AFTER VIRTUALLY EVERY major war, a significant

Transcription

1 ORIGINAL ARTICLE Health Outcomes Associated With Military Deployment: Mild Traumatic Brain Injury, Blast, Trauma, and Combat Associations in the Florida National Guard Rodney D. Vanderploeg, PhD, Heather G. Belanger, PhD, Ronnie D. Horner, PhD, Andrea M. Spehar, DVM, MPH, JD, Gail Powell-Cope, PhD, ARNP, FAAN, Stephen L. Luther, PhD, Steven G. Scott, DO 1887 An audio podcast accompanies this article. Listen at ABSTRACT. Vanderploeg RD, Belanger HG, Horner RD, Spehar AM, Powell-Cope G, Luther SL, Scott SG. Health outcomes associated with military deployment: mild traumatic brain injury, blast, trauma, and combat associations in the Florida National Guard. Arch Phys Med Rehabil 2012;93: Objectives: To determine the association between specific military deployment experiences and immediate and longerterm physical and mental health effects, as well as examine the effects of multiple deployment-related traumatic brain injuries (TBIs) on health outcomes. Design: Online survey of cross-sectional cohort. Odds ratios were calculated to assess the association between deploymentrelated factors (ie, physical injuries, exposure to potentially traumatic deployment experiences, combat, blast exposure, and mild TBI) and current health status, controlling for potential confounders, demographics, and predeployment experiences. Setting: Nonclinical. Participants: Members (N 3098) of the Florida National Guard (1443 deployed, 1655 not deployed). Interventions: Not applicable. From the Mental Health and Behavioral Sciences Psychology Service (Vanderploeg, Belanger), Physical Medicine and Rehabilitation Service (Scott), Research Service (Spehar), and Nursing Service (Powell-Cope, Luther) and Health Services Research and Development (HSR&D)/Rehabilitation Research and Development (RR&D) Center of Excellence: Maximizing Rehabilitation Outcomes (Vanderploeg, Belanger, Spehar, Powell-Cope, Luther, Scott), James A. Haley Veterans Hospital, Tampa, FL; Departments of Psychiatry and Neurosciences (Vanderploeg), Psychology (Vanderploeg, Belanger), Public Health (Powell-Cope, Luther), and Internal Medicine (Scott), University of South Florida, Tampa, FL; Defense and Veterans Brain Injury Center, James A. Haley Veterans Affairs Medical Center, Tampa, FL (Vanderploeg, Belanger, Scott); and Department of Internal Medicine (Horner), University of Cincinnati, Cincinnati, OH. Presented in part to the Federal Interagency Conference on Traumatic Brain Injury, June 15, 2011, Washington, DC. Supported by the Department of Veterans Affairs, Veterans Health Administration, and Veterans Administration HSR&D (grant no. CCN ). Further support was provided by the James A. Haley Veterans Hospital and its HSR&D/RR&D Center of Excellence. The views expressed herein are those of the authors and do not necessarily reflect the views or the official policy of the Department of Veterans Affairs or the US government. No commercial party having a direct financial interest in the results of the research supporting this article has or will confer a benefit on the authors or on any organization with which the authors are associated. Reprint requests to Rodney D. Vanderploeg, PhD, Mental Health and Behavioral Sciences Psychology Service (116B), James A. Haley Veterans Hospital, Bruce B. Downs Blvd, Tampa, FL 33612, Rodney.Vanderploeg@va.gov. In-press corrected proof published online on Jul 24, 2012, at /12/ $36.00/0 Main Outcome Measures: Presence of current psychiatric diagnoses and health outcomes, including postconcussive and non postconcussive symptoms. Results: Surveys were completed an average of 31.8 months (SD 24.4, range 0 95) after deployment. Strong, statistically significant associations were found between self-reported military deployment-related factors and current adverse health status. Deployment-related mild TBI was associated with depression, anxiety, posttraumatic stress disorder (PTSD), and postconcussive symptoms collectively and individually. Statistically significant increases in the frequency of depression, anxiety, PTSD, and a postconcussive symptom complex were seen comparing single to multiple TBIs. However, a predeployment TBI did not increase the likelihood of sustaining another TBI in a blast exposure. Associations between blast exposure and abdominal pain, pain on deep breathing, shortness of breath, hearing loss, and tinnitus suggested residual barotrauma. Combat exposures with and without physical injury were each associated not only with PTSD but also with numerous postconcussive and non postconcussive symptoms. The experience of seeing others wounded or killed or experiencing the death of a buddy or leader was associated with indigestion and headaches but not with depression, anxiety, or PTSD. Conclusions: Complex relationships exist between multiple deployment-related factors and numerous overlapping and cooccurring current adverse physical and psychological health outcomes. Various deployment-related experiences increased the risk for postdeployment adverse mental and physical health outcomes, individually and in combination. These findings suggest that an integrated physical and mental health care approach would be beneficial to postdeployment care. Key Words: Blast injuries; Brain concussion; Military personnel; Neurobehavioral manifestations; Outcome assessment; Rehabilitation; Sequelae by the American Congress of Rehabilitation Medicine AFTER VIRTUALLY EVERY major war, a significant percentage of combatants return with a variety of nonspecific symptoms that include sleep problems, fatigue, irritability, headaches, other bodily aches and pains, concentration difficulties, and memory complaints, 1 with disagreement about CI OR PCS PTSD TBI List of Abbreviations confidence interval odds ratio postconcussion symptom posttraumatic stress disorder traumatic brain injury

