OVER THE PAST 3 DECADES, veterans from the 1990-1991 Gulf War (GW) have consistently reported numerous chronic health symptoms affecting their quality of life.1-14 Despite considerable investment directed at discovering the etiologies of these chronic health symptoms among GW veterans, the cause or causes remain elusive.15 In the more recent wars in this region (Operations Enduring Freedom and Iraqi Freedom [OEF/OIF]), veterans have also reported some similar chronic health symptoms, which have been primarily attributed to mild traumatic brain injuries (mTBI).16-19 However, the relationship between TBI and chronic health symptoms in veterans from the 1990-1991 GW has not been studied. Due to the limited ground war, TBI was not considered a risk factor for chronic multisymptom illness because these types of injuries were thought to be infrequent or nonexistent during the GW.

The constellation of symptoms associated with GW deployment has been termed by the Centers for Disease Control and Prevention as chronic multisymptom illness20 (CMI; also referred to as Gulf War Illness, Gulf War Syndrome, or Medically Unexplained Symptoms).21 Some 12% to 66% of GW veterans have endorsed symptoms consistent with CMI.15 The etiology of CMI has been controversial because of inconsistent and varied results. Theories have speculated on the impact of stress, prophylactic treatments (eg, vaccinations and anti-nerve gas pills), and environmental contaminants (eg, sarin/cyclosarin, pesticides).15,22

In general, military personnel have increased rates of TBIs when compared with the civilian population.23 Since the Department of Defense began documenting TBIs in 2000, nearly 300 000 cases have been reported. Battle-related TBIs only account for a minority of these injuries (~11%) with the most common causes including motor vehicle accidents (about 20%), falls (about 20%), and being struck by or against objects (15%).24 The majority of these injuries are classified as mild and occur when a force to the head or neck results in the transient disruption of brain functioning causing loss of consciousness for less than 30 minutes, alteration of mental state for less than 24 hours, and posttraumatic amnesia for less than 1 day.25 Mild TBIs often go unrecognized when the acute symptoms are of short duration (eg, a few seconds), are underreported, or are attributed to other causes by the patient or clinician26; as such mTBI has been described as a "silent epidemic."27

In response to growing concerns about the effects of unrecognized mTBIs in OEF and OIF veterans, the Department of Veterans Affairs mandated a standardized screening and comprehensive evaluation process for all returning veterans starting in 2007. Based on this postdeployment evaluation, more than 30 000 veterans have received a clinical diagnosis of TBI, but more than 44 000 reported moderate-to-severe postconcussive symptoms.28 Those with TBI also report chronic symptoms, some of which are similar to those experienced by GW-deployed veterans including neck aches, backaches, irritability, fatigue, and nausea. The extent to which these subjective self-reported health symptoms indicate an undiagnosed brain injury is unclear. However, military personnel training to safely use explosives have shown an increase in an objective biomarker of brain injury in addition to reporting an increase in postconcussive symptoms, despite not being diagnosed with TBI.29 Similar findings have also been reported in various nonmilitary cohorts.30-33 Taken together, these findings suggest that although postconcussive symptoms are recognized by the patient and may be correlated with an objective biomarker of brain injury, they do not always lead to a clinician-diagnosed TBI.

The degree to which TBI contributes to chronic health symptoms reported by veterans from the 1990-1991 GW is unknown. Given that the etiology of these symptoms remains unclear, and that TBI has been linked to similar postdeployment symptoms in other veteran populations, this study aimed at examining the cross-sectional relationships among self-reported TBI and health symptoms, CMI, and health-related quality of life in GW veterans. We hypothesized that GW veterans with reported TBI will endorse significantly higher rates of health symptoms and CMI as well as worse health-related quality of life.

