt a strategy for treatment that is independent of diagnosis and rather focuses on treatment of the most salient symptoms present at any given time in order to reduce the overall burden of the patient (Brenner et al., 2009).

PTSD occurs when the experience of a traumatic event causes an individual to re-experience the event, avoid reminders of the event, produce negative thoughts after the event and become reactive in a way not displayed before the event. These symptoms must persist for more than one month and interfere with normal daily functioning and not be the result of medication, substance abuse or another illness (American Psychiatric Association, 2013). Research has shown that up to 43.9% of service members who lose consciousness during injury meet this diagnostic criteria for PTSD, compared to 16.2% with other injuries and 9.3% who are uninjured (Hoge et al., 2008). Additionally, service members who have sustained TBI endorse increased levels of PTSD symptoms (Hoge et al., 2008, Kennedy et al., 2010) and service members with PTSD endorse increased levels of postconcussion symptoms (Hoge et al., 2007, Schneiderman et al., 2008, Vanderploeg et al., 2009). Though not directly linked to TBI, the prevalence of PTSD in service members (30%) is much higher than that of demographically matched civilians (5%) (Koenen et al., 2002). And service members who have been injured show higher rates of PTSD than their non-injured counterparts (Koren et al., 2005, Pitman et al., 1989). Similarly, service members who sustain blast-related TBI report more PTSD symptoms than those who sustain a TBI from other mechanisms (Trudeau et al., 1998, Warden, 2006).

The above-cited evidence suggests comorbid PTSD in military-related TBI likely discriminates it from civilian TBI simply due to the lack of blast and combat exposure in the civilian environment. This cannot be said definitively though, as the documented prevalence of developing PTSD after sustaining TBI in the civilian population vary widely depending on the operationalization of variables and method of data accumulation. See Gill et al. (2014) for review. Civilian estimates that exist, however, range from 0 to 33% (Bryant and Harvey, 1999a, Bryant and Harvey, 1999b, Hoffman et al., 2012), compared to the 43.9% of service members mentioned above. While these values may not be perfectly representative of their respective populations, they arguably display a notable difference between the two groups. Additionally, a systematic review of PTSD development after TBI found evidence of greater prevalence rates of PTSD in military versus civilian samples (Carlson et al., 2011) after mTBI. When analyzing all TBI severity levels, however, little evidence to support a difference between civilians and service members was present. The authors attributed this lack of evidence not to an absence of real-world differences, but to an abundance of differences in study design and lack of standardized measures; shortcomings that are being addressed in TBI research globally. Finally, though not empirically valid, anecdotes present in the literature of civilian PTSD and TBI often mention the higher prevalence of PTSD in service members who sustain a TBI.

Chronic pain is operationalized in many ways across various studies. Much pain research employs a survey or questionnaire type measure, such as the Brief Pain Inventory or Visual Analog Scale or some other self-report scale in which individuals rate their level of pain on some continuum (1–10, none-extreme, grimacing-smiling face). Chronic is attributed to pain that lasts more than 3 (usual minimum), 6 or 12 (maximum) months or more simply, “pain that extends beyond the expected period of healing” (Turk and Melzack, 2010). Pain due to TBI may be perceived when damage to the somatosensory area of the brain responsible for a particular area of the body is damaged.

Like PTSD, the true prevalence of comorbid chronic pain and TBI is difficult to ascertain due to methodological differences between studies. One of the earlier and most widely cited civilian studies found that 95% of their mTBI sample (N = 104) reported chronic pain versus only 22% with moderate to severe TBI (Uomoto and Esselman, 1993). Another sample of civilians with moderate to severe TBI (N = 146) reported that 74% suffered from pain 1-year post injury (Hoffman et al., 2007), while another, reported 91% of their civilian mTBI sample (N = 53) experienced chronic pain an average of 12.9 months post injury (Alfano, 2006). These figures, which vary widely, are similar to many studies within the TBI/chronic pain literature. For a systematic review see Dobscha et al. (2009).