2 1888 MILITARY DEPLOYMENT AND HEALTH OUTCOMES, Vanderploeg causation. 2 A smaller subset meets diagnostic criteria for specific psychiatric conditions including posttraumatic stress disorder (PTSD) and depression. 1,3-5 These postdeployment symptoms and conditions often occur in individuals who experienced multiple deployment-related stresses, traumas, and physical injuries (either minor or more significant), one of which may have been traumatic brain injury (TBI). 3,6,7 Recent efforts to attribute symptoms after deployment to either mental health factors (eg, PTSD) or physical injury (eg, mild TBI) neglect the long history of these unexplained symptoms after war and the role that cultural factors and time period may play in attributing causation. 8 Findings 9,10 showing the importance of both mental health and physical injury factors (eg, PTSD and mild TBI) to predicting somatic symptoms illustrate that adverse health outcomes after military deployment to combat zones likely are the consequence of complex and multifactorial contributions including innate biologic factors, predeployment experiences and conditions, and specific deployment-related events such as witnessing or experiencing psychologically distressing events, combat, blast exposure, and physical injuries (including TBI) and, perhaps, the deployment, per se. These multifactorial contributions of deployment to various co-occurring and overlapping postdeployment health outcomes are yet to be fully elucidated. Existing studies often rely on 1 or more conditions after war (eg, PTSD) to explain the variance associated with other postwar health outcomes (eg, postconcussive symptoms) as though causal. 3,5 For example, PTSD and depression have been suggested as mediators of the effects of mild TBI on health outcomes. 3 However, PTSD and depression are both health outcomes, and both are associated with a variety of emotional, physical, and cognitive symptoms that overlap with other health outcomes. Therefore, when multiple postdeployment health outcomes have overlapping symptom presentations, controlling for 1 or more outcomes (eg, PTSD and depression) will inevitably minimize associations between other postdeployment outcomes and potential etiologic deployment traumas such as mild TBI. 3,5,11 This report presents the findings from an epidemiologic study, the purpose of which was to elucidate the unique contributions of various deployment-related exposures and injuries, including mild TBI, to current postdeployment physical, psychological, and general health outcomes in a large sample of members of the Florida National Guard who served in Iraq or Afghanistan. The frequency of mild TBI in relation to various blast injury mechanisms is also explored, as are the effects of single versus multiple TBIs on health outcomes. METHODS Recruitment In 2009 to 2010, we conducted an anonymous online survey of the currently active members of the Florida National Guard. Approximately 10,400 letters were mailed from the Florida National Guard Headquarters to their members inviting participation in a Web-based research survey. The protocol was approved by the Institutional Review Board of the University of South Florida. Of 10,400 letters mailed, approximately 700 letters were returned for insufficient addresses and 4005 individuals responded for a response rate of 41.3%. Of those deployed (46.6% of respondents), Guard members had returned from Operation Enduring Freedom/Operation Iraqi Freedom deployment an average of months (range 0 95) prior to the survey. Because of the anonymity protections of the survey, it was not possible to compare respondents with nonrespondents. Measures The survey was developed specifically for this study, and the resulting questionnaire assessed 3 domains of possible predictors of health outcomes: (a) demographics, (b) predeployment psychologically traumatic events or history of prior TBI, and (c) Operation Enduring Freedom/Operation Iraqi Freedom deployment and related experiences including combat exposure, and exposure to potentially physically injurious and psychologically traumatic events including mild TBI and blast exposures. Current psychological and physical health outcomes were assessed as well. Predeployment Psychological Trauma or TBI Guard members were asked about any predeployment traumatic events including experiences that were frightening, horrible, or upsetting, and then they were asked to indicate the nature of the trauma. Responses included sexual trauma, domestic abuse, losing a child in an accident, losing family members to violent death, witnessing murder, being involved in a fatal car accident, near drowning, being mugged or assaulted, and similar types of traumas. The interpretation of traumatic was their own. This variable was coded positively if they endorsed any such traumatic event in their past. In addition, predeployment head injury was coded if they endorsed ever having a head injury that knocked them unconscious or resulted in memory gaps for the time immediately after the event. They were then asked to fill in the amount of time any memory gap lasted (most were less than an hour and all were less than a day). Deployment-Experienced Trauma Experiences, Blast, Combat, TBI, and Other Injuries Exposure to potentially traumatic combat experiences was determined by asking whether the respondent had seen others wounded, killed or dead or experienced the death of a buddy to whom you were very close or experienced the loss of a leader whom you valued and trusted. Presence of physical trauma or injury was determined by asking the respondent whether he/she had deployment-related loss of body function or limb, injury in a motor vehicle accident, injury in combat, injury via blast including burns, wounds, broken bones, amputations, breathing toxic fumes, internal injuries, crush injuries, or other injuries. Probable deployment-related TBI was determined by asking the respondent about any event(s) resulting in loss of consciousness, blacking out, or memory gaps. Thus, individuals who were only dazed and confused without memory gaps would not be included. These were not included because combat events such as unexpected ambushes, firefights, or nearby blasts would likely be experienced as confusing, disorienting, or dazing but not necessarily indicate a brain injury. Participants were then asked about the duration of any memory gaps (most were less than an hour and all were less than a day). Participants were not asked about the duration of unconsciousness, because it was felt that it could not be reliably selfreported. Ten percent (n 144) of deployed Guard members reported a deployment-related TBI, 23 of whom reported more than 1. For the purposes of the primary analyses, probable TBI was coded as present (1 or more occurrences) or absent. Supplemental analyses exploring the effects of 1 versus multiple probable TBIs included predeployment TBIs in the analyses, and TBI was coded single versus multiple (2 or more). Blast exposure was determined through 4 questions designed to assess exposure to the primary blast wave ([experiencing] a wave of highly compressed gas... that felt almost like smashing into a wall ), secondary blast effects (after a blast being peppered or hit by debris, shrapnel, or other items ), tertiary