METHOD

Participants

Participants included 1 274 GW veterans from the Devens Cohort Study, which is composed of approximately 100 US Army units that returned from the war in 1991 through Fort Devens in Massachusetts. Approximately 92% of the participants reported being enlisted at the time of the GW, and the remaining 8% reported being officers. The Devens Cohort Study was the first research investigation of GW veterans following their return from deployment.34 The initial purpose of the study was to assess psychological readjustment postdeployment. Later assessments of the cohort included both physical and emotional health concerns. The cohort completed postdeployment questionnaires at the following 3 time points: 1991, 1992 to 1993, and 1997 to 1998. In addition, a stratified random subset of the cohort was assessed in-person in 1994 to 1996.3,5,7,8,10 Cross-sectional data from the 1997 to 1998 survey were analyzed for this study.3-10 To determine TBI history, participants reported "yes" or "no" if they have ever had a "head injury, concussion, or period of being knocked unconscious." One hundred fifty-six participants self-reported a history of TBI (12.2% of the sample; TBI group) and 1 118 denied a TBI history (no-TBI group). The groups were similar in age, education, race, and ethnicity (see Table 1).

TABLE 1 Demographic characteristics

Health symptom checklist

The health symptom checklist (HSC) is a 52-item self-report questionnaire assessing the presence or absence of bothersome health symptoms over the past 4 weeks. Participants were instructed to check "yes" if the symptom was present during that time frame and "no" if it had been absent. If present, the participants indicated whether the symptom was bothersome either "sometimes" or "a lot." The focus of this study was the presence or absence of symptoms over the past 4 weeks.

CMI criteria

Centers for Disease Control and Prevention criteria for CMI include the presence of persistent health symptoms for 6 months or longer in 2 of the 3 following categories: fatigue, musculoskeletal factors, and mood and/or cognition. Participants self-reported the presence or absence of symptoms in each of these domains over the past 6 months and were classified accordingly as exhibiting CMI or not.20

RAND 36-Item Short Form Health Survey 1.0

The RAND 36-Item Short Form Health Survey 1.0 (SF-36) is a 36-item self-report questionnaire used to evaluate health-related quality of life. This commonly used measure has well-established reliability and validity for a diverse range of populations including persons with TBI.35-38 Higher scores represent better functioning. Responses load onto 2 component scores, Physical Component Summary (PCS) and Mental Component Summary (MCS); each comprises 4 individual subscales. The PCS includes these subscales-(1) Physical Functioning, which evaluates limitations in performing physical activities; (2) Role-Physical, which evaluates limitations with work or other daily activities due to physical health; (3) Bodily Pain, which evaluates limitations with work or other daily activities due to pain; and (4) General Health, which evaluates the respondents' perception of overall health. The MCS includes the following subscales: (1) Vitality, which evaluates feeling tired or worn out; (2) Social Functioning, which concerns problems with social activities due to physical and emotional problems; (3) Role-Emotional, which evaluates problems with work or other activities due to emotional problems; and (4) Mental Health, which addresses feelings of nervousness or depression. Scores were calculated using the method described by Ware et al.39

Statistical analysis

Analysis of HSC

The chi-square test of independence was used to compare the rates of each of the 52 health symptoms in the TBI and no-TBI groups. To control for type I errors, alpha level was conservatively adjusted to 0.001 (0.05/52 = 0.001) using the Bonferroni method.40,41 Eight participants in the no-TBI group, who completed less than 80% of the HSC, were excluded from the analysis.

Analysis of CMI

The chi-square test of independence was used to compare the rate of CMI between the TBI and no-TBI groups. An alpha level of 0.05 was adopted. The CMI status was available for all participants in the TBI group, but not for 7 participants in the no-TBI group, and therefore they were excluded from the analysis.

Analysis of SF-36

Because of the nonnormal distribution of the data, between-group comparisons were performed using the nonparametric Mann-Whitney U test. For the PCS and MCS, an alpha level of 0.05 was adopted. For follow-up analyses of the 4 individual subscales, which make up each component score, alpha was adjusted to 0.0125 (0.05/4 = 0.0125). Fifteen of the 1 274 participants (1.2%; 1 in the TBI group; 14 in the no-TBI group) had at least 1 missing data point on the SF-36 and therefore were excluded from the analysis.

RESULTS

Comparison between the TBI and no-TBI groups on the HSC

Rates of the following 8 health symptoms were significantly higher in the TBI group when compared with the no-TBI group-racing heart, nausea or upset stomach, muscle weakness or fatigue, neck aches or stiffness, backaches, thickened saliva, fatigue or easily tired, and excessive anger or irritability (all P <= .001) (see Table 2).