The same factors (study design, standardized categorization and measures, etc.) that make determining prevalence of comorbid civilian TBI and chronic pain difficult also apply to the military population. As such, studies within the military also report widely varying prevalence statistics. For instance, Bosco et al. (2013) provide a range of 30% to 90% prevalence of comorbid pain (mostly headache) and mTBI in a recent review. No prevalence figures are given for moderate/severe TBI. A recent Veteran's Affairs medical record review of Iraq and Afghanistan veterans revealed that 70% of those diagnosed with a TBI of any severity (N = 22,053) also received diagnoses of chronic head, neck or back pain (Taylor et al., 2012). Hoge et al. (2008) report that 34% of service members who lost consciousness (mTBI) due to injury claimed to have back or joint pain 3 to 4 months post injury.

The most applicable evidence with which to compare chronic pain prevalence in civilians and service members post TBI comes from Nampiaparampil (2008) who conducted a systematic review of 23 studies showing 52% of 3289 civilian patients with TBI have chronic pain, while 43% of 917 veterans with TBI suffer from chronic pain. Among the civilians in this review chronic pain was present more frequently in mTBI (73%) than moderate or severe TBI (32%).

According to this literature, there is not much evidence to support that chronic pain dissociates TBI in civilians from TBI in service members. If anything, evidence suggests the severity of TBI may be negatively associated with the presence of chronic pain. Though, if this is true, the lack of delineation between TBI severity levels in many chronic pain studies may obscure this finding as well as whether actual chronic pain differences exist between military and civilian occurrences of mild, moderate or severe TBI. Future pain/TBI studies will benefit from adding TBI severity measures to their data.

Persistent headache after TBI is the most common type of chronic pain reported by this population (Nicholson and Martelli, 2004). Between 30% and 90% of civilians who sustain mTBI report persistent headaches compared to 27% to 52% who suffer moderate/severe TBI (Couch and Bearss, 2001, Evans, 2004, Lahz and Bryant, 1996). Prevalence of persistent headache in service members following mTBI is reported to range between 17% and 63%, and 38% among those with moderate/severe TBI (Hoge et al., 2008, Walker et al., 2005). While chronic pain studies often include instances of headache within their data, headache measured independently may be a more robust indicator of the potential differences between civilian and military TBI. A recent review by Theeler et al. (2013) addressed this possibility, but did not find evidence of a difference in prevalence of posttraumatic headache (PTH, defined as those beginning within one week after TBI) between civilians and service members. They did report, however, a large difference in the proportion of civilians and service members with comorbid PTH and PTSD (30% and 80%, respectively). The increased prevalence of PTSD in military related PTH compared to civilian PTH is intriguing, especially when paired with the evidence of greater prevalence rates of PTSD in service members versus civilians after mTBI. These similar findings suggest that PTSD, but not chronic pain or headache, differentiates civilian and military instances of mTBI, though not moderate or severe TBI. Theeler et al. used PTH as their headache measure, but other studies operationalize headache in many other ways, such as frequency and duration of occurrence, specific symptoms experienced and length of time between injury and onset as well. So while differences are not supported in the previous example another method of headache operationalization may potentially reveal differences between service members and civilians with TBI.

CTE was first discovered in the early 20th century among boxers who sustained repeated brain trauma and was originally termed “dementia pugilistica” (Winterstein, 1937) and informally known as “punch drunk” (Martland, 1928). It is a neurodegenerative condition associated behaviorally with increased aggression and irritability, memory deficits and impaired cognition and emotional lability. Later stages of CTE involve motor and speech disturbances and dementia (McKee et al., 2013). A cure does not presently exist.

The current special issue has a focus on repetitive brain injury and CTE; however, very little research has been conducted on CTE in military samples. Regardless, a summary of the existing literature regarding service members, veterans and CTE follows.

The idea that repetitive non-injurious or mildly injurious impacts to the head can result in a degenerative brain disease that ultimately causes dementia and death is worrisome to say the least. Not surprisingly, CTE in sports has become a topic of intense debate and discussion and evidence is mounting that suggests CTE in sports is a real phenomenon. At present, the only evidence that CTE is distinct from other types of dementia is the cellular and topographical pattern of tau neurofibrillary tangle (NFT) and neurite deposition within the brain (Hof et al., 1992, McKee et al., 2013) and the only means of confirming a diagnosis of CTE is upon autopsy.