3 MILITARY DEPLOYMENT AND HEALTH OUTCOMES, Vanderploeg 1889 effects ( thrown to the ground or against some stationary object like a wall or vehicle by the explosion ), and quaternary effects ( experiencing any of the following injuries as a result of an explosive blast: burns, wounds, broken bones, amputations, breathing toxic fumes, or crush injuries from collapsed structures falling onto you ). Blast exposure was recategorized into none, nonprimary blast, and primary blast. It is important to note that the primary blast category did not exclude individuals with other blast mechanisms; it simply indicates that a primary blast effect was reported as part of the experience. Combat intensity was measured with the use of 4 items from the military Post-Deployment Health Assessment 12 (DD Form 2976, version APR 2003), resulting in scores ranging from 0 to 8 that were recategorized into quartiles from none to high combat intensity. Postdeployment Psychological Health Outcomes: Depression, PTSD, Generalized Anxiety, and Alcohol Abuse Presence of current symptoms (ie, symptoms during the past month) of major depressive disorder, general anxiety, and PTSD was assessed by the 9-question, depression-assessment module of the Patient Health Questionnaire, 13,14 the 7-question Generalized Anxiety Disorder scale, 15 and the 17-item National Center for PTSD Checklist, 16 Civilian Version, respectively. (The civilian version was used because the goal was to assess current symptoms, and then examine the association between current symptoms and potential deployment-related factors.) For probable major depressive disorder, subjects had to meet the criteria of the American Psychiatric Association s Diagnostic and Statistical Manual of Mental Disorders, fourth edition, 17 and report impairment in work, home, or interpersonal functioning at the very difficult level. 13 For probable anxiety syndrome, there had to be nervousness or anxiety more than half the time; 3 of 6 other anxiety symptoms more than half the time; and impairment in work, home, or interpersonal functioning at the very difficult level. 15 For probable PTSD, subjects had to meet the Diagnostic and Statistical Manual of Mental Disorders, fourth edition, 17 symptom criteria (1 intrusion symptom, 3 avoidance symptoms, and 2 hyperarousal symptoms experienced at the moderate level or higher) and have substantial distress, as measured by a total score of at least 50 (range 17 85, with higher scores indicating a greater number and severity of symptoms). Following the PTSD Checklist, respondents were asked how difficult these problems made it for them to do their work, take care of things at home, or get along with other people. A response at the very difficult level or higher was required to be classified as having probable PTSD. 18 Participants also completed the Alcohol Use Disorder Identification Test, which identified those with excessive drinking. 19,20 Postdeployment Physical Health Outcomes As part of the 12-Item Short Form Health Survey (SF-12), Guard members were asked to rate their overall health (from excellent to poor ), which was recategorized into excellent/ very good/good versus fair/poor. Current physical symptoms were measured by the Post-deployment Health Assessment 12 (DD Form 2976, version APR 2003), which asks about the presence of 20 symptoms. This was supplemented by a review of systems that assessed 38 symptoms across 10 body systems, and the 22-item Neurobehavioral Symptom Inventory, 21 which assesses physical, cognitive, and emotional symptoms associated with concussions or TBI. The Neurobehavioral Symptom Inventory was also used to calculate a symptom-based postconcussion syndrome variable based on the International Statistical Classification of Diseases and Related Health Problems, 10th Revision, 22 criteria of at least 1 physical symptom (dizzy, headaches, noise sensitivity, fatigue), 1 cognitive symptom (problems with concentration, memory, or thinking), 1 emotional symptom (anxiety, depression, poor frustration), and sleep problems, all at a moderate level of severity or higher. This same binary recategorization of postconcussion symptom (PCS) severity was used when examining individual symptoms. Statistical Analysis Data analysis involved chi-square and analysis of variance for assessing associations in the univariate analyses while the multivariate analyses were conducted with the use of multiple logistic regression to calculate the odds ratios (ORs) of various predictor variables on current health variables, controlling for all other predictors (ie, demographic, predeployment psychological trauma or TBI, and deploymentrelated factors including combat, physical injuries, exposure to potentially traumatic combat experiences, blast exposure, and TBI [ie, variables listed in table 1]). As this was an exploratory study, the alpha error rate was set at P.01. SPSS software a was used for all analyses. To assess the interrelatedness of our key outcomes, intercorrelations were examined. The majority of the correlations were in the midrange of Spearman correlation (.42.55). Anxiety was correlated more highly with both depression (r.76) and PTSD (r.62). In exploratory analyses when we adjusted for anxiety, ORs were nonsignificant. This was to be expected in that anxiety is an aspect of PTSD and overlaps in symptomatology with major depression. Hence, our strategy of examining outcomes without adjustment for other outcomes was utilized. RESULTS Participants Of the 4005 individuals who responded, we excluded 423 who did not fully complete the survey, 371 who completed the survey more than 1 time, and 113 who gave inconsistent or impossible responses, resulting in a final sample of 3098 respondents. Of these respondents, 1655 (53.4%) had not deployed, 1123 (46.6%) had 1 deployment, and 320 (10.3%) had 2 or more deployments. The deployed group differed from the nondeployed group in several characteristics, including more likely to be men, married, and with some college education as well as having a history of psychological trauma and TBI (see table 1). Deployment Experiences or Traumas Exposure to deployment-related traumatic combat experiences was common (57.5%), as was exposure to blasts (51.5%), the latter about equally divided between those reporting exposure to a primary blast wave and those reporting blast exposure without primary blast wave contact. One hundred forty-four (10%) of the deployed sample reported a probable deployment-related TBI, all of which was categorized as mild, with memory gaps of less than a day. Other deployment-related physical injuries were reported by 14.3% (n 206) of the sample; 80 (39%) of these also reported a probable deployment-related TBI. Three hundred thirty (22.9%) reported no combat-related experience while deployed.