TABLE 2 Rates of health symptoms in Gulf War veterans with and without TBI

Comparison between TBI and no-TBI groups on rate of CMI

Rate of CMI in the TBI group (76%) was significantly higher than in the no-TBI group (59%; P < .000) (see Figure 1).

Figure 1. Rate of CMI for the TBI group and no-TBI group. The TBI group had higher rate of CMI when compared with the no-TBI group.CMI, chronic multisymptom illness; TBI, traumatic brain injury.

Comparison between TBI and no-TBI groups on the SF-36

The TBI group reported significantly worse functioning for the PCS compared with the no-TBI group (TBI = 47.7 +/- 10.06; no-TBI = 44.7 +/- 11.18; P = .000), but not for the MCS (TBI = 45.3 +/- 12.0; no-TBI = 43.5 +/- 12.8; P = .143). Significant between-group differences were found on 3 of the 4 individual subscales that make up the PCS. Relative to the no-TBI group, the TBI group reported worse functioning for Physical Functioning (P = .001), Bodily Pain (P = .001), and General Health (P = .001) individual subscales. Groups were similar on the Role-Physical individual subscale (P = .015). Significant between-group differences were found on 1 of the individual subscales that make up the MCS. Relative to the no-TBI group, the TBI group reported worse symptoms on Vitality (P = .009), but not Social Functioning (P = .025), Role-Emotional (P = .118), or Mental Health (P = .056) (see Figures 2 and 3).

Figure 2. Mean scores for the Physical and Mental Health summary scores of the SF-36. Error bars represent standard deviation. The TBI group reported significantly more problems on Physical Health, but not Mental Health.

Figure 3. Mean score for each of the 8 individual subscales for the SF-36. Error bars represent standard deviation. The TBI group reported significantly more problems on the following individual subscales: Physical Functioning, Bodily Pain, General Health, and Vitality.

DISCUSSION

This study examined the relationships among self-reported TBI and health symptoms, CMI, and health-related quality of life in a cohort of 1990-1991 GW veterans. Traumatic brain injury has not previously been considered a contributing factor to the chronic health symptoms experienced by GW veterans, but it has been connected to poorer health in OEF/OIF veterans and civilian populations.16-19,42-44 Overall, we found that GW veterans with a self-reported history of TBI report higher rates of health symptoms and are more likely to meet criteria for CMI when compared with veterans without a self-reported history of TBI. These symptoms are not inconsequential, as results from the SF-36 indicate that they appear to interfere with everyday activities of daily living, including performing household chores, working, and leisure activities.

Consistent with our hypotheses, rates of self-reported health symptoms (as measured by HSC) were higher overall for GW veterans with a self-reported history of TBI, with 8 specific symptoms spanning 5 body systems (ie, cardiac, gastrointestinal, musculoskeletal, neurological, and psychological) reported at a significantly higher rate compared with GW veterans without this history. These findings remained even after conservatively adjusting for multiple comparisons. Five of the 8 health symptoms correspond to 1 of the 3 categories (ie, fatigue, musculoskeletal, and mood-cognition) that define CMI criteria. Specifically, the symptom "fatigue or easily tired" is consistent with the CMI category of fatigue. The remaining 4 symptoms are consistent with CMI categories of musculoskeletal (ie, muscle weakness or fatigue, neck aches or stiffness, and backaches) and mood-cognition (ie, excessive anger or irritability).

Taken together, these findings suggest that TBI in this cohort of GW veterans is associated with specific health symptoms, several of which likely contribute to CMI diagnosis. This accords with our additional observation that a history of self-reported TBI in this veteran population was associated with an increased rate of CMI. Previously, CMI has been primarily attributed to certain war-related environmental exposures, such as sarin/cyclosarin, pesticides, and the use of anti-nerve gas pills.15,22 In addition to these exposures, results from this study indicate that TBI may also contribute to CMI. This is particularly relevant, as multiple brain insults (eg, pesticides and TBI) may cause an abnormally persistent neuroimmune response with each subsequent insult, contributing to chronic health symptoms such as those seen in GW veterans.45-48