CTE is most closely associated with individuals who have sustained repetitive mTBIs while engaged in contact sports. Recently, however, 22 ex service members have received confirmed diagnoses of CTE (Goldstein et al., 2012, McKee et al., 2009, McKee et al., 2013, Omalu et al., 2011), prompting many to ask whether exposure to multiple explosive blasts and/or sustaining multiple blast-related mTBIs results in CTE. While this serious possibility needs to be thoroughly researched, at present, an association between CTE and blast exposure is only speculation. There are far too few cases that have been verified and very little research to support such a claim. Causality is still largely unknown in sports related CTE and the fact that 16 of the 22 soldiers diagnosed with CTE were also athletes warrants pause. While blast exposure is the leading cause of TBI in service members, blunt-force trauma is theorized to drive the onset/progression of CTE. The characteristic pattern of NFT distribution used to diagnose CTE is suggested to relate “to direct mechanical injury from blows to the side or top of the head, given their multifocal dorsolateral frontal and parietal, inferior frontal and occipital, and lateral distribution” (McKee et al., 2009). Of course, secondary and tertiary blast injuries accompany almost all blast-related TBIs, but the heterogeneity of associated impact locations preclude the establishment of a pattern of typical impacts, unlike those that occur while playing a given sport. Additionally, the unpredictable nature of combat likely prevents any series of injuries in a single individual from being similar enough to be considered repetitive or confer the necessary damage to identify an accumulation of common effects from those injuries. One injury will likely be very different from the next. However, it is not impossible that future research will reveal that repeated blast exposure results in predictable patterns of brain damage. Only time will tell.

Until then, any individual at increased risk for head trauma should be aware of their susceptibility to neurodegenerative diseases. The unknown prevalence of CTE in at-risk populations coupled with the observation that millions of former athletes and veterans have aged normally suggests that one or more additional factors need to be present in order for repeated mTBIs to trigger the development of CTE. Continued media exposure of the association between playing contact sports and the development of CTE ensures a great deal of research will rightly be devoted to answering these questions.

The most poignant need in TBI treatment is a reliable and sensitive diagnostic test(s) that can discern the presence, and level of severity, of TBI. Several biomarkers have been explored (see Jeter et al., 2013, Marion et al., 2010 for review), including serum-, cerebrospinal fluid- and blood-based proteins, neuroimaging methods, and behavioral and cognitive measures. No single method, however, has proven to be both sensitive and specific enough to delineate the presence or severity of TBI accurately and the majority of work in this field has only addressed severe TBI. Attempts at using several methods in combination with individual symptoms to create more effective “biomarker signatures” are underway, but at present, an easily administered, cost effective and accurate method of TBI (especially mild) diagnosis is lacking. Behavioral indices of brain injury that utilize reaction time, motor functions and autonomic responses such as pupil dilation and visual convergence are being developed and show promise as diagnostic tools for the detection of TBI that may have great ecological utility. Such a biomarker is greatly needed for the diagnosis of TBI in all its severities. Recommendations as to which biomarkers are the best candidates for practical diagnostic use were delineated in a workshop in 2010 (Marion et al., 2010). This workshop suggested that a minimum of three tests will be needed to accurately assess TBI severity and ideally would take no longer than 15 min to administer. Candidate diagnostic tests range from serum and blood based biomarkers to indices of behavior (i.e. balance, reaction time) and cognition (i.e. memory tests) to neuroimaging measures (i.e. MRI, EEG). The precise combination of which is the focus of ongoing research.

Heterogeneity of TBI occurrences and symptoms make successful treatment strategies that can be applied to the majority of TBI patients difficult to develop. What may work for one individual will not necessarily work for another. A means of assessing who is likely to respond to a particular type of treatment would be a valuable rehabilitative tool. A recent study showed that when comparing acute symptoms to chronic symptoms (postconcussion and mental health) after mTBI that chronic symptoms are a better predictor of outcome 6–12 months after injury (Mac Donald et al., 2015). This is not surprising; however, it underscores the need to discover variables with high predictive value of outcome after mTBI acutely. Such variables may be genetic, epigenetic or environmental factors not yet explored in the context of outcome after injury. For instance, psychosocial factors such as social status and early-life environment are associated with overall mortality rates and prevalence rates of many diseases (Adler and Stewart, 2010, Braveman et al., 2011). Such an association may be present in terms of outcome after TBI as well, but to our knowledge no empirical attempts have been made to explore this possibility.

In lieu of such predictive information, there is a movement to tailor treatment interventions to an individual's most salient symptoms, regardless of TBI severity or comorbid diagnoses (Brenner et al., 2009). Such intervention strategies combined with rigorous experimental design will undoubtedly improve evidence based treatment for future brain injured patients.