4 1890 MILITARY DEPLOYMENT AND HEALTH OUTCOMES, Vanderploeg Characteristic Table 1: Characteristics of the Study Population* Total Sample (N 3098) Nondeployed (n 1655) Deployed (n 1443) Demographic characteristics, n (%) Female sex 527 (17.0) 345 (20.8) 182 (12.6).001 Minority race 1184 (38.2) 701 (42.4) 483 (33.5).001 High-school education or less 816 (26.3) 529 (32.0) 287 (19.9).001 Marital status.001 Single/never married 942 (30.4) 677 (40.9) 265 (18.4) Married 1628 (52.6) 706 (42.7) 922 (63.9) Divorced, separated, or other 528 (17.0) 272 (16.4) 265 (17.7) Predeployment characteristics, n (%) Prior psychological trauma 702 (22.7) 310 (18.7) 392 (27.2).001 Prior probable TBI 365 (11.8) 175 (10.6) 190 (13.2).030 Deployment characteristics, n (%) Deployed 1443 (100.0) Physical injury 206 (14.3) Exposure to potentially traumatic combat experiences 830 (57.5) Quartile of combat intensity 1 (none) 330 (22.9) 2 (low) 308 (21.3) 3 (medium) 348 (24.1) 4 (high) 457 (31.7) Blast exposure None 700 (48.5) Nonprimary blast 380 (26.3) Primary blast 363 (25.2) Probable TBI 144 (10.0) *P values were calculated with the use of the 2 test. Predeployment traumatic events experienced as frightening, horrible, or upsetting, and any predeployment head injury with unconscious or postevent memory gaps. Psychological trauma exposure was report of seeing others wounded or killed, or experiencing the death of a buddy or leader. Physical trauma or injury was deployment-related self-reported loss of limb or body function, or fractures, wounds, burns, or other injuries. Deployment-related TBI was any event(s) resulting in loss of consciousness, blacking out, or memory gaps. Combat intensity was assessed using 4 items from the Post-Deployment Health Assessment, resulting in scores ranging from 0 to 8 categorized into quartiles. Blast exposure was determined through 4 questions asking about exposure to the primary blast wave, secondary blast, tertiary, and quaternary effects; recategorized into none, nonprimary blast, and primary blast. P Postdeployment Health Outcomes In unadjusted associations of deployment experiences with various health outcomes, those deployed had significantly poorer health outcomes than did nondeployed Guard members on all measures considered (table 2). However, once demographic, predeployment, and other specific deploymentrelated predictors were taken into account, the association between deployment, per se, and all health outcomes disappeared (table 3). Having sustained a physical injury during deployment was significantly associated with many current health conditions, including self-ratings of poor overall postdeployment health (OR 2.37 [95% confidence interval (CI) ]), PTSD (OR 2.43 [95% CI ]), and PCS collectively (PCS complex) (OR 2.61 [95% CI ]). Supporting the notion that PCS are etiologically nonspecific, non-tbi physical injury was a significant predictor of most of these symptoms, with significant ORs ranging from 1.84 for headaches (95% CI ) to 3.12 for memory problems (95% CI ). Physical injury was also associated with arthritis (OR 2.66 [95% CI ]) and many non postconcussive pain and sensory symptoms. The experience of seeing others wounded or killed or experiencing the death of a buddy or leader during deployment was not significantly associated with postdeployment depression, PTSD, anxiety disorder, or excessive drinking. It was associated only with headaches (OR 1.69 [95% CI ]) and indigestion (OR 1.81 [95% CI ]) (see table 3). Combat experiences were significantly associated with a variety of postdeployment health conditions, sometimes monotonically related to combat intensity, but not consistently so. Only the highest level of combat intensity was significantly associated with probable PTSD (OR 4.83 [95% CI ]), although a monotonic association with intensity appears to be present (see table 3). The highest level of combat exposure was also associated with the PCS complex (OR 3.01 [95% CI ]) and with 3 specific postconcussive symptoms (ie, insomnia, balance, and memory problems). Moderate to high levels of combat exposure were significantly associated with back and muscle pain, indigestion, tinnitus, and irritability. Blast exposure was significantly associated with a number of postdeployment health conditions. Primary blast exposure was associated with a relatively higher likelihood of adverse health outcomes than was nonprimary blast exposure for pain on deep breathing (nonprimary blast OR 2.58 [95% CI ]; primary blast OR 3.25 [95% CI ]), hearing loss (nonprimary blast OR 1.63 [95% CI ]; primary blast OR 2.32 [95% CI ]), and ringing in the ears (nonprimary blast OR 1.77 [95% CI ]; primary blast OR 2.92 [95% CI ]). Only primary blast exposure was associated with abdominal pain (OR 2.28 [95% CI 1.32