Gulf War veterans with self-reported TBI also reported worse health-related quality of life associated with physical but not mental health as indicated by scores on the SF-36. Evidence suggests that clinically or socially meaningful differences in MCS and PCS scores range from 2.5 and 5 points.49-51 Gulf War veterans with a self-reported history of TBI endorsed average scores approximately 3 points lower than veterans without this history on the PCS, indicating a clinically meaningful difference in physical functioning. This finding adds to the existing literature in other populations such as OEF/OIF veterans, indicating that TBI contributes to a wide range of health symptoms and poorer quality of life.16,52 Gulf War veterans and OEF/OIF veterans report some similar health symptoms, which may stem, in part, from exposure to TBI.53 Gulf War veterans initially reported health concerns immediately upon returning from war, and, in general, these symptoms have persisted or worsened over the past 3 decades.13,14,54,55 The long-term trajectory of the health and functioning of OEF/OIF veterans is unknown. However, if TBI is contributing to the health symptoms reported by both veteran groups, then OEF/OIF veterans may also experience persistent health outcomes similar to those of GW veterans. There is already some evidence indicating that OEF/OIF veterans are developing chronic long-term debilitating health problems. For example, a cohort of these veterans reported a slowly progressive worsening of health-related quality of life, as indicated by scores on the same measure used in this study (ie, SF-36).52 Factors contributing to these problems were not explored, but Falvo and colleagues52 suggest that mTBI and respiratory-related illnesses may be contributing to these findings. Change in self-reported quality of life is particularly worrisome, as it is associated with increased healthcare utilization and mortality in veterans.56-58 However, it is important to highlight that the present findings may or may not apply to OEF/OIF veterans due to differences in TBI assessment. In general, studies of OEF/OIF veterans have used external verification of TBI, whereas this study relied on self-report, which is more likely to be influenced by systematic response bias.

Although this study has its strengths, there are a number of limitations that warrant discussion. Retrospective self-report data were used with no external confirmation of TBI. Furthermore, this study only asked veterans to indicate the presence or absence of a TBI. No information was collected on the nature of the injury such as timing (ie, before, during, or after the war), severity, frequency, or cause. This information would be useful to collect in future studies to determine whether health outcomes vary depending on the nature of injury (eg, 1 TBI vs multiple TBIs, mild vs moderate). Self-report of TBI may be unreliable if veterans were unaware that they had a TBI. It is also possible that some veterans had a tendency to endorse items indiscriminately (ie, systematic response bias). Future studies would benefit from external verification of TBIs, such as clinical evaluations to increase accuracy of reporting. Use of an updated definition of TBI would also be beneficial to future studies in increasing accuracy of reporting, as scientific understanding of TBI has changed over the past few years.59 Studies with other populations have demonstrated significant increases in self-reports of TBI when a clear definition is provided.60

This study is also limited by the use of a single diagnostic criterion for defining the cluster of symptoms related to the 1990 to 1991 Gulf War. More than a half-dozen diagnostic criteria for this syndrome have been developed,61 including recent recommendations by the Department of Defense and Department of Veterans Affairs.21 This study used the most commonly cited diagnostic criteria,20 but different outcomes may be found if other classification systems are used.

In addition, the findings may be confounded by higher rates of psychiatric conditions in veterans with TBI. Both TBI and psychiatric conditions can result in increased health symptoms and decreased quality of life,62-67 though the temporal relationship appears to be bidirectional. In this study, participants indicated the presence or absence of "frequent anxiety or nervousness" or "frequently feeling depressed." The rates of self-reported anxiety and depression between veterans with and without a self-reported history of TBI did not reach statistical significance. Likewise, there were no significant differences between groups on the MCS from the SF-36 or the Mental Health subscale of the SF-36, the latter of which assesses anxiety and depression. Taken together, these findings suggest that psychological health is unlikely to explain the present findings, though comprehensive psychiatric evaluations were not performed.

To our knowledge, this is the first study to examine whether reported TBI is associated with health symptoms, CMI, and health-related quality of life in GW veterans. Our results indicate that those reporting a history of TBI compared with no TBI also report significantly more health symptoms, increased rate of CMI, and poorer health-related quality of life. Past research has focused on a wide range of causes for CMI such as environmental exposures and contaminants, which have also been the primary focus for treatment.15 Further research is needed to determine whether history of TBI is a causative factor of longer-term CMI and related health symptoms, either alone or as part of a multiple-hit hypothesis. Potential research could include investigating the role of TBI in combination with the environmental exposures unique to GW deployment.

REFERENCES