5 MILITARY DEPLOYMENT AND HEALTH OUTCOMES, Vanderploeg 1891 Table 2: Reported Health Status After Deployment* Current Health Nondeployed (n 1655) Deployed (n 1443) Overall health rating, n (%) Fair/poor overall health 67 (4.0) 193 (13.4).001 Mental health conditions, n (%) Probable major depression 16 (1.0) 47 (3.3).001 Probable anxiety disorder 24 (1.5) 61 (4.2).001 Probable PTSD 32 (1.9) 100 (6.9).001 Excessive drinking 494 (29.8) 554 (38.4).001 Postconcussive symptom complex 54 (3.3) 162 (11.2).001 Physical health conditions, n (%) Hypertension 93 (5.6) 159 (11.0).001 Allergies 218 (13.2) 239 (16.6).009 Arthritis 54 (3.3) 127 (8.8).001 Non PCS, n (%) Chest pain 50 (3.0) 122 (8.5).001 Gait problems 25 (1.5) 59 (4.1).001 Back & muscle pain 262 (15.8) 612 (42.4).001 Abdominal pain 52 (3.1) 124 (8.6).001 Indigestion 72 (4.4) 231 (16.0).001 Shortness of breath 74 (4.5) 177 (12.3).001 Pain on deep breathing 34 (2.1) 81 (5.6).001 Hearing loss 142 (8.6) 423 (29.3).001 Ringing in ears 161 (9.7) 457 (31.7).001 PCS, n (%) Headaches 146 (8.8) 338 (23.4).001 Dizziness 45 (2.7) 138 (9.6).001 Fatigue 113 (6.8) 295 (20.4).001 Insomnia 157 (9.5) 370 (25.6).001 Balance problems 19 (1.1) 63 (4.4).001 Memory problems 93 (5.6) 262 (18.2).001 Concentration problems 103 (6.2) 257 (17.8).001 Irritability 107 (6.5) 270 (18.7).001 *P values were calculated with the use of the 2 test. Rating of overall health (from poor to excellent) recategorized into fair to poor versus good to excellent. Major depression was measured with the Patient Health Questionnaire; classified as major depression if Diagnostic and Statistical Manual of Mental Disorders, fourth edition, criteria were met and functional impairment was reported at the very difficult level. PTSD was measured with the Post-Traumatic Stress Disorder Checklist; meeting psychiatric-symptom criteria of the Diagnostic and Statistical Manual of Mental Disorders, fourth edition, and had a total score of at least 50 on the Post-Traumatic Stress Disorder Checklist. Anxiety disorder was measured with the 7-question Generalized Anxiety Disorder scale; there had to be worry and anxiety more than half the time, 3 other anxiety symptoms more than half the time, and impairment in work, home, or interpersonal functioning at the very difficult level. Excessive drinking was assessed with the Alcohol Use Disorder Identification Test using published criteria for excessive drinking. 17,18 Postconcussive symptom complex was assessed with the Neurobehavioral Symptom Inventory; symptom-based postconcussion syndrome criteria based on the International Statistical Classification of Diseases and Related Health Problems, 10th Revision, had to be met at least 1 physical symptom, 1 cognitive symptom, 1 emotional symptom, and sleep problems, all at a moderate level of severity or higher. Health conditions and non postconcussive symptoms were selfreported on a review of systems and conditions checklist, or on the Post-deployment Health Assessment. 10 PCS were assessed with the Neurobehavioral Symptom Inventory with symptoms reported at the moderate level of severity or higher. 3.94]), dizziness (OR 2.26 [95% CI ]), and back and muscle pain (OR 1.64 [95% CI ]). Finally, any blast exposure was associated with shortness of breath and 2 postconcussive symptoms fatigue and insomnia (see table 3). P Having experienced probable mild TBI or concussion during deployment was significantly associated with selfratings of poor overall postdeployment health (OR 2.52 [95% CI ]), mental health conditions (probable depression OR 4.89 [95% CI ]), probable anxiety (OR 3.91 [95% CI ]), probable PTSD (OR 2.63 [95% CI ]), the PCS complex (OR 2.47 [95% CI ]), and all the specific postconcussive symptoms examined with the exception of hearing loss and ringing in the ear. However, having experienced probable mild TBI during deployment was not significantly associated with any of the non-pcs examined. Supplemental TBI Analyses: Blast-Related TBI and the Effects of Multiple TBIs Of those individuals reporting a deployment-related probable mild TBI (n 144), 69 reported blast as the mechanism of injury. Of these, 46 reported a single blast-related TBI (66.7%), 13 reported 2 blast-related TBIs (18.8%), 5 reported 3 blastrelated TBIs (7.2%), and 5 reported more than 3 blast-related TBIs (7.2%). The potential blast-related mechanisms of injury were explored. Twenty of the 69 individuals (29%) reported that their blast-related probable mild TBI occurred in the context of combined primary (ie, blast wave), secondary (ie, hit by debris), and tertiary (ie, thrown against something by the blast) blast mechanisms, while 19 (27.5%) reported experiencing all 4 blast injury mechanisms when they sustained their probable mild TBI. Only 6 (8.7%) reported experiencing only a primary blast injury mechanism for their mild TBIs, and of these only 4 reported no other injuries. Thus, individuals who report primary blast exposure without other blast injury mechanisms rarely report a probable mild TBI (4 of 363 or 1% of those reporting experiencing a primary blast wave and 4 of 69 or 6% of those reporting a blast-related mild TBI). The remainder reported exposure to 2 or 3 mechanisms of blast injury (table 4). The cumulative effects of more than 1 probable mild TBI were explored on the major mental health outcomes associated with mild TBI in the primary analyses. For these analyses, all 3098 respondents were included, and probable TBIs prior to deployment were also considered in the analyses of multiple TBI effects. Combining deployment and non deploymentrelated probable TBI resulted in 2603 respondents reporting no TBI, 461 reporting a single TBI, and 34 reporting 2 or more TBIs. OR for single versus multiple TBI associations were calculated controlling for demographic characteristics, predeployment psychological trauma, deployment per se, and other deployment-related trauma exposures (ie, other physical injuries, psychological trauma exposures, combat intensity, and blast exposure). Each of the 4 major mental health outcomes (ie, probable major depression, probable generalized anxiety, probable PTSD, and PCS complex) associated with a probable mild TBI effect in the primary analyses showed a cumulative TBI effect in these supplemental analyses (table 5). To examine the potential increased vulnerability to sustaining a subsequent TBI once an earlier TBI was sustained, those individuals reporting exposure to a blast (n 363) were examined. No difference in rates of sustaining a probable TBI in a blast event were found in those with (n 66) or without (n 297) a history of a TBI, 15.2% versus 19.9%, respectively (P.38). DISCUSSION In this study, strong associations between military deployment of Guard members and presence of adverse health out-

6 Table 3: Adjusted ORs (95% CI)* for Each Postdeployment Health Outcome According to Deployment-Related Factors Current Health Deployment Physical Injury Traumatic Experiences Combat Intensity Blast Low Medium High Nonprimary Primary Overall health rating Poor overall health 1.58 ( ) 2.37 ( ) 1.46 ( ) 0.94 ( ) 1.42 ( ) 1.44 ( ) 1.18 ( ) 1.47 ( ) 2.52 ( ) Mental health conditions Major depression 0.94 ( ) 1.65 ( ) 1.04 ( ) 1.00 ( ) 0.61 ( ) 1.89 ( ) 2.08 ( ) 2.30 ( ) 4.89 ( ) Anxiety disorder 0.70 ( ) 1.73 ( ) 0.90 ( ) 0.72 ( ) 1.40 ( ) 3.17 ( ) 1.84 ( ) 2.35 ( ) 3.91 ( ) PTSD 0.40 ( ) 2.43 ( ) 1.84 ( ) 1.07 ( ) 2.09 ( ) 4.83 ( ) 2.04 ( ) 2.05 ( ) 2.63 ( ) Excessive drinking 0.75 ( ) 1.28 ( ) 1.36 ( ) 1.44 ( ) 1.49 ( ) 1.60 ( ) 1.07 ( ) 1.26 ( ) 1.17 ( ) PCS complex 0.70 ( ) 2.61 ( ) 1.64 ( ) 1.40 ( ) 2.31 ( ) 3.01 ( ) 1.57 ( ) 1.66 ( ) 2.47 ( ) Health conditions Hypertension 1.66 ( ) 1.15 ( ) 1.19 ( ) 0.73 ( ) 0.61 ( ) 0.64 ( ) 1.19 ( ) 1.21 ( ) 1.53 ( ) Allergies 1.43 ( ) 1.47 ( ) 1.13 ( ) 0.63 ( ) 0.76 ( ) 0.89 ( ) 0.92 ( ) 1.02 ( ) 0.69 ( ) Arthritis 1.77 ( ) 2.66 ( ) 1.07 ( ) 0.80 ( ) 0.76 ( ) 1.01 ( ) 1.41 ( ) 1.16 ( ) 1.29 ( ) Non PCS Chest pain 1.01 ( ) 2.06 ( ) 1.92 ( ) 1.29 ( ) 1.14 ( ) 1.40 ( ) 1.53 ( ) 1.39 ( ) 1.00 ( ) Gait problems 1.09 ( ) 4.77 ( ) 1.08 ( ) 0.76 ( ) 1.83 ( ) 2.04 ( ) 0.46 ( ) 0.97 ( ) 2.33 ( ) Indigestion 1.05 ( ) 1.38 ( ) 1.81 ( ) 1.57 ( ) 2.15 ( ) 2.46 ( ) 0.99 ( ) 1.65 ( ) 1.56 ( ) Back & muscle pain 1.48 ( ) 1.80 ( ) 1.32 ( ) 1.46 ( ) 1.72 ( ) 1.77 ( ) 1.34 ( ) 1.64 ( ) 1.36 ( ) Abdominal pain 1.50 ( ) 1.16 ( ) 1.49 ( ) 0.71 ( ) 0.97 ( ) 1.12 ( ) 1.62 ( ) 2.28 ( ) 1.24 ( ) Shortness of breath 0.92 ( ) 1.93 ( ) 1.52 ( ) 1.22 ( ) 1.62 ( ) 1.22 ( ) 2.05 ( ) 2.36 ( ) 1.18 ( ) Pain on deep breath 0.82 ( ) 1.47 ( ) 1.40 ( ) 1.05 ( ) 1.34 ( ) 1.24 ( ) 2.58 ( ) 3.25 ( ) 1.42 ( ) PCS Headaches 1.18 ( ) 1.84 ( ) 1.69 ( ) 1.16 ( ) 1.59 ( ) 1.61 ( ) 1.40 ( ) 1.37 ( ) 2.20 ( ) Dizziness 1.25 ( ) 1.90 ( ) 1.43 ( ) 0.54 ( ) 1.01 ( ) 1.87 ( ) 1.70 ( ) 2.26 ( ) 2.52 ( ) Fatigue 1.17 ( ) 2.21 ( ) 1.43 ( ) 1.13 ( ) 1.24 ( ) 1.77 ( ) 2.04 ( ) 1.85 ( ) 1.90 ( ) Insomnia 1.14 ( ) 1.97 ( ) 1.35 ( ) 1.58 ( ) 1.53 ( ) 1.99 ( ) 1.63 ( ) 1.61 ( ) 1.77 ( ) Balance problems 0.50 ( ) 1.73 ( ) 2.81 ( ) 0.54 ( ) 1.65 ( ) 3.10 ( ) 1.21 ( ) 2.21 ( ) 2.54 ( ) Memory problems 0.78 ( ) 3.12 ( ) 1.44 ( ) 1.56 ( ) 2.32 ( ) 2.50 ( ) 1.87 ( ) 1.62 ( ) 2.34 ( ) Concentration problems 0.97 ( ) 2.22 ( ) 1.39 ( ) 1.05 ( ) 1.87 ( ) 1.86 ( ) 1.63 ( ) 1.80 ( ) 2.64 ( ) Irritability 0.68 ( ) 1.90 ( ) 1.51 ( ) 1.57 ( ) 2.34 ( ) 3.30 ( ) 1.62 ( ) 1.90 ( ) 2.10 ( ) Hearing loss 1.47 ( ) 1.84 ( ) 1.36 ( ) 1.09 ( ) 1.20 ( ) 1.54 ( ) 1.63 ( ) 2.32 ( ) 1.18 ( ) Ringing in ears 1.21 ( ) 1.86 ( ) 1.23 ( ) 1.23 ( ) 1.73 ( ) 1.89 ( ) 1.77 ( ) 2.92 ( ) 1.58 ( ) *Adjusted analyses were conducted by logistic regression to calculate the OR for various covariate predictor variables, controlling for all other covariates, as well as demographic and predeployment factors. Reference group for combat intensity was no combat; reference group for blast was no blast. All psychiatric diagnostic categories should be considered probable as they were based on self-report to standardized questionnaires. PCS Complex denotes International Statistical Classification of Diseases and Related Health Problems, 10th Revision, symptom criteria based on the Neurobehavioral Symptom Inventory at least 1 physical symptom, 1 cognitive symptom, 1 emotional symptom, and sleep problems, all at a moderate level of severity or higher. All dependent variables (self-reported current physical health problems or symptoms after deployment) are shown in the first column. P.001. P.005. P.01. Mild TBI 1892 MILITARY DEPLOYMENT AND HEALTH OUTCOMES, Vanderploeg

7 MILITARY DEPLOYMENT AND HEALTH OUTCOMES, Vanderploeg 1893 Table 4: Blast Injury Mechanisms Associated With Blast-Related Mild TBI Mechanism of Injury Blast-Related Mild TBIs, n (%) 1 6 (8.7) (14.5) (8.7) (1.4) (29.0) (10.1) (27.5) NOTE. 1 Primary (ie, blast wave). 2 Secondary (ie, hit by debris). 3 Tertiary (ie, thrown against something). 4 Quaternary (ie, building collapse, burn, or inhalation injuries). comes were related to specific aspects of the deployment experience and not to deployment, per se. Most of the observed postdeployment health outcomes had associations with multiple deployment-related factors. As with other recent studies, 3,4,10,23-25 strong associations were found between incurring probable mild TBI during deployment and postdeployment probable depression, anxiety, and PTSD. In the case of PTSD, having experienced high levels of combat exposure or physical injuries during deployment was concomitantly associated with postdeployment probable PTSD. Others have hypothesized that mild TBI initially impairs information processing and executive inhibitory control functions. As a result, individuals are conjectured to have difficulty cognitively processing the initial trauma experience and inhibiting intrusive recollections, thereby strengthening associations and setting the stage for PTSD development. 23,26 In contrast, the current findings can be interpreted as reflecting a common pathway of perceived threat to one s physical integrity or life to the development of PTSD. That is, the deployment-related events found to be associated with probable PTSD physical injury, high levels of combat, and probable TBI are each potent psychological traumas that likely are experienced as a threat to one s physical integrity or life. However, this threat to one s physical integrity explanation does not account for the strong association between probable mild TBI and both probable depression and generalized anxiety, where mild TBI was the only significant factor occurring during deployment. Therefore, these findings may be suggestive of long-term central nervous system dysfunction, though clearly additional studies and longitudinal data are needed. Findings from the current study are consistent with multiple reports in which PCS are not specific to TBI 3,10 ; both mild TBI and PTSD have been found to be independently associated with at least some postconcussive symptoms. 3,4,8,10 In the current study, postconcussive symptoms were associated with a variety of deployment experiences physical injury, combat, and blast exposure, as well as a probable mild TBI. Some prior studies have conflated what may well be deployment-related etiologic mechanisms (eg, mild TBI, physical trauma, and combat) with potentially separate postdeployment health outcomes (eg, PTSD, depression, and postconcussive symptoms). For example, both depression and PTSD are potential consequences of war and both have been suggested as mediators between the presence of mild TBI and PCS. 3 Examining the variance in PCS associated with mild TBI while controlling for PTSD symptoms is arbitrary and obscures other potential explanations. PTSD and depression are both health outcomes, and both are associated with a variety of symptoms that overlap with PCS. Therefore, controlling for PTSD or depression will inevitably minimize associations between PCS and mild TBI. 3,5,11 Given the overlap in symptoms across various health outcomes, it should not be surprising that such outcomes may be more highly associated with each other than they are with possible deployment-related etiologies (eg, mild TBI or combat). 3,4,11 The current study attempted to clearly distinguish between possible deployment-related etiologic mechanisms and potentially overlapping postdeployment health outcomes and found mild TBI along with some other deployment-related experiences to be separately associated with both the PCS complex and most of the specific postconcussive symptoms examined. We also attempted to account for the possible confounding effects of any predeployment psychological trauma or TBI on current postconcussive symptoms. However, retrospective observational studies may never be able to completely disentangle these relationships. Other important mild TBI findings were statistically significant monotonic increases in the frequencies of probable depression, anxiety, PTSD, and PCS complex in those reporting multiple versus only a single mild TBI. However, a predeployment self-reported TBI did not increase the likelihood of sustaining another self-reported TBI in a blast, compared with those with no prior TBI. That is, having a prior mild TBI did not increase one s vulnerability to sustaining a subsequent TBI. This suggests that a prior TBI, at least in the current sample, did not lower the intraindividual threshold of physical force needed to incur a subsequent mild TBI. These findings may obviously differ according to the time between injuries. In the current sample, this time frame was unknown but was likely months to years. Another finding was that only 1% of the respondents reported sustaining a mild TBI from exposure to a primary blast wave without any other physical injury. A selfreported mild TBI from exposure to a primary blast wave is rare if no additional injuries are incurred. Unique to the current study were the observed associations between blast and specific health outcomes in a pattern suggestive of subtle residual barotrauma: abdominal pain, pain on deep breathing, hearing loss, shortness of breath, ringing in the ears, and dizziness, especially if the Guard member experienced primary blast exposure. Although additional deployment-related factors were associated with many of these symptoms, the blast-related associations are consistent with damaged air or fluid-filled organ systems (ie, abdomen, lung, gastrointestinal tract, and ear) in which primary blast waves are known to cause injuries Relatively fewer associations were found between the experience of seeing others wounded or killed or experiencing the death of a buddy or leader and postdeployment adverse health conditions, and no associations were found with probable depression, anxiety, or PTSD. This is somewhat surprising and may be due to how this psychological trauma exposure variable was operationally defined. Respondents were not asked if they Table 5: Cumulative Effect of Mild TBI on Mental Health Outcomes Current Mental Health Condition Single TBI Multiple TBIs Major depression 2.55* ( ) 4.73* ( ) Anxiety disorder 2.72 ( ) 4.02 ( ) PTSD 2.41 ( ) 3.44* ( ) PCS Complex 1.81 ( ) 3.29* ( ) NOTE. Values are OR (95% CI). *P.005. P.001. P.01.

8 1894 MILITARY DEPLOYMENT AND HEALTH OUTCOMES, Vanderploeg felt psychologically traumatized. Instead they were asked about exposure to events that may be experienced as psychologically traumatic. It may be that their training helped prepare them for such exposures. Future research should inquire specifically about the personal experience of psychological trauma and broaden this psychological trauma construct to other potential traumas. Despite this limitation, indigestion and headaches were associated with psychological trauma exposure. These findings suggest that psychological factors may play a significant, but not exclusive, role in postdeployment pain and gastrointestinal complaints in this sample. Study Limitations The principal limitation of this study is the cross-sectional design based on self-report of deployment experiences and current health outcomes. Causality can only be inferred, and reporting bias is possible. However, the temporal sequence of the variables assessed in the data (eg, deployment-related experiences preceding current health status) helps to minimize this limitation. Another limitation is that some of the elements of the survey were developed specifically for this study by the investigators and were not previously validated. These variables included self-report of physical health conditions (ie, hypertension, arthritis, and allergies), physical injuries sustained, and non postconcussive symptoms. The assessment of possible TBI and blast exposure was also developed specially for this study, and the combat intensity variable was also novel. Similarly, in this study psychological trauma was defined and measured in an overly constrictive manner. Deploymentrelated psychological trauma was defined as exposure to seeing others wounded, killed, or dead, or experiencing the death of a close friend or the loss of a valued and trusted leader. Clearly, there are many additional types of psychological trauma exposure that may be experienced in a war zone. Furthermore, trauma severity was not assessed. The diagnostic classifications used in the current study can only be considered probable or screening cases because they were assessed with self-report screening tools, not clinically based diagnostic interviews. In addition, the recency of the deployment-related trauma events (eg, mild TBI and blast exposure) was not taken into account, although in virtually all cases the time since trauma would be considered postacute or chronic (ie, more than 3mo since the event). However, analyses of primary outcomes were rerun using time since deployment as an additional predictor variable. Time was not significantly related to any outcome, and its inclusion did not change any of the reported findings. Another limitation is that no information was collected regarding events since deployment that may have had an impact on the current findings, such as subsequent TBIs. However, postdeployment adverse events would likely be randomly distributed across the deployed sample, and therefore unlikely to change the associations reported in this study. Finally, because of the anonymous nature of the survey, we were able to collect only limited information about respondents. In addition, information was unavailable from the Florida National Guard on their cohort for comparative purposes. Thus, no information was available to help determine whether respondents differed from nonrespondents, and the response rate was somewhat low. However, the available sample demographic characteristics and rates of combat experience of the subjects are consistent with those in other studies. 4,5,30-32 In addition, those findings that do overlap with previous studies were consistent with but also extend the existing literature 3-6,9-11 because potential deployment-related etiologies are not confounded with postdeployment outcomes. Therefore, the findings are likely to apply to deployment experiences and traumas across units deployed to Iraq and Afghanistan. However, it is important to recognize that deployment-related experiences and traumas diverge across different phases of conflict, military occupation or role, service branch (again because of different military roles), and other factors. As a result, although the relationships between various deployment experiences and traumas and subsequent postdeployment outcomes likely would be similar across cohorts, the frequencies and the intensity of deployment experiences and injuries is not likely to generalize. CONCLUSIONS Complex relationships were found between multiple deployment-related mechanisms and numerous overlapping and cooccurring subsequent adverse health conditions. This multifactorial pattern suggests that an integrated interdisciplinary health care approach would be beneficial, such as the interdisciplinary system of care currently being used within the Veterans Health Administration Polytrauma programs. 33,34 At most sites, polytrauma programs are also integrated with postdeployment primary care clinics utilizing a medical home model of care. 35,36 Both programs include embedded mental health professionals. Such programs are best situated to assess, treat, and manage these complex postdeployment conditions. Acknowledgments: We thank Audrey Nelson, PhD, RN, FAAN, Peter A. Toyinbo, MBChB, PhD, CPH, William A. Lapcevic, MSST, MPH, and Britta Neugaard, PhD, MPH, for their invaluable assistance with the original grant application, data collection, data management, and statistical consultation and the Florida National Guard Adjutant General, State Surgeon, Deputy State Surgeon, and the Guard members for their participation and service. References 1. Hoge CW, Castro CA, Messer SC, McGurk D, Cotting DI, Koffman RL. Combat duty in Iraq and Afghanistan, mental health problems, and barriers to care. N Engl J Med 2004;351: Jones E, Hodgins-Vermaas R, McCartney H, et al. Post-combat syndromes from the Boer war to the Gulf war: a cluster analysis of their nature and attribution. BMJ 2002;324: Hoge CW, McGurk D, Thomas JL, Cox AL, Engel CC, Castro CA. Mild traumatic brain injury in U.S. Soldiers returning from Iraq. N Engl J Med 2008;358: Schneiderman AI, Braver ER, Kang HK. Understanding sequelae of injury mechanisms and mild traumatic brain injury incurred during the conflicts in Iraq and Afghanistan: persistent postconcussive symptoms and posttraumatic stress disorder. Am J Epidemiol 2008;167: Polusny MA, Kehle SM, Nelson NW, Erbes, CR, Arbisi PA, Thuras P. Longitudinal effects of mild traumatic brain injury and posttraumatic stress disorder comorbidity on postdeployment outcomes in National Guard Soldiers deployed to Iraq. Arch Gen Psychiatry 2011;68: Tanielian T, Jaycox LH, editors. Invisible wounds of war: psychological and cognitive injuries, their consequences, and services to assist recovery. Santa Monica: RAND Corporation; Terrio H, Brenner LA, Ivins BJ, et al. Traumatic brain injury screening: preliminary findings in a US Army Brigade Combat Team. J Head Trauma Rehabil 2009;24: Jones E. Historical approaches to post-combat disorders. Philos Trans R Soc Lond B Biol Sci 2006;361: Brenner LA, Ivins BJ, Schwab K, et al. Traumatic brain injury, posttraumatic stress disorder, and postconcussive symptom reporting among troops returning from Iraq. J Head Trauma Rehabil 2010;25: Vanderploeg RD, Belanger HG, Curtiss G. Mild traumatic brain injury and posttraumatic stress disorder and their associations with health symptoms. Arch Phys Med Rehabil 2009;90: