Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
5 MAJOR COHORT STUDIES This chapter provides an overview of the major cohort studies identified by the committee that examined long-term health outcomes related to traumatic brain injury (TBI). The studies are categorized by population, including military and veteran populations, general population, and people who sustained sports-related TBIs. Most major cohorts, once established, led to additional health outcome studies, which the committee refers to as derivative studies. Table 5.1, at the end of this chapter, provides information on each original cohort study, including the study design, the recruitment method, the eligible population, the study population, and the percentage of subjects who were enrolled. Information on the derivative studies appears in the table under the information on the original cohort studies from which they drew their populations and includes purpose, design, enrollment of subjects, sample size, response rates, and other characteristics if provided by the study authors. The information on derivative studies helped the committee to identify the populations that have been studied frequently and to understand which studies were independent of each other so that it could avoid evaluating studies of health outcomes in the same population repeatedly. Studies discussed below might not be included as primary studies in Chapters 6 through 10, because they did not meet the committeeâs strict criteria for primary studies. However, many of those not included as primary studies have been included as secondary studies and informed the committeeâ decisions about the long-term consequences of TBI. GENERAL LIMITATIONS OF COHORT STUDIES A number of limitations were encountered when the committee was reviewing the studies that are detailed below. Among them are self-reporting of exposure and health outcomes, lack of representativeness, selection bias, and failure to include a reference or control population. Many of the cohort studies relied on self-reporting of exposure (such as TBI, concussion, and loss of consciousness) and outcomes (such as headache and memory problems) rather than clinical evaluation or medical-record review. Self-reporting of exposure in retrospective cohort studies introduces the possibility of recall bias, the tendency for participants who have an outcome to overestimate (or underestimate) their exposure. That can limit the usefulness of study findings because outcomes may not necessarily be attributable to the exposure in question (TBI). 117
118 GULF WAR AND HEALTH Self-reporting of outcomes can introduce reporting bias. Reporting bias, which occurs when the group being studied reports more frequently what it remembers than a comparison group, can potentially lead to an overestimation of the incidence or prevalence of symptoms or diagnoses in the exposed populations. Self-reporting of outcomes based solely on symptoms might also introduce misclassification bias, in which there are errors in how symptoms are classified into outcomes. Low participation rates, which can introduce selection bias, can severely limit the ability to generalize study results because the study population may not be representative of the larger population to which the results are meant to be generalized. A related issue is the use of inappropriate controls, such as comparison of military populations with civilian populations; military personnel may be healthier than the general population, so the two populations may be noncomparable. That is referred to as the healthy-warrior effect; there may have been nonrandom assignment of those selected and not selected for participation in the military. It is possible to measure the potential for such biases and to adjust for them in the analysis. Another important limitation of some of the cohort studies is that they lack unexposed control groups. An unexposed group is a necessary component of a well-designed cohort study because it permits comparisons of rates of disease between exposed and unexposed populations and understanding of how an exposure affects the incidence of an outcome. Some of the studies discussed below are registries of participants who presented for care. These studies are not intended to be representative of the symptoms and diagnoses of an entire population. Although this is not necessarily a limitation, many studies discussed below were not designed with the committeeâs research question in mind. It was therefore difficult to use their findings to assess the broader question of the relation of long-term health outcomes to TBI. ORGANIZATION OF THE CHAPTER This chapter has sections on military cohort studies, population-based studies, other cohort studies, and sports-related studies. For each major cohort study, the methods for selecting the study population, the outcomes assessed, and the general findings are discussed. The committee was most interested in studies of long-term health outcomes related to TBI in military and veteran populations, so this group of studies is given primary consideration below. MILITARY STUDIES Studies of TBI have been conducted in nearly all the major conflicts of the 20th century, including World Wars I and II, the Korean War, and the Vietnam War; many of the studies evaluated seizure as the outcome of interest. Meirowsky (1982) noted that studying military populations âoffers the advantage of similarity in age and general health of the subjects at the time of injury and the relative ease with which they can be followed in subsequent years.â The committee paid particular attention to studies that assessed TBI in military populations because these were generally long-term prospective assessments of the population of interest.
MAJOR COHORT STUDIES 119 E. A. Walkerâs Studies of Head-Injured Bavarian World War I Veterans Walker and Erculei (1970) examined a cohort of head-injured Bavarian World War I veterans. The veterans were patients at a medical center established in Munich in 1916 for head injuries. Medical records, including field medical records and neurology reports on 5,500 men who had sustained head injuries, were reviewed in 1964â1966. The records also included information on the men for up to 50 years after injury. About 1,000 records were randomly selected from the 5,500, and death certificates were sought from social-welfare offices in Bavaria and West Germany. Vital statistics were obtained for about 600 of the 1,000 men; the remainder could not be located. Controls were about 600 uninjured Bavarian World War I veterans. Men who were born before 1880, who died before the age of 35 years, or whose dates of death were not known were excluded. Head injuries were diagnosed on the basis of demonstration of immediate posttraumatic neurologic disturbance or evidence of a contusion, laceration, or compound wound injury of the scalp. Posttraumatic epilepsy was diagnosed on the basis of absence of preinjury seizures and the occurrence of seizures at some time after injury. Seizures were verified by a physician, nurse, or family member; if no outside party could verify the seizure occurred, this was noted. The authors noted that most of the patients had their first seizure within a year after the injury and others many years after the injury (Walker et al., 1971). Walker and colleagues (1971) compared life expectancy of those with injuries and unwounded Bavarian veterans of World War I who had been awarded service medals carrying small pensions. The injured group had 1.8% more deaths than expected in the general male population, and a 4-year shorter life expectancy than the control group. In 1965, 73% of men at least 65 years old with TBI and 80% of those at least 65 years old without TBI were alive. Weiss et al. (1982) used the same data and found that in 1972, 497 (76.8%) of 647 TBI veterans and 483 (78.4%) of 616 of the control group had died. E. A. Walkerâs Studies of Head-Injured World War II Veterans Walker and Ercluei (1969) also conducted a cohort study of 364 severely head-injured World War II veterans 15 years after injury. Of these, 241 were originally studied at the Army Posttraumatic Epilepsy Center at Cushing General Hospital in Framingham, Massachusetts, in 1945â1946; these patients experienced at least one posttraumatic seizure. A battery of medical, psychologic, and electroencephalographic (EEG) tests were administered 1â3 years after injury; a 10-year followup consisted of examination, phone interview, or questionnaire. The authors reported that annual contact was made with nearly all subjects. The other 123, unselected head- injured men were studied as part of a followup in Baltimore from 1950 to 1954 and identified through Army and Veteransâ Administration (VA) pension rosters; the population was comparable with the Cushing General Hospital group in severity of injury. Medical records were not as detailed and complete as those on the group described previously. Neurologic, social, psychometric, and EEG tests were administered 6â9 years after injury. In general, the study participants had more severe brain wounds than would typically be seen in civilian or military hospitals. Dural penetrating frontal wounds tended to be included in the series although occipital and temporal injuries tended to be excluded (Walker and Erculei, 1969); dural penetration was found in 87% of the Cushing General Hospital group and in 71% of the Baltimore group (Walker and Erculei, 1970).
120 GULF WAR AND HEALTH The authors contacted participants by mail, and information was obtained through interviews and examination. They collected data on time from injury to examination, socioeconomic and work status, clinical symptoms, state of cranium and scalp, neurologic examination, epileptic status, EEG examinations, and psychometric testing (Wechsler-Bellevue, Minnesota Multiphasic Personality Inventory, Goddard Form Board, and McGill Picture Anomaly Test). Information was obtained on 343 (94%) of the 364 men originally identified; 31 died, leaving 313 living patients. 2 Of the 313, 243 (78%) were examined in person (Walker and Erculei, 1969). The authors assessed a variety of outcomes related to head injury, including neurologic symptoms (nervousness, headache, and other posttraumatic symptoms, such as nervousness and headache), posttraumatic epilepsy, employment status and other social-function outcomes, and psychologic outcomes. Of the 313 men, 212 (68%) reported some form of nervousness, from mild to severe, and 200 (64%) reported that they had experienced headaches. The authors also assessed posttraumatic syndrome, defined as a complex of symptoms that follow a minor head injury, including dizziness, nervousness, anxiety, and emotional instability. Of the 306 veterans on whom information was available, 34 (11%) reported no complaints, 65 (21%) had isolated symptoms, and 207 (68%) had posttraumatic syndrome (Walker and Erculei, 1969). Walker and Erculei (1969) also studied the prevalence of posttraumatic epilepsy. Two primary groups were defined for the analysis: 199 posttraumatic epileptic men from Cushing General Hospital and 114 men with posttraumatic encephalopathy from the Baltimore group who were matched by class of injury. Clinical examinations were conducted on 154 of those with epilepsy and 95 of those with encephalopathy. The authors found a statistically significantly lower survival rate in those with posttraumatic epilepsy. Patients with posttraumatic epilepsy were more likely to be unemployed (57%) than those with posttraumatic encephalopathy (14%). Similarly, neurologic deficits were more likely in the group with epilepsy than in the group with encephalopathy. Hemiplegia, for example, was found in 64% of the men with epilepsy and 40% of those with encephalopathy. Increased mental impairment was also noted in the epileptic group; in the memory tests, two errors or fewer were recorded by 85% of the posttraumatic- encephalopathy group and 66% of the epilepsy group. Walker and Erculei (1970) also assessed posttraumatic epilepsy and found that 15 years postinjury, 40% (n = 92) had no seizures of any time between the 5th and 15th year; 3% had no seizures from the 10th to 15th year. Twenty-three percent or 52 men continued to have 1 to 6 episodes annually; 68 had more than 6 episodes per year. In another analysis, Walker and colleagues (1969) assessed employment status after head injury in 303 subjects (nine were omitted because they were hospitalized during the followup assessment) and found that 182 (60%) men were regularly employed and 121 (40%) were unemployed or working irregularly. Neurologic deficits were also assessed (Walker and Erculei, 1969). Of the 249 men examined, 199 (80%) exhibited abnormality of neurologic function. Of those with abnormal neurologic function, hemiplegia was present alone or in combination with other findings in 118 (59%), hemianesthesia in 121 (61%), hemianopsia in 43 (22%), aphasia in 36 (18%), mental impairment in 13 (7%), and cranial nerve defect in 137 (69%). 2 Numbers as reported in study.
MAJOR COHORT STUDIES 121 Walker and Erculei (1969) found that few of the patients who had psychologic conditions in the early posttraumatic years recovered, 19 men developed mental abnormalities 10â15 years after injury, and severe mental disturbances occurred in only a small percentage of the patients. Finnish Studies Since 1948, the treatment, rehabilitation, and study of all head-injured Finnish war veterans have been monitored by one central institution, the Rehabilitation Institute for Brain- Injured Veterans. Achte and colleagues (1969), in an uncontrolled series, followed 3,552 men who suffered head injury in the Finnish wars of 1939â1945 to identify the prevalence of posttraumatic psychoses that developed up to 22â26 years after injury. On admission, mild injuries accounted for 19%, moderately severe for 59%, and severe for 22% of the sample; open head injuries were present in 42% of patients. Patientsâ initial medical records and examination and treatment records were reviewed to ensure the presence of a brain injury; questionable cases were excluded. In addition, moderate and severe injuries may have been overrepresented inasmuch as patients with mild traumas were often left untreated. Most patients were examined personally by the authors; otherwise, records of psychiatric treatment were obtained in addition to personal communication with the patients. Between the time of injury and 1966, 317 (8.9%) of the original population experienced at least one psychotic episode; schizophrenic psychosis was the most prevalent at 21.1% of the 317 (it appeared to be more frequent in the mildly injured and those under 20 years old), followed by paranoid psychosis (17.6%), epileptic psychosis (14.6%), and concussion psychosis (13.7%). Psychosis began in 24.0% of the cases less than 1 year, in 16.0% 1â5 years, in 17.7% 5â10 years, and in 42.3% over 10 years after injury. In a more inclusive series by Achte and colleagues (1991), roughly 10,000 Finnish veterans were followed for 50 years after brain injury. Of them, 2,907 suffered some type of psychiatric disturbance throughout their lives, 26% (762) of which were classified as psychotic. At the time of study publication, 251 of those veterans were assigned a detailed diagnosis with the following distribution: delusional psychosis, 28%; major depression, 21%; delirium, 18%; and paranoid schizophrenia, 14%. Delusional psychosis tended to develop 15â19 years after injury and persisted for over 5 years in 40% of cases; paranoid schizophrenia and schizophreniform generally had a shorter latencyâless than a year in 23% of casesâand persisted for over 5 years in 63% of cases. Teuberâs Cohort In the late 1940s and 1950s, Teuber and colleagues at the New York UniversityâBellevue Medical Center recruited and examined over 300 World War II veterans (and some from World War I and the Korean War) who lived near New York City and had sustained penetrating injuries of the brain or the peripheral nervous system (Weinstein, 1954). All the veterans in this longitudinal cohort study, originally identified from rosters maintained by the VA, incurred traumatic lesions in the service. Sensory, motor, and cognitive tests were administered to the veterans, first in Teuberâs New York laboratory and later by investigators at the Massachusetts Institute of Technology. In one early study by Teuber and Weinstein (1954), 35 brain-injured veterans and 12 controls with arm or leg peripheral nerve injury were selected for assessment of performance on a modified Seguin-Goddard formboard task by area of brain injury. The brain-injured veterans
122 GULF WAR AND HEALTH made significantly more errors, recalled fewer forms, and had greater variability in the time it took to place the correct form in the correct opening. The investigators did not observe greater impairment in those with frontal lobe injuries than in those with lesions in the parietal, temporal, or occipital lobes. In another study, Weinstein and Teuber (1957a) obtained preinjury scores on the Army General Classification Test (AGCT) for 62 men who subsequently sustained penetrating brain injuries and 50 controls who incurred nerve injuries of the arm or leg. Both groups of men were retested with a comparable form of the AGCT (First Civilian Edition). The preinjury scores of the two groups were similar: the mean score was 106.4 in the controls and 105.0 in the brain- injured group. Scores on the postinjury test showed some gains in the controls (48 of the 50 controls increased their mean scores to 119.4) while there was little or no change in the brain- injured menâs test scores. In the same sample of veterans, Weinstein and Teuber (1957b) reported that the findings were independent of any effects for differences in preinjury education and preinjury AGCT score. A study of roughness discrimination was also conducted with Teuberâs cohort (Weinstein et al., 1958), in which 43 veterans with brain injury were compared with 20 controls with leg peripheral nerve injuries. The study participantsâ task was to touch a patch of sandpaper and then attempt to find in a comparison array of 18 patches the one that was identical in roughness. Four sets of experiments were conducted: unilateral-successive for ipsilateral hand, unilateral- successive for contralateral hand, bilateral-successive, and bilateral-simultaneous. In all groups, there was a significantly lower average error in the unilateral experiments. However, there was a deficit in roughness discrimination in veterans who had sustained a penetrating brain injury. Under unilateral testing conditions, the left hand appeared to be more vulnerable to error than the right hand, regardless of the location of the brain injury. In more recent studies by Corkin et al. (1984, 1989), the investigators examined whether life expectancy or cognitive decline late in life is associated with having survived a penetrating brain injury. To study factors that might influence life expectancy, the authors evaluated 190 men who had sustained a brain injury during World War II and 106 men who had sustained peripheral nerve injuries during the war. Survival information was obtained from the VA, and the KaplanâMeier method was used to estimate cumulative survival distributions for the two groups. Although sustaining a penetrating head injury alone did not shorten life expectancy, the risk of death increased when it was coupled with posttraumatic epilepsy. As of May 1, 1983, mortality was 3.6 times higher in veterans with brain injury and epilepsy than in veterans with peripheral nerve injury. Corkin et al. (1989) also evaluated the interaction between aging and effects of brain injury in a similar series of veterans. To study whether head injury exacerbates cognitive decline in later years, the authors evaluated 57 World War II veterans with head injury and 27 with peripheral nerve injury matched on age, premorbid intelligence, and education. The participants received two timed cognitive tests: the ACGT (Total, Vocabulary, Arithmetic, and Block Counting) and the Hidden Figures Test, in which participants trace a simple geometric figure embedded in another geometric figure. On all five cognitive measures, the group with brain injury was statistically significantly inferior to the control group 10 years after injury. Over another 30-year period (that is, 40 years after injury), cognitive decline was observed in the brain-injury group on every AGCT measure except Vocabulary as compared to the control group (Corkin et al., 1989).
MAJOR COHORT STUDIES 123 W. F. Caveness Studies of Korean War Veterans In 1951, W. F. Caveness, then chief of the neurologic service in the US Naval Hospital, in Yokusaka, Japan, initiated a study of craniocerebral injuries in male military personnel wounded during or immediately after the Korean War. The participants were seen at Yokusaka or in the US Navy hospital ships off the coast of Korea. The investigators identified 467 cases in 1951â1954, many of them in Navy or Marine Corps personnel. During the initial period, investigators conducted a review of medical records, gave periodic physical examinations, distributed supplemental questionnaires and personal letters, and conducted interviews. The head injuries were categorized as related to missiles (resulting from contact with small-arms fire, grenades, land-mine mortar, or heavy artillery) or not related to missiles (resulting from blunt or sharp objects or vehicle accidents or secondary to blast); missile-related injuries accounted for more than half the injuries in the cohort (Caveness, 1963). A followup study was conducted 8â11 years after the initial injury. During the followup period, 356 of the original cases participated (76% of the total and 87% of those eligible for followup). Information was collected with a mailed questionnaire, a physical examination, interviews with the American Red Cross, and review of VA records. During the period 1957â 1958, additional VA information was available on 85% of the participants. Questionnaires were obtained in 1961â1962 from 91% of the participants, personal letters from 22%, and telephone replies from 10% (Caveness, 1963). Additional studies of this cohort evaluated the prevalence of posttraumatic epilepsy as diagnosed by EEG. Seizures were diagnosed if they fulfilled the criteria for focal somatomotor, somatosensory, special sensory, or adversive seizures. Other less well-defined focal events, such as patterns attributed to the temporal lobe, were included if accompanied by other overt phenomena of seizures (Caveness, 1963). The investigators noted that generalized seizures, characterized by a loss in consciousness with or without bilateral motor expression, âwere recognized either as a progression from a focal onset, in conjunction with focal seizures, or as a principal expression of the convulsive disorder.â Evans (1962) evaluated the prevalence of seizures in 422 of the head-injured Korean War veterans 3â11 years after injury. The authors found the overall prevalence of seizures to be 19.7%. Those with penetrating head injuries had a prevalence of seizures of 32%, those with blunt head injuries 8%, and those with blast wounds 2%. Caveness and colleagues (1962) assessed the prevalence of seizures in five retrospective cohorts from three wars (World War I, World War II, and the Korean War). In a random sample of 407 cases from the Korean War 5 years after injury, 24.1% had seizuresâ35.1% of those with missile head wounds and 12.2% with blunt or blast wounds. Caveness (1963) also found that 8â9% of the 467 men initially included in the study population had seizures within the first 2 weeks. Of the 356 men followed up 8â11 years after injury, 109 (30.6%) had seizures. Forty-two percent of those with penetrating head wounds and 16.4% with blunt head wounds had seizures. There was no significant difference in seizure incidence between the total original group and those followed for 8â11 years.
124 GULF WAR AND HEALTH Vietnam Head Injury Study The Vietnam Head Injury Study (VHIS) is a long-term, prospective followup study of head-injured Vietnam veterans, originally organized by William Caveness at the National Institutes of Health (NIH). The ultimate goal of the study is to evaluate the long-term neuropsychologic and other health outcomes of patients with penetrating head injury to learn about the role of head injury in the etiology of dementia and posttraumatic epilepsy, mechanisms of motor and cognitive recovery, and functions of various regions of the brain. The initial registry phase, conducted during the Vietnam War, consisted of military physiciansâ entering demographic, injury, and outcome data on registry forms for about 2,000 head-injured soldiers who had survived the first week after sustaining injury. Data were entered during 1967â1970. Over 95% of the patients enrolled were male, had penetrating head injuries, and were 18â25 years old (Grafman, 2007). Phase I of the study was a detailed medical-records review conducted some 5 years after injury. At that time, the VHIS cohort consisted of 1,200 men who had either closed or penetrating head injuries. Field records and records of acute hospitalization, rehabilitation, and followup were available for all subjects. The VHIS cohort allowed tracking over long followup periods and included preinjury vocational and intelligence testing (National Naval Medical Center, 2008). Phase II, conducted primarily by Grafman and Salazar, was a collaborative effort of the VA, NIH, and the American Red Cross and consisted of a comprehensive inpatient evaluation at the Walter Reed Army Medical Center of 520 head-injured subjects from the original cohort of 1,200 and 85 matched normal controls who were evaluated in 1981â1984 (12â15 years after injury) (National Naval Medical Center, 2008). The controls were recruited from the VA files of non-head-injured soldiers who had served in Vietnam in the same years and were within the same age range as the head-injured soldiers. Many patients were lost to followup and were no longer receiving medical care or were not honorably discharged, because of behavioral changes related to their head injuries. Phase II was also used to identify these patients and refer them for appropriate medical care. During phase II, researchers conducted a number of tests of neurologic, motor, speech and language, and neuropsychologic outcomes. Phase II also identified veterans with specific lesions (such as orbitofrontal and dorsal frontal) or with particular cognitive or neurobehavioral deficits and studied the prevalence of posttraumatic epilepsy and cognitive function after head injury (National Naval Medical Center, 2008). Two such studies in phase II assessed seizures after head injury (Salazar et al., 1985; Swanson et al., 1995). Swanson and colleagues (1995) assessed interictal personality traits in 238 veterans who had developed seizure disorders and compared them with personality traits in 229 veterans with penetrating head injuries but without seizures and 84 uninjured controls. Of the 238 with seizure disorders, 39 had simple partial seizures, 59 had complex partial seizures, 76 had partial seizures with secondary generalization, and 64 had generalized seizures. The authors assessed history of psychiatric treatment, preinjury intelligence, brain-volume loss, seizure frequency, and duration of epilepsy. Statistically significant increases in interictal psychopathology were observed in the groups with complex partial seizures, partial seizures with secondary generalization, and generalized seizures (but not simple partial seizures) compared with controls. No group differences between groups with seizure types were found.
MAJOR COHORT STUDIES 125 In an evaluation of 421 veterans from the VHIS cohort, Salazar and colleagues (1985) found that 53% had posttraumatic epilepsy. The relative risk of epilepsy in the head-injured veterans was 580 times higher than that in the general age-matched population in the first 6 months after injury and fell to 25 times higher after 10 years. Hemiparesis (p = 0.03), organic mental disorder (p = 0.015), aphasia (p = 0.009), headache (p = 0.001), and visual-field loss (p = 0.015) were associated with seizures. The authors found that the incidence of posttraumatic epilepsy was 86% in patients who had residual aphasia. Of patients with seizures, 57% had attacks within a year and 25% 1â5 years after injury; in about 18%, the first seizure occurred more than 5 years after injury; and in 7%, the first seizure occurred 10 years or more after injury. Patients with frequent seizures in the first year after injury were more likely to have a longer duration of epilepsy (p < 0.001). Of all those who sustained head injuries, 28% had persistent seizures 15 years after injury. The major limitations of the study include its lack of a reference group. It is also unclear whether seizures occurred before head injury. Phase III, conducted from 2004 to 2006, examined the role of head injury in cognitive neuroplasticity of the aging brain, memory and amnesia, such neurologic disorders as epilepsy, and social functioning. Phase III testing included elective neuroimaging, such as positron- emission tomography, and quantitative EEG. Of the 520 patients in phase II, 182 (35%) participated in phase III, and 17 were newly recruited. Of the 85 controls in phase II, 32 (38%) participated in phase III, and 23 were newly recruited (Grafman, 2007). Two studies from phase III have been published: TBI and cognitive outcomes and TBI and posttraumatic stress disorder (PTSD). Raymont and colleagues (2008) studied the relationship of preinjury intelligence, brain-tissue volume loss, lesion location, demographic variables, and the role of genetic markers in long-term cognitive decline. They found that subjects with penetrating head injury had a greater degree of overall cognitive decline than controls. Preinjury intelligence was the most consistent predictor of cognitive outcomes. Koenigs and colleagues (2007) studied the relationship between TBI and PTSD and found a âreduced occurrence of post traumatic stress disorder . . . following ventromedial prefrontal cortex damage and the complete absence of PTSD following amygdala damage.â Phase IV will begin in 2015, about 45 years after injury. The VHIS will provide baseline premorbid and injury information that can be used to asses the effects of penetrating head injury on the development of a variety of neurologic disorders in old age, the rate of physical and cognitive decline, and the effects of various variables on performance data. The investigators will re-examine the patients on some of the tasks (including standardized tests and the Armed Forces Qualification Test) administered during the phase II evaluation to assess cognitive, mood, personality, and neurologic functions. Vietnam Experience Study The Vietnam Experience Study (VES) was a multidimensional health assessment that began with data collection from Vietnam-era veterans in the middle 1980s, about 16 years after discharge (Luis et al., 2003). The VES included four components: medical and psychologic examinations, mortality assessments, telephone interviews, and reproductive-outcome assessments (CDC, 1989). The eligible population consisted of male US Army veterans who first entered the military during the period January 1965âDecember 1971, served at least 4 months on active duty, served only one tour of duty, obtained a military occupational specialty, and
126 GULF WAR AND HEALTH achieved a pay grade no higher than E-5 (sergeant) on discharge. On the basis of those requirements, random sampling of military records found 9,324 men who had been members of the US Army and served a single tour in Vietnam and 8,989 who served elsewhere (CDC, 1989). From the total eligible population, 4,462 veterans were randomly selected. A comprehensive 3- day medical and psychologic evaluation was administered to ascertain what health-related events occurred from time of military discharge to the study date (Luis et al., 2003). Numerous studies were published on the basis of extracted data on the cohort; the three described below evaluate the effects of mild TBI. Luis and colleagues (2003) compared the prevalence of persistent postconcussion symptom complex (PPCSC) in veterans with and without a history of mild TBI. Of the 4,462 randomly selected veterans, 329 were excluded because they met criteria for PPCSC in the 10th edition of the International Classification of Diseases (ICD-10) or in the fourth edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV) but not both; 55 were excluded because of hospitalization, and 121 were excluded because data on them were incomplete. The remaining 3,957 were categorized as follows: 2,937 with no history of motor-vehicle accident (MVA) and no history of TBI, 488 with a history of MVA but no history of TBI, 323 with a history of TBI unaccompanied by loss of consciousness (LOC), and 209 with a history of TBI accompanied by LOC. Results indicated that the group that had had TBI with LOC had significantly greater odds of having PPCSC than the unexposed control group (odds ratio [OR], 2.84; 95% confidence interval [CI], 2.12â3.80). No significant difference was found between the group with a history of an MVA but no TBI and the group with TBI but no LOC. Multiple factors (demographic, psychiatric, and social support) accounted for 33% of the variance in PPCSC in participants with TBI (history of TBI with or without LOC), but significantly less variance was found in the no-TBI group (23.9%). Vanderploeg and colleagues (2007) used a cross-sectional cohort sample to examine the long-term psychiatric, neurologic, and psychosocial outcomes resulting from self-reported mild TBI. A subsample of 4,384 veterans (excluding 40 who were hospitalized after injury and 38 on whom data were incomplete) were categorized into three groups: no history of MVA and no history of TBI (normal control, 3,214, 73%), injured in an MVA but no history of TBI (MVA control, 539, 12.3%), and TBI with altered consciousness (mild TBI, 254, 5.8%); those who reported a TBI without LOC were excluded (n = 377, 8.6%). Age, education, enlistment General Technical Test scores, and medical and psychiatric conditions varied among the three groups and were statistically controlled for in later analyses. The mild-TBI group had higher odds of having depression than the normal control group (OR, 1.77; 95% CI, 1.13â2.78). Antisocial personality disorder was twice as prevalent in the veterans with mild TBI as in the normal control group, but this outcome probably reflects a risk factor for obtaining an injury, given the similar rates of preinjury conduct disorder. The odds of postconcussion symptoms (PCSs) were about doubled in patients with a history of mild TBI by both DSM-IV and ICD-10 criteria (OR, 2.00 and 1.80, respectively; 95% CI, 1.49â2.69 and 1.33â2.43, respectively). The odds of peripheral visual imperceptions were twice as high (OR, 1.98; 95% CI, 1.21â3.24) and of impaired tandem gait were about three times as high (OR, 2.93; 95% CI, 1.34â6.38). People with TBI had higher odds of being unmarried (OR, 2.01; 95% CI, 1.57â2.75) and higher odds of employment issues (OR, 1.89; 95% CI, 1.36â2.64), low income (OR, 1.88; 95% CI, 1.29â2.74), and self-reported disability (OR, 2.90; 95% CI, 1.63â5.15).
MAJOR COHORT STUDIES 127 Vanderploeg and colleagues (2005) used the same cohort to conduct a cross-sectional study of neuropsychologic outcomes. A full 15-measure neuropsychologic battery with neurologic measures of tandem gait and peripheral visual attention was administered. Results revealed no statistically significant difference in any of the measures among the three groups. In examining more subtle differences in attention, concentration, and memory, it was found that the mild-TBI group had significantly higher odds of being unable to continue the Paced Auditory Serial Addition Test (PASAT) than either of the two control groups (comparison with normal control group: OR, 1.32; 95% CI, 1.00â1.73; comparison MVA control, OR, 1.53; 95% CI 1.10â 2.13). With respect to working memory, the mild-TBI group had excessive proactive interference (comparison with normal control group: OR, 1.66; 95% CI, 1.11â2.47). PASAT continuation problems were associated with left-side visual imperceptions, and excessive proactive interference was associated with impaired tandem gait in the mild-TBI group. In another study of the VES cohort, Vanderploeg and colleagues (2003) conducted logistic regression analyses to survey long-term outcomes of work and marital status in people who had mild TBI and any pre-existing factors that may perpetuate the symptoms of the injury. The author notes that the subsample (after exclusion of 53 people because they were hospitalized and 87 because data on them were incomplete) consisted of 4,322 people: 626 (14%) who had a mild TBI (373 without LOC, and 253 with LOC) and 3,896 (86%) who did not have a TBI. Psychiatric disorders were assessed with the Diagnostic Interview Schedule (DIS-II-A), and psychosocial outcomes were gathered by trained examiners. Results indicate that the outcome of a mild TBI may be influenced by the presence of any pre-existing demographic, medical, or psychiatric factors. Factors associated with work and marital status accounted for 23% and 17%, respectively, of the variance in those with head injury. Variance was significantly lower in those without head injury: 13.6% and 9.4%, respectively. POPULATION-BASED STUDIES Rochester Epidemiology Project The Rochester Epidemiology Project is a medical-records-linkage system that encompasses detailed health-care information on residents of the City of Rochester and Olmsted County, Minnesota. Funded initially in 1966 with medical records dating back to 1910, the project was designed to link all medical data and clinical information developed by the Mayo Clinic with data obtained by community health providers, including Olmsted Medical Group, the Olmsted Community Hospital, the University of Minnesota Hospital, and the Minneapolis VA Medical Center. Each patient was assigned a unique identifier, and information on all medical visits has been recorded for each patient. The database includes thorough medical histories, clinical assessments, consultation reports, surgical procedures, laboratory and radiology results, death certificates, and autopsy reports (Flaada et al., 2007). The medical information is continuously updated into an electronic format. By maintaining complete medical histories, the Rochester Epidemiology Project provides the capability to conduct population-based studies of disease risk factors and health outcomes and can be used to study long-term secular trends in disease incidence (Melton, 1996). As of 1996, the project included medical records on a population with more than 3.6 million person-years of experience in 1950â1995 (Melton, 1996). The demographic characteristics of Olmsted County residents largely resemble those of the US white population (Melton, 1996). Over 1,500 publications have resulted from the project,
128 GULF WAR AND HEALTH including studies related to TBI and a variety of health outcomes, such as seizures, Alzheimer disease, and Parkinson disease. A number of studies evaluated the incidence of seizures after head trauma in this population. In 1980, Annegers and colleagues conducted a retrospective population-based cohort study to determine the risk of posttraumatic seizures in people in the Rochester Epidemiology Project who had sustained head injuries. The authors found that the risk of posttraumatic seizures after severe injury was 7.1% within 1 year and 11.6% within 5 years. In 1995, Annegers and colleagues assessed the incidence of seizures in the same population, including cases in 1935â 1984. The age-adjusted incidence from 1955 to 1984 was 39/100,000 person-years. The age- adjusted incidence was higher in men (52.0) than in women (29.5). A third study conducted by Annegers et al. (1998) evaluated seizure cases in 1975â1984 with a followup period through 1994. The overall standardized incidence ratio (SIR) of seizures in patients with brain injury was 3.1 (95% CI, 2.5â3.8). The SIR was 1.5 (95% CI, 1.0â2.2) in those with mild TBI, 2.9 (95% CI, 1.9â4.1) in those with moderate TBI, and 17.0 (95% CI, 12.3â23.6) in those with severe TBI. Studies of a number of neurodegenerative diseasesâincluding Alzheimer disease, Parkinson disease, and multiple sclerosis (MS)âwere also conducted. Bower and colleagues (2003) assessed the risk of Parkinson disease in the head-injured people and found that for any head trauma the odds of developing Parkinson disease were 4.3 (95% CI, 1.2â15.2); subjects who had a mild head injury with LOC or a more severe trauma had odds of 11.0 (95% CI, 1.4â 85.2). Williams and colleagues (1991) assessed the risk of Alzheimer disease, dementia, and Parkinson disease in the same head-injured population and found that the standardized morbidity ratio was 1.00 (95% CI, 0.63â1.50) for Alzheimer disease, 1.06 (95% CI, 0.74â1.46) for dementia, and 0.94 (95% CI, 0.38â1.94) for Parkinson disease. Siva and colleagues (1993), using data from the Rochester Epidemiology Project, identified an incidence and prevalence cohort with MS, a head-injury cohort, and a lumbar-diskâsurgery cohort to evaluate the association between mechanical trauma and MS onset or exacerbation. Of a cohort of 819 people with head injury in Olmsted County, none developed MS within 6 months of the trauma. Of a lumbar-diskâ surgery cohort of 942 local residents, five developed MS, but onset of the disease preceded the spinal surgery in four of the five. The authors found no association between head injury or lumbar-disk surgery and the onset of MS. Brown et al. (2004) carried out a study to determine whether mortality from TBI was affected by the severity of the injury and the survival time by using medical records from the Rochester Epidemiology Project. There were 68 deaths in the moderate-to-severe TBI group for a KaplanâMeier estimated 30-day case-fatality rate of 29.3%. During the followup period, the 68 deaths demonstrated a significantly higher risk of mortality as compared to the 1990 Minnesota white population (risk ratio, 5.29; 95% CI, 4.11â6.71); but in people with moderate or severe TBI who survived for 6 months after injury, the risk ratio for death was not significantly greater than the expected (14 deaths vs 12.7 expected; risk ratio, 1.10; 95% CI, 0.60â1.85). Findings in those with mild TBI were similar. Community-Based Study of Injuries in the Aquitaine, France A population-based prospective study was conducted over a 1-year period from 1985 to 1986 in the Aquitaine, France, to assess the incidence, causes and severities of injuries. The Aquitaine, one of the 22 administrative regions in France, contains both urban and rural areas.
MAJOR COHORT STUDIES 129 The population in 1986 (extrapolated from the 1982 census) was estimated at 2.7 million. The cases were people with unintentional or intentional injuries sustained from December 1985 to December 1986 that resulted in death or required hospitalization. Public and private hospitals participating in the study were asked to complete a questionnaire for each injury case that required admission. The questionnaire included queries about demographic information and time and place, cause, origin, and clinical nature of the injury. Causes and clinical nature of the injuries were coded according to a classification system designed specifically for the study. All 21 public hospitals and 38 of the 43 private hospitals contributed to the study (Tiret et al., 1989). From 1985 to 1986, there were 391 deaths and 2,116 hospital admissions due to head trauma; the case-fatality rate was 4%. In the nonfatally head-injured patients, 80% of the head injuries were classified as mild, 11% moderate, and 9% severe (Tiret et al., 1990). The overall annual incidence was 281 per 100,000, and the annual mortality was 22 per 100,000 (Tiret et al., 1990). Masson and colleagues (1996) assessed the effects of cognitive, behavioral, and somatic impairments on disability and recovery after TBI. The study population included 231 TBI patients 5 years after injury and 80 lower-limbâinjured controls. Sixty-four lower-limbâinjured patients and 176 TBI patients were assessed. The severity of the head injuries was defined as severe if a patient had a Glasgow Coma Scale (GCS) score of 8 or less for at least 6 hours in the first 24 hours after injury; moderate if the patients had one of the following, a GCS score of 8 or less between 1 and 6 hours, a GCS score of 9â12 on the first day after injury, an abnormal computed tomography scan, or a need for a neurosurgical procedure; and minor if neither of those categories was appropriate. A number of complaints were more commonly reported in TBI patients than in the lower-limbâinjured patients, such as headache (OR, 4.6), memory problems (OR, 4.01), dizziness (OR, 3.35), anxiety (OR, 6.11), and sleep disturbance (OR, 3.10). Regarding mild, moderate, and severe TBI, there was no significant difference in the prevalence of headaches (44%, 54%, 44%), anxiety (47%, 49%, 63%), and dizziness (33%, 37%, 26%) among the three TBI severity groups, respectively. Mental impairments were reported frequently in patients with severe TBI (18â40% of patients); however, most impairments in patients with minor and moderate TBI were related to associated injuries. Masson and colleagues (1997) conducted a study to assess long-term disabilities related to TBI in 407 patients. The authors found that 5 years after injury 64 of the patients had died and 36 were lost to followup. Of those who sustained severe head injury, 56% died, and 50% of the survivors were disabled. The authors note that âhead injuries induce long-lasting handicap in 9 per 100,000 habitants which is severe in 2 per 100,000.â Canadian Study of Health and Aging The Canadian Study of Health and Aging (CSHA) is a population-based cohort study designed to assess the prevalence and incidence of dementia and risk factors for it in the Canadian population. Planned originally in 1989, the CSHA includes people 65 years old and older sampled from 36 communities around the country. The study population was selected to be representative of the general population (Lindsay et al., 2002). The study was conducted in three phases: CSHA-1 in 1991â1992, CSHA-2 in 1996â 1997, and CSHA-3 in 2001â2002 (studies of which were not directly applicable to this report). Initial contact with study participants was made by telephone, and they were asked questions
130 GULF WAR AND HEALTH about current health issues. Participants who were living in long-term-care institutions were given a clinical examination. Those living in the community were interviewed in their homes and then screened for dementia. A community interview was held to discuss general health issues, disability, and the presence of chronic ailments. Screening tests for cognitive impairment were administered, including the Modified Mini-Mental State Examination. A person who screened positive for impairment was asked to participate in a clinical examination; independent diagnoses were also made by a physician and a neuropsychologist. The clinical examination included neuropsychologic tests and medical assessment to permit a preliminary diagnosis of dementia. The diagnoses were the basis of the incidence and prevalence rates and were used in followup studies to identify risk factors. Followup studies included the same diagnostic criteria; cases were reassessed to be diagnosed according to new criteria based on the DSM-IV. In CSHA-1, 10,263 participants were identified; 9,008 lived in the community, and 1,255 lived in institutions. Participants who were included were interviewed about health issues and limitations in performing basic daily-life activities as assessed by the Older Americansâ Resources and Services Activities of Daily Living Scale (Lindsay et al., 2002). A caseâcontrol study was conducted to assess risk factors associated with Alzheimer disease or other dementias (CSHA, 1994). The participants were given a questionnaire that included questions about family and medical history, behavior, occupational and environmental exposures, and lifestyle. To avoid recall bias, questionnaires for participants with dementia were given to persons who knew them well. The study included 258 patients with probable Alzheimer disease, 129 with vascular dementia, and 535 normal controls. All participants who did not have cognitive impairment during this first phase were asked to complete a risk-factor questionnaire; these included 6,628 participants without cognitive impairment. A caseâcontrol study based on data from CSHA-1 (CSHA, 1994) found that age, family history of dementia, educational level, arthritis, and use of nonsteroidal anti-inflammatory drugs were significantly related to Alzheimer disease. In 1996â1997, the subjects who had agreed to participate in CSHA-2 were interviewed to determine changes in health status during the prior 5 years and underwent a process similar to that in CSHA-1. During CSHA-2, investigators conducted a nested caseâcontrol study of 194 participants who had recently diagnosed Alzheimer disease and 3,894 controls. For vascular dementia, 105 newly diagnosed cases were compared with controls (Lindsay et al., 2002). The authors found that increasing age, fewer years of education, and the apolipoprotein e4 (APOE 4) allele were significantly associated with increased risk of Alzheimer disease. However, they found no statistically significant association between Alzheimer disease and family history of dementia, history of depression, sex, estrogen-replacement therapy, head trauma, antiperspirant or antacid use, smoking, high blood pressure, heart disease, or stroke. Traumatic Brain Injury Model Systems The Traumatic Brain Injury Model Systems (TBIMS) program was established in 1987 by the National Institute on Disability and Rehabilitation Research of the US Department of Education. It consists of 16 nationwide centers, which provide acute hospital and rehabilitation care. Inclusion criteria for the TBIMS database, from which numerous prospective and retrospective conditional studies have been derived, are as follows: moderate to severe TBI
MAJOR COHORT STUDIES 131 (posttraumatic amnesia [PTA] > 24 hours or LOC > 30 minutes or GCS in emergency department [ED] < 13 or intracranial neuroimaging abnormalities), admitted to ED within 72 hours of injury, age 16 years or greater at time of injury (NDSC, 2008). Harrison-Felix et al. (2004) studied moderately to severely injured patients identified from the TBIMS database, which covered 15 TBI Model Systems Centers (TBIMSCs). Using vital-status information obtained from the Social Security Death Index, the authors found 161 deaths among 2,178 people with TBI followed for 17 daysâ12.8 years after injury for a mortality of 7.4%. Compared with age-sex-rateâspecific mortality in the general population, the standardized mortality ratio (SMR) was 2.00 (95% CI, 1.69â2.31). The SMR for people with TBI who survived for more than 1 year after injury was only slightly lower, 1.95 (95% CI, 1.61â 2.29). The median interval between injury and death was 2 years; of the 161 deaths, 38 occurred between rehabilitation discharge and 1 year after injury. Of the TBIs, 62% resulted from motor- vehicle crashes and 20% from acts of violence. The study participants had a mean age of 37 years; 76% were men, 60% were white, and 37% had severe TBI according to 24-hour postinjury GCS scores. Life expectancy was reduced by 5â9 years (average, 7 years), depending on age at injury, race, and sex. Analysis of causes of death of the 124 people who survived more than 1 year after injury to a maximum followup of 13 years showed that 29% died from circulatory diseases, mainly ischemic heart disease (12%) or other heart disease (9%); 18% from external causes of injury and poisoning (5% from homicide, 7% unintentional, and 1% suicide); 14% from respiratory disease (7% pneumonia); 11% from infectious disease (9% septicemia); 9% from neoplasms (5% lung cancer); 7% from digestive diseases; and 4% in association with seizures (Harrison-Felix et al., 2006). People with TBI were at increased risk of dying from seizures (SMR, 37.17; 95% CI, 12.07â86.74), septicemia (SMR, 11.63; 95% CI, 5.58â21.38), pneumonia (SMR, 3.88; 95% CI, 1.68â7.65), unintentional injuries (SMR, 3.39; 95% CI, 1.81â5.80), or a digestive condition (SMR, 3.29; 95% CI, 1.42â6.49). The authors noted that the power of the study to detect specific outcomes was low. Cifu et al. (1999) studied 665 enrolled patients (inclusion criteria differed in that patients presented to the ED within 24 hours after injury) in 1989â1996 in four geographically diverse TBIMSCs to assess incidence of rehospitalization. The rate of 1- and 2-year postinjury followup was 53% (351 of 665 and 281 of 534 eligible patients, respectively), and the 3-year followup rate was 47% (199 of 424 eligible patients). The rate of rehospitalization ranged from 20% to 22.5% over the 3-year period. Orthopedic or reconstructive surgery was the primary reason for rehospitalization (44%) during the first year; infections were also important, ranging from 8% to 17% during the 3-year period. After the first year, the rate of rehospitalization due to seizures and psychiatric issues increased from 6% to 15% 3 years after injury. Marwitz et al. (2001) conducted a prospective study that assessed the cause and incidence of rehospitalization 1 and 5 years after head injury. A total of 1,547 patients admitted within 24 hours of injury from the 17 TBIMS in 1989â1999 were studied. Data indicated a rehospitalization rate of 23% (58% followup rate) at 1 year and a rate of 17% (55% followup rate) at 5 years. As seen in the work of Cifu et al. (1999), orthopedic or reconstructive surgery was the primary cause, 25%, at 1 year, decreasing to 13% at year 5. Rehospitalization due to infections went from 10% at 1 year to 8% at 5 years. In addition, rehospitalization related to seizures and psychiatric disorders ranged from 12% to 19%, increasing over the followup period.
132 GULF WAR AND HEALTH Nonelective rehospitalization accounted for 66% of admissions at 1 year, which increased to 83% at 5 years. Brown et al. (2007) assessed physical impairment at time of rehabilitation admission and 1 year after injury. Patient data were drawn from the TBIMS database during 1988â2002. Mean GCS score of the sample (n = 3,463) was 10.6, mean duration of hospitalization was 51.5 days, and mean length of stay in the rehabilitation center was 30.3 days. Normal hearing (88%) and normal vision (82%) were common and did not differ between rehabilitation admission and 1 year. Normal swallowing improved from 61% to 95% of patients at 1 year. The percentage of patients with normal limb strength among the four limbs ranged from 42% to 44% at admission, and increased to 82â84% at 1 year. Normal coordination and tone improved from the time of rehabilitation entry to 1 year (60.1â60.8% to 87.2â90.4% and 80.6â81.9% to 90.9â92.2%, respectively). Fewer than 19% of patients had normal standing balance and 48% normal sitting balance at admission; 1 year after injury, 76% had normal standing and 95% normal sitting balance. Seel and colleagues (2003) evaluated the frequency of depression in a sample of 666 outpatients from the 17 TBIMSCs. Mean GCS score at the time of admission was 8.6 Â± 4.6, mean PTA duration was 31.7 Â± 26.2 days, and average evaluation time was 35.3 Â± 26.9 months after injury. As specified in the DSM-IV, the Neurobehavioral Functioning Inventory was administered to identify symptoms of major depressive disorder. Of the participants, 27% reported problems in more than five of the nine criteria A symptoms and thus had a diagnosis of a DSM-IV major depressive disorder. The most commonly reported symptoms were fatigue (29%), distractibility (28%), anger or irritation (28%), and rumination (25%). A significant relationship was found between unemployment and low income status at the time of injury and postinjury depressive symptoms. National Institutes of Health Traumatic Coma Databank The Traumatic Coma Data Bank (TCDB) was a collaborative project between the National Institute of Neurological Disorders and Stroke and four clinical centers around the nation (the Medical College of Virginia at Richmond; the University of California, San Diego; the University of Virginia at Charlottesville; and the University of Texas Medical Branch at Galveston). All severely head-injured patients admitted to any of the centers from April 1983 to April 1988 were prospectively studied (Chesnut et al., 1993). Severe head injury was defined by a GCS of 8 or less at the time of or during the first 48 hours after admission and corresponded with lack of eye opening, lack of comprehensible speech, and the inability to obey commands (Levin et al., 1991b). A total of 1,030 patients were admitted to the four centers, with some variation in exclusion criteria, as described in the four studies below. Preinjury, hospital, and rehabilitation data on each participant were collected and analyzed; no external controls were included (Chesnut et al., 1993). Levin et al. (1991b) investigated the outcome of vegetative states and consciousness after severe closed head injuries. A sample of 650 was obtained after several exclusions: 167 had gunshot wounds (16%), 121 were brain-dead on admission (14%), and 92 were younger than 16 years old (12%). Data were analyzed at the time of discharge and at followup of 6 months, 12 months, 2 years, and 3 years after injury. According to GCS scores and pupillary findings, the 93 patients (14% of the 650) who were discharged in a vegetative state had sustained more severe
MAJOR COHORT STUDIES 133 head injury than their conscious counterparts. Eighty-four patients were able to be followed up adequately, and the results are as follows: 40% became conscious by 6 months, 52% regained consciousness by 1 year, and 58% recovered consciousness within 3 years. Of the remaining 35 (42%) patients, 20 had died and 15 remained in their vegetative state. Analyses of neurologic and demographic features did not indicate any predictive factors for recovery. Levin et al. (1991a) investigated the relationship between intracranial hypertension and memory deficit 6 months (n = 133) and 1 year (n = 126) after severe closed head injury, using neurobehavioral data from the TCDB. Memory was assessed by administering auditory verbal tests (prose recall from the Wechsler Memory Scale, the Selective Reminding Test, and the Digit Span subtest of the Wechsler Adult Intelligence Scale) and nonverbal visual tests. Those with intracranial hypertension in the first 72 hours after injury displayed some memory impairment at the 6-month assessment; impairment diminished with time and was not significant at the 1-year followup. Chestnut et al. (1993) prospectively studied the effects of early and late hypotension (defined by systolic blood pressure 90 mm Hg) on mortality in patients admitted to the TCDB. Of the 1,030 patients admitted, 284 were brain-dead, did not survive resuscitation, or had suffered gunshot wounds, and data on 47 were inadequate (with respect to prehospital course or initial blood pressure or arterial blood-gas results); therefore, 699 patients were eligible for study at the time of hospital arrival. Early shock, defined as hypotension from the time of injury to resuscitation, was experienced by 35% of patients. Mortality was associated with the occurrence of shock: no shock, 27%; shock any time during the early phase, 50%; and presenting with shock, 60%. The association between outcome and hypotension when age, hypoxia, and severe multiple trauma were controlled for was extremely significant (p < 0.0001). Late shock captured the remainder of the patientsâ stay after the first intensive-care-unit (ICU) shift (8 hours). Of the 493 eligible patients, 32% experienced late shock. There was a significant difference (p < 0.001) in mortality and morbidity between no shock (17%) and a recording of hypotension after the first ICU shift (66%). Lu and colleagues (2005) retrospectively studied the decrease in mortality due to severe brain injury from 1984 to 1996. The sample consisted of extracted data from numerous study populations, including 635 patients from the TCDB in 1984â1987 (patients who had penetrating injury, who were deceased on arrival, or who were under 16 years old were excluded). The cohort also included 382 patients from the Medical College of Virginia and 822 from clinical- trial databases. In the total cohort of 1,839 patients, 526 (29%) died. Mortality in the severely brain-injured decreased from 39% in 1984 to 27% in 1996; there was a significant difference (p < 0.0001) in head-injury mortality between 1984â1987 (37%) and 1988â1996 (24%). After adjustment for a variety of factorsâincluding age, admission motor score, and pupillary responseâthe difference remained significant (p < 0.05). OTHER COHORT STUDIES Bryant and Harvey Studies Bryant and Harvey (1999a) conducted a prospective cohort study to compare rates of acute stress disorder (ASD) and PTSD in MVA survivors who sustained a mild TBI with rates in MVA survivors who did not have a TBI. A major trauma center in New South Wales, Australia,
134 GULF WAR AND HEALTH assessed admissions of adults who had been involved in an MVA over a 10-month period. Exclusion criteria for the cohort were inability to speak English, PTA for over 24 hours, not being medically fit or not being on narcotic analgesia other than codeine 4 weeks after trauma, and inability to be contacted. After application of exclusion criteria, 79 (55 male and 24 female) patients with mild TBI and 92 (61 male and 31 female) without mild TBI were evaluated 2â25 days after trauma. An assessment was administered 6 months after injury to 63 (80%) mild-TBI patients and 71 (77%) controls; 37 patients were lost to followup. The presence of ASD was assessed with the Acute Stress Disorder Interview (ASDI), and the presence of PTSD at 6 months was assessed with the PTSD module of the Composite International Diagnostic Interview (CIDI); both were administered by clinical psychologists. Mean injury severity score (ISS) was greater in mild-TBI patients (9.28) than in nonâmild-TBI patients (4.0) (p < 0.001). Non-TBI patients reported fear and helplessness more often than mild-TBI patients during the acute and 6- month followup assessments; intrusive memories were also more common during the acute phase in non-TBI patients. There was no significant difference between 11 mild-TBI and 12 non- TBI patients in the rates of diagnosed ASD (14% and 13%, respectively) or between 15 TBI and 18 non-TBI patients in rates of diagnosed PTSD (24% and 25%, respectively). In a separate analysis of the same MVA population, Bryant and Harvey (1999b) investigated the relationship between PTSD and PCS in a population of mild-TBI patients. Over the period of study, 126 patients were initially identified; at the 6-month followup, 46 (32 male and 14 female) mild-TBI patients (mean ISS, 8.96; SD, 6.08) and 59 (31 male and 28 female) non-TBI patients (mean ISS, 3.92; SD, 3.74) were captured, representing 83% of the original sample. Assessments administered at 6 months were the PTSD module of the CIDI and the Postconcussive Symptom Checklist. Results indicated that 20% (n = 9) of mild-TBI patients and 25% (n = 15) of non-TBI patients met the criteria for PTSD diagnosis. Concentration deficits, dizziness, fatigue, headache, sensitivity to sound, and visual disturbances were reported more often by patients with PTSD than those without it in the mild-TBI sample; concentration deficits and irritability were reported more often in patients with PTSD than those without it in the non- TBI sample. Subjects with PTSD reported more frequent irritability than in those without PTSD in the mild-TBI sample. With the same population as described above, Bryant and Harvey (1998) and Harvey and Bryant (2000) prospectively studied the frequency of ASD after mild TBI and its utility in predicting PTSD. Of 79 patients who sustained a mild TBI and were administered the ASDI within 1 month after trauma, 11 (14%) met the criteria for ASD. The CIDI module for PTSD was administered at 6 months (n = 63) and 2 years (n = 50) after injury; this represented a 63% retention rate of the original study group. At 6 months and 2 years after trauma, 24% (n = 15) and 22% (n = 11) of patients, respectively, met the criteria for PTSD. Of those with ASD, nine (82%) had a diagnosis of PTSD at 6 months and eight (80%) at the 2-year followup. Of those without ASD, six (12%) and three (8%) had PTSD at 6 months and 2 years, respectively. University of Washington Longitudinal Traumatic Brain Injury Studies A number of studies were based on a series of longitudinal investigations of health outcomes related to TBI conducted at the University of Washington by Dikmen and colleagues. The data from the studies have been formed into a repository and have been used to address questions about outcomes. The studies include Behavioral Outcome of Head Injury, Patient Characteristics and Head Injury Outcome, Dilantin Prophylaxis of Post-Traumatic Seizures, and
MAJOR COHORT STUDIES 135 Valproate Prophylaxis of Post-Traumatic Seizures. More information on the individual studies can be found in the literature (Temkin et al., 1990, 1999a, 1999b; Dacey et al., 1991; Dikmen et al., 1991, 2000; McLean et al., 1993). Participants in all the studies used in the present report were English-speaking adults who were admitted to the level I trauma center at Harborview Medical Center in Seattle, Washington, with head injuries and were followed for at least a year. Study subjects were consecutively admitted and met at least the following criteria: any period of LOC, PTA for at least an hour, or other objective evidence of head trauma; an injury that was serious enough to require hospitalization; and survival of at least a month after the injury, at which time the first assessment was done. The Behavioral Outcome of Head Injury and Dilantin Prophylaxis of Post-Traumatic Seizures studies both excluded people who had prior hospitalization for head injury, alcoholism, cerebral disease, a psychiatric disorder, or mental retardation. The Patient Characteristics and Head Injury Outcome study did not exclude subjects on the basis of those conditions. The Dilantin Prophylaxis of Post-Traumatic Seizures and the Valproate Prophylaxis of Post-Traumatic Seizures studies enrolled patients who had more severe head injuries that posed an increased risk of seizures, such as intracranial hematoma, cortical contusion, and depressed skull fracture. Studies to assess a variety of neurocognitive and social function outcomes have been conducted with these study populations. Dikmen et al. (1995b) conducted a prospective study of 436 adults with TBI recruited at the time of injury from a level 1 trauma center for the Behavioral Outcome of Head Injury, Patient Characteristics and Head Injury Outcome, or Dilantin Prophylaxis of Post-Traumatic Seizures study. The subjects made up 85% of the 514 subjects recruited into these studies. The controls were 132 patients enrolled as part of the Patient Characteristics and Head Injury Outcome study who were admitted to the emergency room at Harborview Medical Center of University Hospital after trauma to any part of the body except the head; they were group- matched on age, sex, and education. Analyses weighted the cases to adjust the mix of severity and pre-existing conditions to approximate that of the unselected Patient Characteristics and Head Injury Outcome Study. A variety of neuropsychologic tests were conducted a year after injury. A year after injury, the TBI group performed significantly worse than controls on all the neurocognitive tests except the category test, on which the groups did not differ. A doseâ response relationship between length of coma and level of performance on neurocognitive tests was observed at 1 year after injury; increasing degree of impairment was associated with increasing severity of brain injury. Subjects with the most severe TBI were significantly impaired on all neurocognitive measures. Several other studies by Dikmen and colleagues used subsets of the same cohort of head- injured patients, but controls were friends of the head-injured patients (Dikmen et al., 1986, 1990) so they might not have controlled as well for general health effects of trauma. A study based on mild TBI found mild subtle neuropsychologic effects at 1 month that could no longer be detected at 1 year (Dikmen et al., 1986), whereas those with moderate to severe injuries demonstrated significant impairments at 1, 12, and 24 months after injury compared with healthy controls (Dikmen et al., 1990). Motor dexterity and speed were found to be sensitive to the effects of TBI even at 1 year after injury (Haaland et al., 1994). Memory functions were examined at 1 and 12 months after injury. At 1 year, only those with deep or prolonged impaired consciousness (represented by more than 1 day of coma, GCS score of 8 or lower, and PTA of 2 weeks or longer) were performing significantly worse than controls (Dikmen et al., 1987).
136 GULF WAR AND HEALTH Dikmen et al. (1995c) examined 466 people who were enrolled as part of the Behavioral Outcome of Head Injury Study (21% of the subjects), the Patient Characteristics and Head Injury Outcome Study (50%), or the Dilantin Prophylaxis of Post-Traumatic Seizures Study (29%). The controls were 124 trauma controls from the Patient Characteristics and Head Injury Outcome Study who had sustained bodily injury but not to the head and 88 healthy friends from the Behavioral Outcome of Head Injury Study. Analyses weighted the cases to adjust the mix of severity and pre-existing conditions to approximate that of the unselected Patient Characteristics and Head Injury Outcome Study. Social function was evaluated with the Glasgow Outcome Scale; a structured interview was conducted to collect information on independent living, school, employment, and income. The Sickness Impact Profile (SIP) was also administered. The head- injured were stratified by severity of injury. More severe TBI was related to worse outcome on all measures of social functioning except return to school, in which no difference was detected between TBI patients and trauma controls. Return to work and other neurocognitive outcomes after head injury were assessed in subgroups of the same population (Fraser et al., 1988; McLean et al., 1993; Dikmen et al., 1994; Doctor et al., 2005). Fraser et al. (1988) found poorer neuropsychologic test scores and more dysfunction on the SIP physical scales at 1 month after injury in those who failed to return to work by 1 year after injury compared with those who had returned to work. McLean et al. (1993) assessed employment issues related to head trauma and found that in addition to a lower rate of return to work, participants with TBI were less likely to have remained at the same or similar position (36%) than the friend controls (60%). Dikmen et al. (1994) assessed time to return to work in 366 head-injured patients and 95 trauma controls who worked preinjury. The study participants were drawn from the three studies mentioned previously (45% from the Patient Characteristics and Head Injury Outcome Study, 33% from the Dilantin Prophylaxis of Post- Traumatic Seizures Study, and 22% from the Behavioral Outcome of Head Injury Study). Preinjury workers were followed for 1â2 years after injury to measure the time from injury to first return to work regardless of the length of the employment. Time to return to work was related to severity of TBI. Doctor et al. (2005) used a longitudinal inception cohort design and the same population as Dikmen et al. (1994) and additional TBI subjects from the Valproate Prophylaxis of Post-Traumatic Seizures Study. Employment was assessed at 1 year after injury in 418 TBI subjects who were working before their injuries and compared with expected unemployment rates from a current population survey. The authors found a substantial increase in risk of unemployment after TBI that increased with severity. Several secondary studies have addressed other social outcomes. Patients with moderate to severe TBI were examined over a 2-year period. In spite of improvement, many subjects were unable to return to work, to support themselves financially, to live independently, or to participate in leisure activities for at least 2 years after injury (Dikmen et al., 1993). The authors examined alcohol use before and after injury and in relation to ED blood alcohol concentrations (Dikmen et al., 1995a); 42% of the subjects were intoxicated on arrival at the ED. The amount of drinking and associated problems decreased immediately after injury but were followed by an increase by 1 year although not to the same levels. Patients with more severe injuries decreased their drinking more than those with mild TBI. Blood alcohol in the ED was a good indicator of a history of problem drinking. Burden to spouse and significant others was examined in the same cohort at 6 months after injury (Machamer et al., 2002). Significant others reported both favorable and unfavorable
MAJOR COHORT STUDIES 137 care-giving experiences. Adverse experience or burden was systematically related to increased TBI severity, worse cognitive outcome, increased dependence on others, reported change in the injured, and changes in the life of the significant other as a function of care-giving. Dikmen et al. (2003) examined preinjury to postinjury changes in various facets of everyday life at 3â5 years after injury in subjects who had mild to moderate to severe TBI. Limitations were seen in all activities, including personal care, ambulation, travel, home management, and social relationships; the most affected were work, leisure and recreation, social relationships, and ambulation. The degree of limitations was related to the severity of TBI. Machamer et al. (2005) used the same cohort to examine stability of work up to 3â5 years after injury. Amount of time worked after injury was related to severity of injury and associated impairments, in addition to preinjury work stability and earnings. Once a person returned to work, the ability to maintain uninterrupted employment was related to premorbid characteristics, such as being older, having higher income, and having had a preinjury job with benefits. Pagulayan et al. (2006) used a subset of the same cohort as Dikmen et al. (2003) to examine recovery of function on the SIP at 1, 6, and 12 months and 3â5 years after injury. Significant limitations in all activities were seen at 1 month after injury compared with both friend controls and trauma controls. By 1 year, however, the TBI group still had problems compared with healthy friend controls but not with trauma controls except for leisure and recreation. McLean and colleagues (1993) assessed psychosocial recovery at 1 and 12 months after head injury in 102 hospitalized patients (they were a subsample of Dikmen et al., 1995c). The reference group included 102 friend controls matched for age, education, sex, and race. At 12 months after injury, the head-injured patients differed significantly in seven symptoms on the Head Injury Symptom Checklist: dizziness (p < 0.01), blurred vision (p < 0.001), concentration (p < 0.001), noise (p < 0.05), irritability (p < 0.01), temper (p < 0.01), and memory (p < 0.001). The median number of symptoms presented at 1 year was 5 in those with severe head injury, 2 in those with moderate head injury, 3 in those with mild head injury, and 2 in controls. The severely injured had significantly more symptoms than those with moderate injury or friend controls. Jennett (Oxford, Rotterdam, Cardiff, and Manchester) Studies Jennett and Lewin (1960) studied traumatic epilepsy in 1,000 patients (infants to 65 years old) who sustained nonmissile head injuries and were admitted to the Radcliffe Infirmary in Oxford from November 1948 to February 1952. The cohort consisted of the first 1,000 cases of the Roberts (1979) studies below. Criteria for admission included some period of unconsciousness; 58% of the patients had PTA lasting less than 1 hour and fewer than 20% over 24 hours, and fewer than 50% had a fractured skull (Jennett, 1975). Of the 1,000, 46 (5%) had no history of epilepsy but experienced early epilepsy within a week of admission (the 14 patients with a history of epilepsy were excluded). An unselected series of 821 patients was admitted directly from the accident site and immediately placed under care; 31 (4%) experienced early epilepsy within a week of admission. A selected series of 179 patients was transferred from other hospitals and in general was considered more severe and complicated; 15 (8%) experienced early epilepsy. Of the total population, 90 patients died, including 8 (9%) who had early epilepsy. Of the 75 children under 5 years old and the 122 who were 6â15 years old, 9% and 3%, respectively, had early epilepsy. Early epilepsy was more frequent in patients who experienced
138 GULF WAR AND HEALTH PTA for over 24 hours. On followup after 4 years, late epilepsy was present in 28 (10%) of 275 patients. The latter population included all who experienced early epilepsy and 100 randomly selected patients with uncomplicated injures and PTA of less than 24 hours. Of those with early epilepsy, 29% experienced late epilepsyâan incidence 4 times higher compared to those without (Jennett and Lewin, 1960). Jennett (1969) expanded the Oxford series to include a total of 189 patients with epilepsy (within 8 weeks of injury); 150 cases of epilepsy within the same period were also added and known as the Glasgow series. Results were consistent with previous studies: an increased risk of late epilepsy was found in people with nonmissile injuries who had early epilepsy. Early epilepsy (within 1 week of injury) occurred 30 times more often than in the following 7 weeks (Jennett, 1969, 1973). A group of 73 patients with missile injuries was included for comparison; results indicated that early epilepsy is not necessarily predictive of late epilepsy in such patients, inasmuch as the baseline risk of late epilepsy is already high; 45% of those with missile injuries develop late epilepsy (Jennett, 1969, 1973). Jennett (1962) examined early and late epilepsy by studying 381 patients who sustained blunt head injuries: 139 had early epilepsy, 282 late epilepsy, and 40 both early and late epilepsy. The population was drawn from the 46 patients with early epilepsy in the Oxford series, 93 patients with early epilepsy in Manchester and Cardiff, England, and patients at the Oxford Infirmary outside the study dates. The late-epilepsy series consisted of 58 followup patients with late epilepsy (drawn from 75 patients with early epilepsy and 240 without early epilepsy in the 1,000-patient Oxford series) and 224 patients who presented with a history of head injury and epilepsy. Results were consistent with previous and later studies of this cohort in that a relationship was found between more severe injury (longer PTA, depressed fracture, and early epilepsy) and development of late epilepsy. The effect of depressed fractures on the incidence of epilepsy was studied in over 600 patients from both the Oxford and Glasgow seriesâ333 patients were followed for over 1 year after injury, and 219 were followed for more than 4 years. Early epilepsy was seen in 10% of those with depressed fractures and 4% of those without; late epilepsy was seen in 21% of those with depressed fractures and 8% of those without (Jennett, 1969). Jennett (1973) studied patients from the original Oxford series, patients from the Glasgow series, and 250 patients with depressed fractures from Rotterdam to investigate known risk factors for late epilepsy: early epilepsy, intracranial hematoma (evacuation within 14 days of injury), and depressed fracture. The results supported those of previous studies. In addition, 75% of patients who had one late epileptic episode experienced seizures over the following 2 years, and over 33% experienced at least one seizure per month. Jennett (1975) summarized previous findings on nonmissile injuries from the combined Oxford, Glasgow, and Rotterdam series. Roberts (Oxford, England) Studies Roberts (1979) examined the relationship between a single nonmissile head injury and characteristics of mental and physical disability 3â25 years after injury in two groups of patients. The study population consisted of 548 patients (11 eventually lost to followup) from a total population of 7,000 patients admitted after accidental head injury to the Accident and Neurosurgical Services of the Radcliffe Infirmary, Oxford, England, in 1948â1961. The study
MAJOR COHORT STUDIES 139 population included patients 5â83 years old who remained unconscious or amnestic for a week or longer. Those who developed intracranial infection or sustained spinal-cord or brachial plexus lesion in addition to head injury, American ex-service personnel for whom there was no adequate method of followup, and foreign nationals isolated from their native culture were excluded. Eighty percent of patients were admitted within hours and directly from the accident scene, and 20% were transfers from facilities that lacked adequate neurosurgical treatment. A consecutive series consisted of 479 patients, and a selected series consisted of 69 patients who had presented at the Addenbrooke Hospital in Cambridge in 1948â1970. The latter group consisted of severely head-injured patients who had remained unconscious for 1 month or longer after injury. Eleven patients were lost to followup, and 206 were no longer alive; that left 331 surviving patients (291 from the consecutive series and 40 from the selected series), who were invited for interview and re-examination at the Addenbrooke Hospital or the Radcliffe Infirmary. Close relatives, spouses, or parents were interviewed for 82% of the patients in the consecutive series and 93% of the patients in the selected series. Forty-two patients declined to attend and were visited in their homes. Six survivors were not examined by the authors, but were seen by another neurologist. Intellectual, personality, and neural deficits in each patient were assessed through interview; patients were then administered a neurologic examination while relatives, if available, were questioned by a psychiatric social worker with regard to past and present behavioral issues. Most were assessed by neuropsychologists who tested memory and intellectual deficits. Intellectual function and memory tests (including verbal memory functions) and tests of visuospatial function were administered by a clinical psychologist to assess cognitive deficits. All patients underwent neurologic examination, and 217 were given a series of cognitive-function tests. Roberts (1979) also assessed patients for disordered hypothalamic and pituitary function. Results indicated that anterior hypopituitarism did not increase in frequency because of head trauma; at the time of the study, only one patient had that diagnosis (a 10-year-old boy). The incidence of diabetes insipidus (diagnosed on the basis of polyuria) in the consecutive series was 8 of 291 patients (3%). Hypothalamic hyperphagia was diagnosed in 16 of 291 (5%) in the consecutive series and 6 of 40 (15%) in the selected series. Lower age and greater severity of injury seem to contribute to increased rates of diabetes insipidus and hyperphagia among the injured. Positional vertigo was identified in 71 (24%) of the consecutive-series patients 10 years after injury; 58 (20%) reported persistent headaches (all four types, as classified by severity) several years after injury. No clear association was found between symptoms and age, complications, or severity of injury. Incidences, however, are likely to be underestimates, inasmuch as vertigo and headache are underreported in cases of severe head injury (Roberts, 1979). Posttraumatic epilepsy was assessed 10â24 years after injury in the consecutive series: 75 (26%) had one or more seizures during the study period; 22 (29%) of those were early (within 1 week after injury). The results were consistent with those of the Jennett studies described above (Roberts, 1979). Lewin et al. (1979) determined the causes of death of 75 severely injured patients who had been discharged from the Radcliffe Infirmary 10â24 years earlier with severe TBI (the patients had remained unconscious for a week or more). No rates of death were given. Compared with the general population of England and Wales, deaths from meningitis (SMR, 65), epilepsy (SMR, 40), drowning (SMR, 20), and respiratory diseases (SMR, 2) were increased. No excesses
140 GULF WAR AND HEALTH of deaths from suicide, accidents, cardiovascular or cerebrovascular disease, or malignancies were observed. In addition, the overall prevalence of posttraumatic epilepsy was 28% in the consecutive series and ranged from 7% in those with uncomplicated injury and PTA of less than a week to 61% in those with complicated injury (compression and traumatic or surgical penetration). An increase in the prevalence of seizures was seen with length of PTA (uncomplicated) and with duration of coma (complicated and uncomplicated). STUDIES OF SPORTS-RELATED TRAUMATIC BRAIN INJURY Concussions are relatively common in people who participate in contact sports, so studies of such people afford a unique opportunity to assess the short-term and long-term consequences of head injury. It is estimated that over 300,000 sports-related concussions occur each year in the United States (Guskiewicz et al., 2005). The studies discussed below focus on the long-term health outcomes related to sports-related TBI. Although not all the studies include large cohorts, the committee believed that it was important to discuss the strengths and limitations of studies of sports-related TBI. Contact sports provide a useful laboratory for assessing the influence of recurring mild TBI on such health outcomes as dementia and Alzheimer disease. One advantage of studies of sports-related TBI is that there is a large population to draw from. A major limitation, however, is that many studies are based on self-reported measures. For instance, participantsâ TBIs are commonly ascertained by asking subjects whether they recall having had a concussion or, less commonly, whether they had a clinical diagnosis of a head injury. Football American football is a âcollision sportâ that is widely known for causing a variety of injuries, including cerebral concussions. It has been reported that a large percentage of professional football players have sustained at least one concussion during their careers (Guskiewicz et al., 2005). Guskiewicz and colleagues (2005, 2007) studied the association of recurrent concussions sustained and long-term health outcomes, including mild cognitive impairment and risk of depression in retired professional football players. The authors originally sent a survey to all 3,683 living members of the National Football League Retired Players Association. The survey instrument included questions about musculoskeletal, cardiovascular, and neurologic conditions experienced during and after the football career. Questions were also asked about the number of concussions sustained during the football career and the presence of such health conditions as depression, Parkinson disease, Alzheimer disease, and schizophrenia (Guskiewicz et al., 2005). Concussion history was based on playersâ recall of injury events, and concussion was defined as âinjury resulting from a blow to the head that caused an alteration in mental status and one or more of the following symptoms: headache, nausea, vomiting, dizziness/balance problems, fatigue, trouble sleeping, drowsiness, sensitivity to light or noise, blurred vision, difficulty remembering, and difficulty concentrating.â The mailed questionnaire included the SF-36 Measurement Model for Functional Assessment of Health and Well-Being to assess daily-living functioning. A physical-health composite score was calculated. The authors sent out a second questionnaire to a subset of 1,754 of the original population that included questions on memory and issues related to mild cognitive impairment.
MAJOR COHORT STUDIES 141 Results were cross-tabulated from the initial questionnaire. Of the original 3,683 general health surveys distributed, 2,552 (69%) were completed. Of those who responded, 1,513 (61%) reported having sustained at least one concussion, and 597 (24%) reported three or more concussions. The authors found an association between recurrent concussion and clinically diagnosed mild cognitive impairment (MCI) (p = 0.02) and self-reported significant memory impairment (p = 0.001). People who reported sustaining three or more concussions had a 5-fold prevalence of MCI diagnosis and a 3-fold prevalence of reported significant memory problems compared with those without a history of concussion. The authors found no association between recurrent concussion and Alzheimer disease but observed earlier onset of Alzheimer disease in the retired football players than in the general American male population. Guskiewicz and colleagues (2007) restudied the same population of 2,552 retired football players to investigate the association between prior head injury and diagnosis of clinical depression. They found that 269 (11%) of the respondents reported a prior or current diagnosis of clinical depression and found an association between recurrent concussion and diagnosis of lifetime depression (p < 0.005). Retired players reporting three or more concussions were three times more likely to have a diagnosis of depression than players with no history of concussion. Similarly, those with a history of one or two previous concussions were 1.5 times more likely to have a diagnosis of depression than players with no history of concussion. As mentioned previously, an important limitation of these studies is that they rely on self- reports of exposure (TBI). Because history of head injury is based on recollection of events that occurred many years before the survey, substantial recall bias may be introduced. Boxing Sports literature has been used to evaluate boxers and to assess the effects of repetitive head injury on long-term health outcomes, particularly neurologic and neurocognitive outcomes, such as dementia pugilistica and the relationship between the APOE e4 gene and chronic TBI. A number of the studies discussed below assess long-term health outcomes in brain-injured boxers, including two (Porter and Fricker, 1996; Porter, 2003) that study the same cohort of boxers. Porter and Fricker (1996) conducted a neuropsychologic assessment of 20 amateur boxers 16â25 years old in the six largest boxing clubs in Dublin, Ireland, and 20 controls matched for age and socioeconomic status. Each amateur boxer was to have competed in a minimum of 40 amateur matches. The boxers and the controls were given a battery of neuropsychologic tests by an independent examiner initially in 1992 and then again 15â18 months later. The tests included Trail-Making Tests A and B, the Finger Tapping Test (FTT), and the Paired Associate Learning test. The authors found that the boxers performed significantly better than the controls in Trail-Making Tests A and B. However, the control groupâs scores on the FTT were significantly higher than those of the boxers. The authors noted that there was no evidence of neuropsychologic impairment in the boxers compared with the controls, and they found no association between boxing and performance on any of the neuropsychologic tests. Porter (2003) conducted a followup study of the same population of 20 amateur boxers and 20 matched controls. Again, the subjects underwent a battery of neuropsychologic tests after an initial assessment at 18 months, 4 years, 7 years, and 9 years. The boxers scored higher than the controls on Trail Making Tests A and B at all times and lower on the FTT at all times except baseline for the dominant hand. The authors found no evidence of neuropsychologic impairment
142 GULF WAR AND HEALTH over the 9-year period; in fact, the boxers improved on some of the tests in comparison with the controls. Soccer Studies of soccer players have been conducted to evaluate the association of TBI with long-term health outcomes, particularly neurocognitive outcomes. Soccer is a popular sport that is considered relatively safe for the general population; however, it is designated a contact sport because rates of concussion in soccer players are high and have been found to be equivalent to those in football players (Matser et al., 1998). Matser et al. (1998), Guskiewicz et al. (2002), Rutherford et al. (2005), and Straume- Naesheim et al. (2005) conducted studies of neurocognitive outcomes related to soccer-related head injuries. Two studies found neuropsychologic impairment in head-injured soccer players. Matser and colleagues (1998) assessed neurocognitive impairment in soccer players with chronic TBI and found that soccer players performed worse than controls on neurocognitive tests of planning, memory, and visuoperceptual tasks. The number of concussions was inversely related to scores on neurocognitive tests. Rutherford and colleagues (2005) studied neuropsychologic impairment in amateur soccer, rugby, and non-contact-sports players and found that the number of head injuries was a significant predictor of scores on the Trails B response test (p = 0.014) and the Test of Attention Performance Divided Attention (p = 0.020). The latter study was designed to be exploratory. Two other studies, however, did not find a relationship between soccer-related TBI and neurocognitive outcomes. Guskiewicz and colleagues (2002) evaluated neurocognitive outcomes in collegiate athletes (including participants in soccer and other sports) and found no significant relationship between a history of soccer-related concussions and scholastic aptitude or neurocognitive performance. Straume-Naesheim and colleagues (2005) studied neuropsychologic impairment in head-injured Norwegian elite soccer players; lifetime heading exposure was not associated with neuropsychologic test performance. As with much of the sports literature discussed above, diagnosis of TBI in studies of soccer players was generally based on self-reports of exposure (for example, questionnaires that asked about number of concussions in the past or number of headings in previous matches) or on surrogates of exposure (such as number of games played). Reliance on self-reports of exposure may introduce recall bias, and this should be considered in evaluating the results of the studies.
TABLE 5.1 Major Cohort Studies (Shaded) and Derivative Studies Responded or Subgroup (n Contacted or Enrolled Type of Study or Date(s) of = Eligible Located (% (Response Reference Eligible Population Methods Enrollment Subjects) of Eligible) Rate) Comments Walkerâs Studies of Head-Injured Bavarian World War I Veterans Walker et al., 1,000 Bavarian men with Cohort 1916â1927 1,000 men 555 cases, 563 555 cases, 563 1971 head injuries from World randomly controls controls War I randomly selected selected from from among 5,500 cases at among 5,500 head-injury center 1916â cases at head- 1927 with âsufficient injury center information for analysisâ of in 1916â1927 nature of injury; with 1,000 unwounded Bavarian âsufficient World War I veterans on information pension lists for receiving for analysisâ medal; all born 1880â1900 of nature of injury 1,000 unwounded Bavarian World War I veterans on pension lists for receiving medal Population (where appropriate) Enrolled Reference Purpose Study Design Eligible Located (Response Rate) Comments Weiss et al., Mortality Cohort 1,010 Bavarian men with head injuries from World War I; 1982 1,000 unwounded Bavarian World War I veterans; final numbers: 647 cases, 616 controls 143
144 Responded or Subgroup Contacted or Enrolled Type of Study or Date(s) of (n=Eligible Located (% (Response Reference Eligible Population Methods Enrollment Subjects) of Eligible) Rate) Comments Walkerâs Studies of Head-Injured World War II Veterans Walker and Head-injured men wounded Cohort Group 1 (Cushing 364 Of 364 men 313 (of 343 No external Erculei, 1969 in World War II who were General Hospital, originally men, 21 comparisonâ studied at Cushing General Framingham, identified, refused to all Hospital in Framingham, MA): enrolled information participate, 30 comparisons MA, in 1945â1946 1945â1946 (n = obtained on died) between those (experienced at least one 241) 343 (94%) with and posttraumatic epileptic without seizure) or were identified Group 2 posttraumatic through Army and VA (Baltimore seizures pension rosters as part of group): enrolled followup in Baltimore in 1950â1954 (n = 1950â1954 (unselected) 123) Baltimore group was matched in class and severity of injury Population (Where Appropriate) Enrolled Reference Purpose Study Design Eligible Located (Response Rate) Comments Walker and Posttraumatic symptoms, Cohort 313 men examined, questioned for relevant symptoms Erculei, 1969 including nervousness, headache, irritability, easy fatigability, impaired memory, dizziness, impaired mentation, lack of concentration, insomnia, intolerance to alcohol Walker and Neurologic deficit Cohort Neurologic deficit assessed with two subgroups of population: patients Erculei, 1969 with no neurologic deficit (n = 50), patients with neurologic deficit (n = 199) Walker and Socioeconomic status, Cohort Cohort divided into two groups: employed (n = 182), unemployed (n = Erculei, 1969 employment status 121)
Population (Where Appropriate) Enrolled Reference Purpose Study Design Eligible Located (Response Rate) Comments Walker and Posttraumatic epilepsy Cohort Comparison of subset of Group 1 (Cushing General Hospital), which Erculei, 1969 included 232 men with posttraumatic epilepsy, and Group 2 (Baltimore group), which included 123 men with posttraumatic encephalopathy Walker and Posttraumatic epilepsy Cohort Comparison of subset of Group 1 (Cushing General Hospital), which Erculei, 1970 included 230 men with posttraumatic epilepsy, and Group 2 (Baltimore group), which included 123 men with posttraumatic encephalopathy Responded or Subgroup (n= Contacted or Enrolled Type of Study or Date(s) of Eligible Located (% (Response Reference Eligible Population Methods Enrollment Subjects) of Eligible) Rate) Comments Finnish Studies Achte et al., Veterans who suffered mild, Prospective cohort 1939â1945 1969 moderate, or severe open and closed head injuries in Finnish wars of 1939â1945; 3,552 men with mild, moderate, or severe open and closed head injuries were studied 22â26 years after injury Population (Where Appropriate) Enrolled Reference Purpose Study Design Eligible Located (Response Rate) Comments Achte et al., Psychiatric disturbances Prospective cohort In more inclusive sample, 10,000 men with mild, moderate, or severe No comparison 1991 open, closed head injuries were studied 50 years after injury group 145
146 Responded or Subgroup (n= Contacted or Enrolled Type of Study or Date(s) of Eligible Located (% (Response Reference Eligible Population Methods Enrollment Subjects) of Eligible) Rate) Comments Teuberâs Cohort Weinstein, World War II (and to Caseâcontrol study 1948â1950s Over 300 1954 smaller extent World War I and Korean War) veterans who sustained penetrating brain injuries or peripheral nerve injuries Population (Where Appropriate) Enrolled Reference Purpose Study Design Eligible Located (Response Rate) Comments Teuber and Spatial and motor function Caseâcontrol 35 veterans with penetrating brain injuries, 12 veterans with peripheral Weinstein, nerve injuries from the original series 1954 Weinstein Intelligence scores Caseâcontrol 62 veterans with penetrating brain injuries, 50 veterans with peripheral and Teuber, nerve injuries for whom preinjury Army General Classification Test 1957a score was available Weinstein Intelligence scores Caseâcontrol 62 veterans with penetrating brain injuries, 50 veterans with peripheral and Teuber, nerve injuries for whom preinjury Army General Classification Test 1957b score was available Weinstein et Sensorimotor discrimination Caseâcontrol 40 World War II, 3 Korean War veterans with penetrating head al., 1958 injuries, 20 controls with peripheral nerve injuries from original series Corkin et Life expectancy Caseâcontrol 190 World War II veterans with penetrating head injuries, 106 World al., 1984 War II veterans with peripheral nerve injuries who were in original series Corkin et Cognitive performance Caseâcontrol 57 men with penetrating head injuries, 27 with peripheral nerve injuries al., 1989 who were in original series
Responded or Subgroup (n= Contacted or Enrolled Type of Study or Date(s) of Eligible Located (% (Response Reference Eligible Population Methods Enrollment Subjects) of Eligible) Rate) Comments W.F. Caveness Studies of Korean War Veterans Caveness, Military personnel of Prospective cohort 467 No reference 1963 Korean War who suffered group head injuries and were treated in either US Naval Hospital in Yokusaka or US Navy hospital ships off coast of Korea Population (Where Appropriate) Enrolled Reference Purpose Study Design Eligible Located (Response Rate) Comments Evans, 1962 Posttraumatic epilepsy Cohort 422 Korean War veterans with head injuries treated in US Naval No reference Hospital in Yokusaka or in US hospital ships, assessed 3â11 years after group; no injury screening for preinjury 422 men who had sustained head injury during Korean War; of original seizure disorder 493 in cohort, 33 were excluded because they were not in US armed forces at time of injury, 10 because of inadequate information on original head injury, 12 because injuries were related to face or cervical spine, 2 because seizures were responsible for initial head injury, 1 because head injury was probably early effect of chronic encephalitis, 12 because they died during period, 1 because admission to hospital was determined by first epileptic seizure after previous head injury Caveness et Posttraumatic epilepsy 5 retrospective 407 Korean War veterans with head injuries (214 missile, 52 blast, 141 No reference al., 1962 cohorts from 3 wars blunt), 135 with dura mater rupture, assessed 5 years after trauma group; (World War I, World undetermined War II, Korean War) whether preinjury seizures Caveness, Posttraumatic epilepsy Cohort 356 Korean War veterans with head injuries assessed 8â11 years after No reference 1963 injury group; no screening for preinjury seizure disorder Caveness, Neurologic deficits Cohort 356 Korean War veterans with head injuries assessed 8â11 years after No reference 147 1966 injury group
148 Responded or Subgroup (n= Contacted or Enrolled Type of Study or Date(s) of Eligible Located (% (Response Reference Eligible Population Methods Enrollment Subjects) of Eligible) Rate) Comments Vietnam Head Injury Study Caveness, Phase I: Head-injured Registry; medical- 1967â1970 1,539 1972 Vietnam soldiers who records review survived first week after sustaining head injury Phase IIâ Phase II: Head-injured Retrospective 1981â1984 520 520 National subjects from original cohort Naval registry and 85 matched Medical normal volunteers evaluated Center, 2008 in 1981â1984, 12â15 years after injury Grafman, Phase III: 182 of 520 head- Retrospective 2004â2006 520 484 (93%) 182 from 2007 injured subjects who were cohort phase II assessed in phase II were (38%), 7 new included in phase III; 17 enrollees patients identified in phase I who did not attend phase II were assessed; 32 of original 85 control subjects in phase II attended phase III; 23 were newly recruited for phase III Population (Where Appropriate) Enrolled Reference Purpose Study Design Eligible Located (Response Rate) Comments Weiss et al., Posttraumatic epilepsy Retrospective Participants from phase I, including 378 of 1,221 participants found to 1983 cohort have posttraumatic seizures Rish et al., Mortality, posttraumatic Prospective cohort Participants from phase I, including 1,127 male Vietnam veterans alive 1983 epilepsy 1 week after trauma
Population (Where Appropriate) Enrolled Reference Purpose Study Design Eligible Located (Response Rate) Comments Salazar et al., Posttraumatic epilepsy Retrospective Participants from phase I, including 421 (of 1,131) head-injured men No reference 1985 cohort group; unclear whether had preinjury seizures Grafman et Face discrimination, Retrospective Participants from phase II, including 213 men with penetrating TBI, 49 Assessed al., 1986 memory cohort controls outcomes based on region of brain affected Salazar et al., Intelligence, reasoning, Retrospective Participants from phase II including 15 veterans who suffered unilateral Assessed 1986 attention; memory, verbal cohort penetrating missile wounds to basal forebrain, 49 uninjured controls, outcomes based free recall, nonverbal 113 patients with lesions elsewhere in brain on region of memory, language brain affected Kraft et al., Occupational, educational Retrospective Participants from phase II, including 520 men with penetrating head 1993 achievement cohort injury, 85 uninjured controls Schwab et al., Measured work status 15 Retrospective Participants from phase II, including 520 men with penetrating head Assessed work 1993 years after injury; cohort injury, 85 uninjured controls outcomes neurologic, neurophysiologic, social- interaction impairments Grafman et Violence, aggression Retrospective Participants from phase II, including 279 male veterans, 57 healthy Assessed al., 1996 cohort controls outcomes based on region of brain affected Groswasser et Cognitive, vocational Retrospective Participants from phase II, including 74 with penetrating head injury, al., 2002 outcome cohort 37 with closed head injury Koenigs et al., PTSD Retrospective Participants from phase III, including 193 veterans with lesions 2007 cohort distributed throughout brain (as result of penetrating head injuries sustained during combat), 52 veterans with combat exposure but no brain injury Raymont et Cognitive outcomes Retrospective Participants from phase III, including 520 with head injury from al., 2008 cohort original registry, 85 matched healthy volunteers, evaluated in 1981â 1984, 12â15 years after injury; of 520 from phase II, 484 still alive, 182 attended phase III 149
150 Responded or Subgroup (n= Contacted or Enrolled Type of Study or Date(s) of Eligible Located (% (Response Reference Eligible Population Methods Enrollment Subjects) of Eligible) Rate) Comments Vietnam Experience Study Luis et al., Male US Army veterans January 1965â 4,462 veterans 2003 who entered military and December 1971 randomly served at least 4 mo on selected from active duty: 9,324 who eligible served single tour in population Vietnam, 8,989 who served elsewhere Population (Where Appropriate) Enrolled Reference Purpose Study Design Eligible Located (Response Rate) Comments Luis et al., PPCSC Cross-sectional 3,957 veterans of original population; 329 excluded because they did 2003 cohort sample not meet criteria for either ICD-10 or DSM-IV PPCSC, but not both; 55 excluded because of hospitalization; 121 excluded because of incomplete data Vanderploeg Psychiatric, neurologic, Cross-sectional 4,384 veterans of original population; 40 excluded because of et al., 2007 psychosocial outcomes cohort sample hospitalization after injury; 38 excluded because of incomplete data Vanderploeg Work, marital status Logistical regression 4,322 veterans of original population; 53 excluded because of et al., 2003 analysis hospitalization after injury; 87 excluded because of incomplete data Vanderploeg Neuropsychologic outcomes Cross-sectional 4,384 veterans of original population; 40 excluded because of et al., 2005 cohort sample hospitalization after injury; 38 excluded because of incomplete data Responded or Subgroup (n= Contacted or Enrolled Type of Study or Date(s) of Eligible Located (% (Response Reference Eligible Population Methods Enrollment Subjects) of Eligible) Rate) Comments Rochester Epidemiology Project Melton, Residents of Rochester, Medical-records 1910âpresent Population of 1996 Olmsted County, MN linkage system Olmsted County, MN
Population (Where Appropriate) Enrolled Reference Purpose Study Design Eligible Located (Response Rate) Comments Annegers et Cancer Double cohort All traumatic brain injuries in Olmstead County, MN, 1935â1974; al., 1979 patients must have survived initial trauma and had no known pre- existing tumor; 2,953 patients followed for total of 29,859 person-years Annegers et Seizures Population-based 2,747 patients of Olmstead County, MN, with head injuries sustained al., 1980 retrospective cohort 1935â1974 Chandra et Alzheimer disease Population-based All incident cases of clinically diagnosed Alzheimer disease in al., 1989 retrospective cohort population of Rochester, MN, with onset 1965â1974 (n = 274) Williams et Dementia, parkinsonism, Population-based 821 Olmsted County residents with head trauma and presumed brain al., 1991 ALS, PD retrospective cohort injury occurring 1935â1974 Siva et al., Multiple sclerosis Population-based 225 incident cases of multiple sclerosis 1905â1991, 164 prevalence 1993 retrospective cohort cases (December 1, 1991) of definite MS in population of Olmsted County, MN Kurland, Multiple sclerosis Population-based All cases of MS diagnosed in Olmsted County, MN, 1905â1991 (n = 1994 retrospective cohort 223) Annegers et Seizures Population-based Incidence of acute symptomatic seizures in population of Rochester, al., 1995 retrospective cohort MN, 1935â1984 (696 episodes of incident acute symptomatic seizures in 692 people; 4 people had two episodes with different etiologies) Annegers et Seizures Population-based 4,541 patients residing in Olmstead County, MN, with head injuries al., 1998 retrospective cohort sustained 1935â1984 Nemetz et Alzheimer disease Population-based Cohort consisted of all Alzheimer disease incident cases diagnosed in al., 1999 retrospective cohort 1965â1984 in Olmstead County, MN; 151 cases excluded because previously identified as having history of head trauma Singer, 2001 Seizures Population-based 4,541 patients residing in Olmstead County, MN, with head injuries retrospective cohort sustained 1935â1984 Bower et al., Parkinson disease Caseâcontrol study 196 Parkinson disease patients living in Olmstead County, MN, with 2003 onset 1976â1995 Brown et Mortality after TBI Population-based Any Olmstead County, MN, resident with medically attended TBI in al., 2004 retrospective cohort 1985â1999 (n = 45,831); random 15.7% sample of TBI patients (n = 7,175) reviewed, 1,448 met inclusion criteria Flaada et al., Mortality after TBI Population-based 17% of all TBI patients in (total = 45,791) Rochester Epidemiology 2007 retrospective cohort Project, 1985â1999; sample = 7,800; 1,443 met TBI case definition 151
152 Responded or Subgroup (n= Contacted or Enrolled Type of Study or Date(s) of Eligible Located (% (Response Reference Eligible Population Methods Enrollment Subjects) of Eligible) Rate) Comments Community-Based Study of Injuries in the Aquitaine, France Tiret et al., People living in Aquitaine, Population December 1985â 2.7 million During 1-year 1989 France, who sustained December 1986 (1986 period, 1,181 injury in 1985â1986 population in deaths region) registered from death certificates, 8,190 hospital admissions observed during sampling periods in residents of Aquitaine Population (Where Appropriate) Enrolled Reference Purpose Study Design Eligible Located (Response Rate) Comments Masson et Self-reported functional Prospective cohort 231 head-injured with various degrees of head injury; 80 controls with al., 1996 status lower-limb injury (LLI); 64 LLI, 176 head-injured patients reviewed (114 minor, 35 moderate, 27 severe) Masson et Disability Prospective cohort 407 head-trauma patients; 5 years after injury, 64 patients deceased, 36 al., 1997 lost to followup
Responded or Subgroup (n= Contacted or Enrolled Type of Study or Date(s) of Eligible Located (% (Response Reference Eligible Population Methods Enrollment Subjects) of Eligible) Rate) Comments Canadian Study of Health and Aging CSHA, Representative sample of Population-based 1991â2002 CSHA-1: 2008 Canadian population 65 cohort 9,008; years old or older on CSHA-2: October 31, 1990, in 39 5,703; CSHA- urban centers, rural areas in 3: 3,437 10 Canadian provinces Population (Where Appropriate) Enrolled Reference Purpose Study Design Eligible Located (Response Rate) Comments CSHA, Alzheimer disease, Prospective 258 with Alzheimer disease 1994 CSHA-1 population-based cohort Lindsay et Alzheimer disease, Prospective 194 with Alzheimer disease, 3,894 controls al., 2002 CSHA-2 population-based cohort Responded or Subgroup (n= Contacted or Enrolled Type of Study or Date(s) of Eligible Located (% (Response Reference Eligible Population Methods Enrollment Subjects) of Eligible) Rate) Comments Traumatic Brain Injury Model Systems NDSC, Patients entered into any Prospective, 1987âpresent 2008 nationwide Traumatic Brain longitudinal Injury Model Systems multicenter Centers and meeting the following criteria: moderate to severe TBI (PTA > 24 h or LOC > 30 min or GCS in ED < 13 or intracranial neuroimaging abnor- malities); admitted to ED within 72 h of injury; > 16 years old at time of injury 153
154 Population (Where Appropriate) Enrolled Reference Purpose Study Design Eligible Located (Response Rate) Comments Harrison- Mortality Retrospective 2,178 patients in 15 TBI Model Systems Centers treated 1988â Felix et al., December 31, 2000 2004 Harrison- Mortality Retrospective 2,140 people surviving 1 year after injury in 15 TBI Model Systems Felix et al., Centers treated 1988âDecember 31, 2000 2006 Cifu et al., Rehospitalization Prospective 665 patients admitted to ED within 24 h of injury to four TBI Model 1999 Systems Centers 1989â1996; response rate for both 1 and 2 years after injury 53% of eligible patients, 3-year response rate 47% Marwitz et Rehospitalization Prospective 1,547 patients admitted within 24 h of injury to 17 TBI Model Systems al., 2001 Centers 1989â1999; 1-year followup 895 (58%) patients; 5-year followup 442 (55%) patients Brown et Physical impairment Prospective 3,463 people in TBI Model Systems database 1988â2002 with al., 2007 longitudinal complete physical examination data at rehabilitation admission, 1 year multicenter after injury descriptive analysis Seel et al., Depression Prospective 666 patients derived from 17 TBI Model Systems Centers who 2003 received followup evaluations 1996â2000 (10â26 mo after injury) Responded or Subgroup (n= Contacted or Enrolled Type of Study or Date(s) of Eligible Located (% (Response Reference Eligible Population Methods Enrollment Subjects) of Eligible) Rate) Comments National Institutes of Health Traumatic Coma Databank Levin et al., Severely head-injured Prospective April 1983âApril 1,030 1991b patients (defined by GCS of 1988 8 or less at time of or during first 48 h after admission, corresponding with lack of eye opening, lack of comprehensible speech, and inability to obey commands) admitted to four geographically diverse clinical centers
Population (Where Appropriate) Enrolled Reference Purpose Study Design Eligible Located (Response Rate) Comments Levin et al., Vegetative state, and Prospective 650 patients available for analysis after following exclusions: 167 1991b consciousness (16%) had gunshot wounds, 121 (14%) were brain-dead on admission, 92 (12%) were under 16 years old Levin et al., Intracranial hypertension in Prospective Intracranial pressure recorded at 6-mo followup (n = 149), 1-year Comparison 1991a relation to memory deficits followup (n = 132); 133 patients 6 mo after injury, 126 patients 1 year group, normal after injury assessed with auditory verbal and nonverbal visual memory community tests residents of Galveston (n = 27) matched on age, education Chesnut et Early and late hypotension Prospective 699 patients available for analysis at time of hospital arrival after al., 1993 in regards to mortality following exclusions: 284 were brain-dead on admission, did not survive resuscitation, or had GSW to head; 29 had insufficient information on prehospital course; 18 had insufficient blood-pressure or blood-gas results Lu et al., Mortality Retrospective 635 patients 16â65 years old from TCDB in 1984â1987 (163 with 2005 penetrating injury, 76 dead on arrival excluded); 382 from Medical College of Virginia and 822 from clinical-trial databases also included Responded or Subgroup (n= Contacted or Enrolled Type of Study or Date(s) of Eligible Located (% (Response Reference Eligible Population Methods Enrollment Subjects) of Eligible) Rate) Comments Bryant and Harvey Studies Bryant and 222 adults 16â65 years old Prospective Unspecified 10- 98 had mild 79 (81%) At 6 mo, 63 Exclusion Harvey, (mild TBI, no TBI) admitted mo period TBI included patients criteria: 1998 to major trauma center in captured (80% inability to New South Wales, Australia followup rate) speak English, after MVA PTA > 24 h, not medically fit or on narcotic analgesia other than codeine, inability to be 155 contacted
156 Population (Where Appropriate) Enrolled Reference Purpose Study Design Eligible Located (Response Rate) Comments Bryant and Influence of mild TBI on Prospective 222 patients admitted during 10-mo period; 79 patients with TBI, 92 Harvey, acute stress disorder, PTSD without TBI included in study; at 6-mo followup, 63 (80%) mild-TBI 1999a patients, 71 (77%) non-TBI patients eligible Harvey and Relationship between acute Prospective 2-year assessment of mild-TBI study population from Bryant 1998; 50 Bryant, stress disorder and PTSD mild-TBI patients captured (63% retention rate from original 2000 after MTBI population, 79% retention from 6-mo population) Bryant and PTSD, PCS Prospective 145 of admitted patients during study period; 46 mild-TBI patients, 59 Harvey, no-TBI patients captured at 6-mo assessment (83% of eligible sample) 1999b Responded or Subgroup Contacted or Enrolled Type of Study or Date(s) of (n= Eligible Located (% (Response Reference Eligible Population Methods Enrollment Subjects) of Eligible) Rate) Comments University of Washington Longitudinal Traumatic Brain Injury Studies Behavioral Consecutively Prospective cohort 1980â1982 102 English-speaking Outcome in Head admitted adults to adults; enrolled Injury, McLean et Level I trauma center at time of injury al., 1993 at Harborview; and subjects met prospectively following criteria: followed to 1 any period of loss of year; of 102 consciousness, PTA enrolled and for at least 1 h, or examined at 1 other objective month, 97 evidence of head examined at 1 trauma; injury had to year require hospitalization; age range, 15â60 years; patients with pre- existing conditions excluded; 102 friend controls included in study
Responded or Subgroup Contacted or Enrolled Type of Study or Date(s) of (n= Eligible Located (% (Response Reference Eligible Population Methods Enrollment Subjects) of Eligible) Rate) Comments Patient Consecutively Prospective cohort 1984â1986 352 eligible 242 (69 %) English-speaking Charecteristics and admitted adults to survivors adults; 221 of Head Injury Level I trauma center 242 (91%) Outcome, Temkin at Harborview; followed to 1 et al., 1990; Dacey subjects admitted year after injury; et al., 1991 with TBI and met subjects not following criteria: excluded if they any period of LOC, had pre-existing PTA of 1 h or more conditions or other objective evidence of cerebral trauma; subjects had to be at least 15 years old; 132 trauma controls included in study Dilantin Consecutively Prospective cohort 1983â1987 234 survivors 170 (73%) English-speaking Prophylaxis of admitted head injured adults; 137 of Post-Traumatic adults to Level I 170 (81%) Seizures, Temkin trauma center at followed to 1 et al., 1990; Harborview; subjects year after injury; Dikmen et al., met one or more of subjects with 1991 following criteria: pre-existing GCS score of 10 or conditions below, cortical excluded; contusion docu- subjects had to mented on CT, be at least 15 depressed skull years old fracture, subdural hematoma, epidural hematoma, traumatic intracerebral hema- toma, penetrating head wound, or seizure within 24 h after injury 157
158 Responded or Subgroup Contacted or Enrolled Type of Study or Date(s) of (n= Eligible Located (% (Response Reference Eligible Population Methods Enrollment Subjects) of Eligible) Rate) Comments Valproate English-speaking Prospective cohort 1991â1995 342 survivors 273 of 342 Prophylaxis of adults admitted to (80%) survivors Post-Traumatic level 1 trauma center seen 6 mo after Seizures, Temkin at Harborview injury, 212 of et al., 1999a; Medical Center in 281 (75%) Dikmen et al., Seattle, WA; subjects survivors seen 1 2000 consecutively year after injury; admitted with TBI denominator and met one or more changed at 1 year of following criteria: because of study cortical contusion, decision to depressed skull suspend 1-year fracture, subdural testing in last 61 hematoma, epidural cases hematoma, intra- cerebral hematoma, penetrating head wound, or a seizure within first 24 h after injury; subjects had to be at least 18 years old; subjects excluded if they had pre-existing conditions Population (Where Appropriate) Enrolled Reference Purpose Study Design Eligible Located (Response Rate) Comments Dikmen et al., Neuropsychologic, Prospective cohort 20 hospitalized subjects with mild head injury; 19 uninjured friend 19 of 20 (95 %) 1986 social outcomes of controls; subjects with pre-existing conditions excluded; subjects were mild head injured mild TBI from Behavioral Outcome in Head Injury study seen at 1 year; healthy friend controls may not control for general effects of trauma
Population (Where Appropriate) Enrolled Reference Purpose Study Design Eligible Located (Response Rate) Comments Dikmen et al., Memory Prospective cohort 102 closed head-injured patients selected from Behavioral Outcome of 97 of 102 head- 1987 Head Injury study; 102 friend controls; no pre-existing conditions injured subjects evaluated 1 year after injury Fraser et al., 1988 Comparison of Prospective cohort 48 of 102 closed head-injured patients selected from Behavioral All 48 followed people with TBI who Outcome of Head Injury study who were working for more than 4 to 1 year after had returned to work h/day for at least 5 mo before injury; 102 friend controls; no pre- injury at 1 year and those existing conditions who had not returned in neuropsychologic, psychosocial functioning at 1 mo, 1 year after injury Dikmen et al., Neuropsychologic Prospective cohort 31 adults with moderate or severe head injury investigated over 2 Subgroup 1990 outcomes years (subgroup of sample of 102 consecutive head injured patients analyses based selected from Behavioral Outcome of Head Injury study); 102 on small samples noninjured friend controls 31 of 46 (67%) eligible subjects from Behavioral Outcome of Head Injury study followed to 2 years after injury; no significant differences between 31 who completed 2-year followup and 15 who were lost on demographics and neuropsychologic measures at 1 and 12 mo after injury 159
160 Population (Where Appropriate) Enrolled Reference Purpose Study Design Eligible Located (Response Rate) Comments Dikmen et al., Psychosocial Prospective cohort 31 adults with moderate or severe head injury investigated over 2 31 of 46 (67%) 1993 outcomes years (subgroup of sample of 102 consecutive head injured patients eligible from selected from Behavioral Outcome of Head Injury study), 102 Behavioral noninjured friend controls Outcome of Head Injury study followed to 2 years after injury; no significant differences between 31 who completed 2-year followup and 15 who were lost on demographics or neurologic severity indexes McLean et al., Psychosocial Prospective cohort 102 consecutive head injured patients from Behavioral Outcome of Followed 97 of 1993 outcomes; Sickness Head Injury study, 102 noninjured friend controls 102 (95%) to 1- Impact Profile; year after injury; Head Injury Subgroup Symptom Checklist; analyses based Modified Function on small samples Status Index Dikmen et al., Time to return to Prospective cohort 366 head injured individuals from 3 prospective, longitudinal studies Head injured and 1994 work (Behavioral Outcome of Head Injury, Patient Characteristics and Head controls similar Injury Outcome, Dilantin Prophylaxis of Post-Traumatic Seizures) on demographics, preinjury Mean age, 30 years; mean education, 12 years; 77% male, 89% white; employment 75% working over 20 h/week before injury status, types of jobs held; 95 trauma controls from Patient Characteristics and Head Injury Results presented Outcome study; mean age, 31 years; mean education, 12 years; 75% as weighted male; 81% white; 78% working over 20 h/week before injury averages to adjust for differences in eligibility criteria between studies
Population (Where Appropriate) Enrolled Reference Purpose Study Design Eligible Located (Response Rate) Comments Haaland et al., Motor skills Prospective cohort 40 patients selected from 102 consecutive head-injured patients 40 of 58 selected 1994 selected from Behavioral Outcome of Head Injury study who did not with no have peripheral upper body injuries, 88 healthy friend controls peripheral injuries, had complete data at 1 mo and 1 year after injury; 18 cases presumably excluded primarily because neurologically too impaired to be tested at 1 mo Dikmen et al., Preinjury drinking, Prospective cohort 197 head-injured patients from Patient Characteristics and Head Injury 179 of 197 (91%) 1995a blood alcohol level; Outcome study followed until 1 preinjury, postinjury year after injury; patterns of alcohol 89 also use participated in Dilantin Prophylaxis of Post-Traumatic Seizures study and told by study nurse that they should not drink alcohol Dikmen et al., Neuropsychologic Prospective cohort 436 adult head-injured patients recruited at time of injury in 3 Study subjects 1995b outcomes prospective, longitudinal studies (Behavioral Outcome in Head Injury, included 85% of Patient Characteristics and Head Injury Outcome, Dilantin 514 subjects Prophylaxis of Post-Traumatic Seizures); 121 general trauma controls recruited from 3 enrolled as part of Patient Characteristics and Head Injury Outcome studies; results as study weighted averages to adjust for differences in 161 eligibility criteria
162 Population (Where Appropriate) Enrolled Reference Purpose Study Design Eligible Located (Response Rate) Comments Dikmen et al., Global outcome, Prospective cohort 466 subjects with TBI selected from 3 prospective, longitudinal 91% of 514 1995c independent living, studies (Behavioral Outcome in Head Injury, Patient Characteristics subjects followed employment, income, and Head Injury Outcome, and Dilantin Prophylaxis of Post- to 1 year after Sickness Impact Traumatic Seizures) injury Profile Results presented 124 trauma controls who had bodily injury other than to head as weighted averages to 88 friend controls, friends of TBI patients, with no pre-existing adjust for conditions differences in eligibility criteria between studies Doctor et al., 2005 Employment status Prospective cohort 418 TBI working before injury from 4 longitudinal investigations 374 of 418 (89%) enrolled 1980â1994 (Behavioral Outcome in Head Injury, Patient followed to 1 Characteristics and Head Injury Outcome, Dilantin Prophylaxis of year after injury Post-Traumatic Seizures, Valproate Prophylaxis of Post-Traumatic Seizures) Responded or Subgroup (n= Contacted or Enrolled Type of Study or Date(s) of Eligible Located (% (Response Reference Eligible Population Methods Enrollment Subjects) of Eligible) Rate) Comments Roberts (Radcliffe, UK) Studies Roberts, 548 people (from total Prospective 1948â1961 11 lost to 1979 population of 7,000) 5â83 followup, 206 years old who sustained TBI died, leaving and remained unconscious 331 surviving or amnestic > 1 week patients (291 admitted to Radcliffe from Infirmary, Oxford: 479 consecutive admitted directly of accident series, 40 from (consecutive series), 69 selected series) transfers from Addenbrook Hospital, Cambridge
Population (Where Appropriate) Enrolled Reference Purpose Study Design Eligible Located (Response Rate) Comments Roberts, Hypothalamic, pituitary Prospective 291 patients from consecutive series 1979 dysfunction Roberts, Positional vertigo, Prospective 291 patients from consecutive series 1979 headaches Roberts, Epilepsy Prospective 291 patients from consecutive series 1979 Lewin et al., Epilepsy, mortality Prospective, 291 patients from consecutive series, 75 patients in whom cause of 1979 retrospective death was determined Responded or Subgroup (n= Contacted or Enrolled Type of Study or Date(s) of Eligible Located (% (Response Reference Eligible Population Methods Enrollment Subjects) of Eligible) Rate) Comments Jennett (Oxford, Rotterdam, Cardiff, and Manchester) Studies Jennett and Oxford series: 1,000 head- Prospective, November 1948â 821 unselected Lewin, 1960 injured patients with at least retrospective February 1952 patients brief period of admitted unconsciousness directly from accident site; 179 selected patients transferred from other hospitalsâ these cases were considered more severe and complicated 163
164 Population (Where Appropriate) Enrolled Reference Purpose Study Design Eligible Located (Response Rate) Comments Jennett, Epilepsy Prospective 189 patients from Oxford series, cases from Lewin; 150 patients from 1969 Glasgow series epileptic within 8 weeks after injury; 73 patients with missile injuries as comparison group; 333 patients 1 year after injury, 219 patients 4 years after injury with depressed fractures from Oxford and Glasgow series Jennett, Epilepsy Prospective 381 patients who had blunt head injuries followed by early epilepsy (n 1962 = 139), late epilepsy (n = 282) drawn from Oxford series (additional patients captured outside study dates), Manchester and Cardiff, England Jennett, Epilepsy Prospective Patients with known risk factors for late epilepsyâearly epilepsy, 1973 intracranial hematoma (evacuation within 14 days of injury), depressed fractureâdrawn from Oxford series, Glasgow series; 250 patients with depressed fractures from Rotterdam Jennett, Epilepsy Prospective Summary of previous data and findings 1975 Responded or Subgroup (n= Contacted or Enrolled Type of Study or Date(s) of Eligible Located (% (Response Reference Eligible Population Methods Enrollment Subjects) of Eligible) Rate) Comments Football Players: Guskiewicz et al. 2005, 2007 Guskiewicz All 3,683 living members of Retrospective cohort 2001â2002 3,683 2,552 2,552 et al., 2005 National Football League Retired Players Association Population (Where Appropriate) Enrolled Reference Purpose Study Design Eligible Located (Response Rate) Comments Guskiewicz Depression Retrospective cohort 2,552 (69%) responded to questionnaires et al., 2007
Responded or Subgroup Contacted or Enrolled Type of Study or Date(s) of (n= Eligible Located (% (Response Reference Eligible Population Methods Enrollment Subjects) of Eligible) Rate) Comments Boxing Studies: Porter et al., 1996, 2003 Porter and Male boxers in amateur Prospective, 1991â1992 53 53 20 selected Many lost because Fricker, boxing clubs in Ireland 16â observational randomly of strict exclusion 1996 25 years old; subjects had (38%) criteria; study of to complete minimum of 40 boxing, not brain bouts injury; flawed comparison group in that controls also had concussion; differentiation in rate of concussion in cases, controls Exclusion criteria: excess alcohol consumption (> 20 standard drinks/week or > 4 drinks/day) Population (Where Appropriate) Enrolled Reference Purpose Study Design Eligible Located (Response Rate) Comments Porter, 2003 Neuropsychologic Prospective, 20 male boxers from amateur boxing clubs in Ireland 16â25 years old; See comments impairment observational subjects had to complete a minimum of 40 bouts above 1 case and 2 controls lost to followup NOTE: ALS = amyotrophic lateral sclerosis, CHSA = Canadian Study of Health and Aging, CT = computed tomography, DSM-IV = Diagnostic and Statistical Manual of Mental Disorders, 4th ed., ED = emergency department, GCS = Glasgow Coma Scale, GSW = gunshot wound, ICD-10 = International Statistical Classification of Diseases and Health Related Problems, 10th revision, LLI = lower-limb injury, LOC = loss of consciousness, MA = Massachusetts, MN = Minnesota, MS = multiple sclerosis, MTBI = mild traumatic brain injury, MVA = motor vehicle accident, PCS = postconcussion syndrome, PD = Parkinson disease, PPCSC = predictors of postconcussion symptom complex, PTA = posttraumatic amnesia, PTSD = posttraumatic stress disorder, TBI = traumatic brain injury, TCDB = Traumatic Coma Databank, UK = United Kingdom, US = United States, VA = Veterans Affairs, VHIS = Vietnam Head Injury Study, WA = Washington. . 165
166 GULF WAR AND HEALTH REFERENCES Achte, K., L. Jarho, T. Kyykka, and E. Vesterinen. 1991. Paranoid disorders following war brain damage. Preliminary report. Psychopathology 24(5):309â315. Achte, K. A., E. Hillbom, and V. Aalberg. 1969. Psychoses following war brain injuries. Acta Psychiatrica Scandinavica 45(1):1â18. Annegers, J. F., J. D. Grabow, R. V. Groover, E. R. Laws, Jr., L. R. Elveback, and L. T. Kurland. 1980. Seizures after head trauma: A population study. Neurology 30(7 Pt 1):683â689. Annegers, J. F., W. A. Hauser, S. P. Coan, and W. A. Rocca. 1998. A population-based study of seizures after traumatic brain injuries. New England Journal of Medicine 338(1):20â24. Annegers, J. F., W. A. Hauser, J. R. Lee, and W. A. Rocca. 1995. Incidence of acute symptomatic seizures in Rochester, Minnesota, 1935â1984. Epilepsia 36(4):327â333. Annegers, J. F., E. R. Laws, Jr., L. T. Kurland, and J. D. Grabow. 1979. Head trauma and subsequent brain tumors. Neurosurgery 4(3):203â206. Bower, J. H., D. M. Maraganore, B. J. Peterson, S. K. McDonnell, J. E. Ahlskog, and W. A. Rocca. 2003. Head trauma preceding PD: A case-control study. Neurology 60(10):1610â 1615. Brown, A. W., C. L. Leibson, J. F. Malec, P. K. Perkins, N. N. Diehl, and D. R. Larson. 2004. Long-term survival after traumatic brain injury: A population-based analysis. Neurorehabilitation 19(1):37â43. Brown, A. W., J. F. Malec, N. N. Diehl, J. Englander, and D. X. Cifu. 2007. Impairment at rehabilitation admission and 1 year after moderate-to-severe traumatic brain injury: A prospective multi-centre analysis. Brain Injury 21(7):673â680. Bryant, R. A., and A. G. Harvey. 1998. Relationship between acute stress disorder and posttraumatic stress disorder following mild traumatic brain injury. American Journal of Psychiatry 155(5):625â629. âââ. 1999a. The influence of traumatic brain injury on acute stress disorder and post- traumatic stress disorder following motor vehicle accidents. Brain Injury 13(1):15â22. âââ. 1999b. Postconcussive symptoms and posttraumatic stress disorder after mild traumatic brain injury. Journal of Nervous and Mental Disease 187(5):302â305. Caveness, W. F. 1963. Onset and cessation of fits following craniocerebral trauma. Journal of Neurosurgery 20:570â583. âââ. 1966. Posttraumatic sequelae: Chapter 17. In Head Injury: Conference Proceedings, edited by W. F. Caveness and A. E. Walker. Philadelphia, PA: Lippincott. Pp. 209â219. âââ. 1972. Vietnam Registry of Head and Spinal Cord Injuries. Report to the Surgeon General, USN, USA, and USAF. Caveness, W. F., A. E. Walker, and P. B. Ascroft. 1962. Incidence of posttraumatic epilepsy in Korean veterans as compared with those from World War I and World War II. Journal of Neurosurgery 19:122â129. CDC (Centers for Disease Control and Prevention). 1989. Health Status of Vietnam Veterans: Vol. IV. Psychological and Neuropsychological Evaluation. Atlanta, GA: Centers for Disease Control and Prevention.
MAJOR COHORT STUDIES 167 Chandra, V., E. Kokmen, B. S. Schoenberg, and C. Beard. 1989. Head trauma with loss of consciousness as a risk factor for Alzheimer's disease. Neurology 39(12):1576â1578. Chesnut, R. M., S. B. Marshall, J. Piek, B. A. Blunt, M. R. Klauber, and L. F. Marshall. 1993. Early and late systemic hypotension as a frequent and fundamental source of cerebral ischemia following severe brain injury in the traumatic coma data bank. Acta NeurochirurgicaâSupplementum 59:121â125. Cifu, D. X., J. S. Kreutzer, J. H. Marwitz, M. Miller, G. M. Hsu, R. T. Seel, J. Englander, W. M. High, Jr., and R. Zafonte. 1999. Etiology and incidence of rehospitalization after traumatic brain injury: A multicenter analysis. Archives of Physical Medicine and Rehabilitation 80(1):85â90. Corkin, S., T. J. Rosen, E. V. Sullivan, and R. A. Clegg. 1989. Penetrating head injury in young adulthood exacerbates cognitive decline in later years. Journal of Neuroscience 9(11):3876â 3883. Corkin, S., E. V. Sullivan, and F. A. Carr. 1984. Prognostic factors for life expectancy after penetrating head injury. Archives of Neurology 41(9):975â977. CSHA (The Canadian Study of Health and Aging). 1994. The Canadian Study of Health and Aging: Risk factors for Alzheimerâs disease in Canada. Neurology 44(11):2073â2080. âââ. 2008. Canadian Study of Health and Aging. http://www.csha.ca/contact_us.asp. (accessed September 3, 2008). Dacey, R., S. Dikmen, N. Temkin, A. McLean, G. Armsden, and H. R. Winn. 1991. Relative effects of brain and non-brain injuries on neuropsychological and psychosocial outcome. Journal of Trauma-Injury Infection and Critical Care 31(2):217â222. Dikmen, S., J. Machamer, and N. Temkin. 1993. Psychosocial outcome in patients with moderate to severe head injury: 2-year follow-up. Brain Injury 7(2):113â124. Dikmen, S., J. Machamer, N. Temkin, and A. McLean. 1990. Neuropsychological recovery in patients with moderate to severe head injury: 2 year follow-up. Journal of Clinical and Experimental Neuropsychology 12(4):507â519. Dikmen, S., A. McLean, and N. Temkin. 1986. Neuropsychological and psychosocial consequences of minor head injury. Journal of Neurology, Neurosurgery and Psychiatry 49(11):1227â1232. Dikmen, S., N. Temkin, A. McLean, A. Wyler, and J. Machamer. 1987. Memory and head injury severity. Journal of Neurology, Neurosurgery and Psychiatry 50(12):1613â1618. Dikmen, S. S., J. E. Machamer, D. M. Donovan, H. R. Winn, and N. R. Temkin. 1995a. Alcohol use before and after traumatic head injury. Annals of Emergency Medicine 26(2):167â176. Dikmen, S. S., J. E. Machamer, J. M. Powell, and N. R. Temkin. 2003. Outcome 3 to 5 years after moderate to severe traumatic brain injury. Archives of Physical Medicine and Rehabilitation 84(10):1449â1457. Dikmen, S. S., J. E. Machamer, H. Winn, and N. R. Temkin. 1995b. Neuropsychological outcome at 1-year post head injury. Neuropsychology 9(1):80â90. Dikmen, S. S., J. E. Machamer, H. R. Winn, G. D. Anderson, and N. R. Temkin. 2000. Neuropsychological effects of valproate in traumatic brain injury: A randomized trial. Neurology 54(4):895â902.
168 GULF WAR AND HEALTH Dikmen, S. S., B. L. Ross, J. E. Machamer, and N. R. Temkin. 1995c. One year psychosocial outcome in head injury. Journal of the International Neuropsychological Society 1(1):67â77. Dikmen, S. S., N. R. Temkin, J. E. Machamer, A. L. Holubkov, R. T. Fraser, and H. R. Winn. 1994. Employment following traumatic head injuries. Archives of Neurology 51(2):177â186. Dikmen, S. S., N. R. Temkin, B. Miller, J. Machamer, and H. R. Winn. 1991. Neurobehavioral effects of phenytoin prophylaxis of posttraumatic seizures. JAMA 265(10):1271â1277. Doctor, J. N., J. Castro, N. R. Temkin, R. T. Fraser, J. E. Machamer, and S. S. Dikmen. 2005. Workers' risk of unemployment after traumatic brain injury: A normed comparison. Journal of the International Neuropsychological Society 11(6):747â752. Evans, J. H. 1962. Post-traumatic epilepsy. Neurology 12:665â674. Flaada, J. T., C. L. Leibson, J. N. Mandrekar, N. Diehl, P. K. Perkins, A. W. Brown, and J. F. Malec. 2007. Relative risk of mortality after traumatic brain injury: A population-based study of the role of age and injury severity. Journal of Neurotrauma 24(3):435â445. Fraser, R., S. Dikmen, A. McLean, B. Miller, and N. Temkin. 1988. Employability of head injury survivors: First year post-injury. Rehabilitation Counseling Bulletin 31(4):276â288. Grafman, J. 2007. Vietnam Head Injury Study Phase III: A 30-Year Post-Injury Follow-up Study. Rockville, MD: The Henry M. Jackson Foundation for the Advancement of Military Medicine. Grafman, J., A. M. Salazar, H. Weingartner, and D. Amin. 1986. Face memory and discrimination: An analysis of the persistent effects of penetrating brain wounds. International Journal of Neuroscience 29(1-2):125â139. Grafman, J., K. Schwab, D. Warden, A. Pridgen, H. R. Brown, and A. M. Salazar. 1996. Frontal lobe injuries, violence, and aggression: A report of the Vietnam Head Injury Study. Neurology 46(5):1231â1238. Groswasser, Z., G. Reider II, K. Schwab, A. K. Ommaya, A. Pridgen, H. R. Brown, R. Cole, and A. M. Salazar. 2002. Quantitative imaging in late TBI. Part II: Cognition and work after closed and penetrating head injury: A report of the Vietnam Head Injury Study. Brain Injury 16(8):681â690. Guskiewicz, K. M., S. W. Marshall, J. Bailes, M. McCrea, R. C. Cantu, C. Randolph, and B. D. Jordan. 2005. Association between recurrent concussion and late-life cognitive impairment in retired professional football players. Neurosurgery 57(4):719â726. Guskiewicz, K. M., S. W. Marshall, J. Bailes, M. McCrea, H. P. Harding, Jr., A. Matthews, J. R. Mihalik, and R. C. Cantu. 2007. Recurrent concussion and risk of depression in retired professional football players. Medicine and Science in Sports and Exercise 39(6):903â909. Guskiewicz, K. M., S. W. Marshall, S. P. Broglio, R. C. Cantu, and D. T. Kirkendall. 2002. No evidence of impaired neurocognitive performance in collegiate soccer players. American Journal of Sports Medicine 30(2):157â162. Haaland, K. Y., N. Temkin, G. Randahl, and S. Dikmen. 1994. Recovery of simple motor skills after head injury. Journal of Clinical and Experimental Neuropsychology 16(3):448â456. Harrison-Felix, C., G. Whiteneck, M. DeVivo, F. M. Hammond, and A. Jha. 2004. Mortality following rehabilitation in the traumatic brain injury model systems of care. Neurorehabilitation 19(1):45â54.
MAJOR COHORT STUDIES 169 Harrison-Felix, C., G. Whiteneck, M. J. Devivo, F. M. Hammond, and A. Jha. 2006. Causes of death following 1 year postinjury among individuals with traumatic brain injury. Journal of Head Trauma Rehabilitation 21(1):22â33. Harvey, A. G., and R. A. Bryant. 2000. Two-year prospective evaluation of the relationship between acute stress disorder and posttraumatic stress disorder following mild traumatic brain injury. American Journal of Psychiatry 157(4):626â628. Jennett, B. 1973. Epilepsy after non-missile head injuries. Scottish Medical Journal 18(1):8â13. Jennett, B. W. 1969. Epilepsy after blunt (nonmissile) head injury: Chapter 22. In The Late Effects of Head Injury, edited by A. E. Walker, W. F. Caveness, and M. Critchley. Springfield, IL: Charles C. Thomas. Pp. 201â214. Jennett, W. B. 1962. Epilepsy after Blunt Head Injury. London: William Heinemann. âââ. 1975. Epilepsy after Non-Missle Head Injuries. Chicago: Year Book Medical Publishers. Jennett, W. B., and W. Lewin. 1960. Traumatic epilepsy after closed head injuries. Journal of Neurology, Neurosurgery and Psychiatry 23:295â301. Koenigs, M., E. D. Huey, V. Raymont, B. Cheon, J. Solomon, E. M. Wassermann, and J. Grafman. 2007. Focal brain damage protects against post-traumatic stress disorder in combat veterans. Nature Neuroscience 11(2):232â237. Kraft, J. F., K. A. Schwab, A. M. Salazar, and H. R. Brown. 1993. Occupational and educational achievements of head injured Vietnam veterans at 15-year follow-up. Archives of Physical Medicine and Rehabilitation 74(6):596â601. Kurland, L. T. 1994. Trauma and multiple sclerosis. Annals of Neurology 36 Suppl:S33âS37. Levin, H. S., H. M. Eisenberg, H. E. Gary, A. Marmarou, M. A. Foulkes, J. A. Jane, L. F. Marshall, and S. M. Portman. 1991a. Intracranial hypertension in relation to memory functioning during the first year after severe head injury. Neurosurgery 28(2):196â199. Levin, H. S., C. Saydjari, H. M. Eisenberg, M. Foulkes, L. F. Marshall, R. M. Ruff, J. A. Jane, and A. Marmarou. 1991b. Vegetative state after closed-head injury. A Traumatic Coma Data Bank report. Archives of Neurology 48(6):580â585. Lewin, W., T. F. Marshall, and A. H. Roberts. 1979. Long-term outcome after severe head injury. British Medical Journal 2(6204):1533â1538. Lindsay, J., D. Laurin, R. Verreault, R. Hebert, B. Helliwell, G. B. Hill, and I. McDowell. 2002. Risk factors for Alzheimer's disease: A prospective analysis from the Canadian Study of Health and Aging. American Journal of Epidemiology 156(5):445â453. Lu, J., A. Marmarou, S. Choi, A. Maas, G. Murray, and E. W. Steyerberg. 2005. Mortality from traumatic brain injury. Acta NeurochirurgicaâSupplement 95:281â285. Luis, C. A., R. D. Vanderploeg, and G. Curtiss. 2003. Predictors of postconcussion symptom complex in community dwelling male veterans. Journal of the International Neuropsychological Society 9(7):1001â1015. Machamer, J., N. Temkin, and S. Dikmen. 2002. Significant other burden and factors related to it in traumatic brain injury. Journal of Clinical and Experimental Neuropsychology 24(4):420â 433.
170 GULF WAR AND HEALTH Machamer, J., N. Temkin, R. Fraser, J. N. Doctor, and S. Dikmen. 2005. Stability of employment after traumatic brain injury. Journal of the International Neuropsychological Society 11(7):807â816. Marwitz, J. H., D. X. Cifu, J. Englander, and W. M. High, Jr. 2001. A multi-center analysis of rehospitalizations five years after brain injury. Journal of Head Trauma Rehabilitation 16(4):307â317. Masson, F., P. Maurette, L. R. Salmi, J. F. Dartigues, J. Vecsey, J. M. Destaillats, and P. Erny. 1996. Prevalence of impairments 5 years after a head injury, and their relationship with disabilities and outcome. Brain Injury 10(7):487â497. Masson, F., J. Vecsey, L. R. Salmi, J. F. Dartigues, P. Erny, and P. Maurette. 1997. Disability and handicap 5 years after a head injury: A population-based study. Journal of Clinical Epidemiology 50(5):595â601. Matser, E. J., A. Kessels, B. Jordan, M. Lezak, and J. Troost. 1998. Chronic traumatic brain injury in professional soccer players. Neurology 51(3):791â796. McLean, A., Jr., S. S. Dikmen, and N. R. Temkin. 1993. Psychosocial recovery after head injury. Archives of Physical Medicine and Rehabilitation 74(10):1041â1046. Meirowsky, A. M. 1982. Notes on posttraumatic epilepsy in missile wounds of the brain. Military Medicine 147(8):632â634. Melton, L. J., 3rd. 1996. History of the Rochester Epidemiology Project. Mayo Clinic Proceedings 71(3):266â274. National Naval Medical Center. 2008. Vietnam Head Injury Study: Phase III. http://www.bethesda.med.navy.mil/Professional/Research/Vietnam_Head_Injury.aspx (accessed August 22, 2008). NDSC (National Data and Statistical Center). 2008. Traumatic Brain Injury Model Systems. http://www.tbindsc.org/ (accessed June 24, 2008). Nemetz, P. N., C. Leibson, J. M. Naessens, M. Beard, E. Kokmen, J. F. Annegers, and L. T. Kurland. 1999. Traumatic brain injury and time to onset of Alzheimerâs disease: A population-based study. American Journal of Epidemiology 149(1):32â40. Pagulayan, K. F., N. R. Temkin, J. Machamer, and S. S. Dikmen. 2006. A longitudinal study of health-related quality of life after traumatic brain injury. Archives of Physical Medicine and Rehabilitation 87(5):611â618. Porter, M. D. 2003. A 9-year controlled prospective neuropsychologic assessment of amateur boxing. Clinical Journal of Sport Medicine 13(6):339â352. Porter, M. D., and P. A. Fricker. 1996. Controlled prospective neuropsychological assessment of active experienced amateur boxers. Clinical Journal of Sport Medicine 6(2):90â96. Raymont, V., A. Greathouse, K. Reding, R. Lipsky, A. Salazar, and J. Grafman. 2008. Demographic, structural and genetic predictors of late cognitive decline after penetrating head injury. Brain 131:543â558. Rish, B. L., J. D. Dillon, and G. H. Weiss. 1983. Mortality following penetrating craniocerebral injuries. An analysis of the deaths in the Vietnam head injury registry population. Journal of Neurosurgery 59(5):775â780. Roberts, A. H. 1979. Severe Accidental Head Injury. London: Macmillan Press.
MAJOR COHORT STUDIES 171 Rutherford, A., R. Stephens, D. Potter, and G. Fernie. 2005. Neuropsychological impairment as a consequence of football (soccer) play and football heading: Preliminary analyses and report on university footballers. Journal of Clinical and Experimental Neuropsychology 27(3):299â 319. Salazar, A. M., J. Grafman, S. Schlesselman, S. C. Vance, J. Mohr, M. Carpenter, P. Pevsner, C. Ludlow, and H. Weingartner. 1986. Penetrating war injuries of the basal forebrain: Neurology and cognition. Neurology 36(4):459â465. Salazar, A. M., B. Jabbari, S. C. Vance, J. Grafman, D. Amin, and J. D. Dillon. 1985. Epilepsy after penetrating head injury. I. Clinical correlates: A report of the Vietnam Head Injury Study. Neurology 35(10):1406â1414. Schwab, K., J. Grafman, A. M. Salazar, and J. Kraft. 1993. Residual impairments and work status 15 years after penetrating head injury: Report from the Vietnam Head Injury Study. Neurology 43(1):95â103. Seel, R. T., J. S. Kreutzer, M. Rosenthal, F. M. Hammond, J. D. Corrigan, and K. Black. 2003. Depression after traumatic brain injury: A National Institute on Disability and Rehabilitation Research model systems multicenter investigation. Archives of Physical Medicine and Rehabilitation 84(2):177â184. Singer, R. B. 2001. Incidence of seizures after traumatic brain injuryâa 50-year population survey. Journal of Insurance Medicine (Seattle) 33(1):42â45. Siva, A., K. Radhakrishnan, L. T. Kurland, P. C. O'Brien, J. W. Swanson, and M. Rodriguez. 1993. Trauma and multiple sclerosis: A population-based cohort study from Olmsted County, Minnesota. Neurology 43(10):1878â1882. Straume-Naesheim, T. M., T. E. Andersen, J. Dvorak, and R. Bahr. 2005. Effects of heading exposure and previous concussions on neuropsychological performance among Norwegian elite footballers. British Journal of Sports Medicine 39(Suppl 1):i70âi77. Swanson, S., S. Rao, J. Grafman, A. Salazar, and J. Kraft. 1995. The relationship between seizure subtype and interictal personality: Results from the Vietnam Head Injury Study. Brain: A Journal of Neurology 118(1):91â103. Temkin, N. R., S. S. Dikmen, G. D. Anderson, A. J. Wilensky, M. D. Holmes, W. Cohen, D. W. Newell, P. Nelson, A. Awan, and H. R. Winn. 1999a. Valproate therapy for prevention of posttraumatic seizures: A randomized trial. Journal of Neurosurgery 91(4):593â600. Temkin, N. R., S. S. Dikmen, A. J. Wilensky, J. Keihm, S. Chabal, and H. R. Winn. 1990. A randomized, double-blind study of phenytoin for the prevention of post-traumatic seizures. New England Journal of Medicine 323(8):497â502. Temkin, N. R., R. K. Heaton, I. Grant, and S. S. Dikmen. 1999b. Detecting significant change in neuropsychological test performance: A comparison of four models. Journal of the International Neuropsychological Society 5(4):357â369. Teuber, H. L., and S. Weinstein. 1954. Performance on a formboard-task after penetrating brain injury. Journal of Psychology 38:177â190. Tiret, L., B. Garros, P. Maurette, V. Nicaud, M. Thicoipe, F. Hatton, and P. Erny. 1989. Incidence, causes and severity of injuries in Aquitaine, France: A community-based study of hospital admissions and deaths. American Journal of Public Health 79(3):316â321.
172 GULF WAR AND HEALTH Tiret, L., E. Hausherr, M. Thicoipe, B. Garros, P. Maurette, J. P. Castel, and F. Hatton. 1990. The epidemiology of head trauma in Aquitaine (France), 1986: A community-based study of hospital admissions and deaths. International Journal of Epidemiology 19(1):133â140. Vanderploeg, R. D., G. Curtiss, and H. G. Belanger. 2005. Long-term neuropsychological outcomes following mild traumatic brain injury. Journal of the International Neuropsychological Society 11(3):228â236. Vanderploeg, R. D., G. Curtiss, J. J. Duchnick, and C. A. Luis. 2003. Demographic, medical, and psychiatric factors in work and marital status after mild head injury. Journal of Head Trauma Rehabilitation 18(2):148â163. Vanderploeg, R. D., G. Curtiss, C. A. Luis, and A. M. Salazar. 2007. Long-term morbidities following self-reported mild traumatic brain injury. Journal of Clinical and Experimental Neuropsychology: Official Journal of the International Neuropsychological Society 29(6):585â598. Walker, A. E., and F. Erculei. 1969. Head Injured Men Fifteen Years Later. Springfield, IL: Charles C. Thomas. âââ. 1970. Post-traumatic epilepsy 15 years later. Epilepsia 11(1):17â26. Walker, A. E., H. K. Leuchs, H. Lechtape-Gruter, W. F. Caveness, and C. Kretschman. 1971. Life expectancy of head injured men with and without epilepsy. Archives of Neurology 24(2):95â100. Weinstein, S. 1954. Weight judgment in somethesis after penetrating injury to the brain. Journal of Comparative and Physiological Psychology 47(1):31â35. Weinstein, S., and H. L. Teuber. 1957a. Effects of penetrating brain injury on intelligence test scores. Science 125(3256):1036â1037. âââ. 1957b. The role of preinjury education and intelligence level in intellectual loss after brain injury. Journal of Comparative and Physiological Psychology 50:535â539. Weinstein, S., J. Semmes, L. Ghent, and H. L. Teuber. 1958. Roughness discrimination after penetrating brain injury in man: Analysis according to locus of lesion. Journal of Comparative and Physiological Psychology 51(3):269â275. Weiss, G. H., W. F. Caveness, H. Einsiedel-Lechtape, and M. L. McNeel. 1982. Life expectancy and causes of death in a group of head-injured veterans of World War I. Archives of Neurology 39(12):741â743. Weiss, G. H., D. M. Feeney, W. F. Caveness, D. Dillon, J. P. Kistler, J. P. Mohr, and B. L. Rish. 1983. Prognostic factors for the occurrence of posttraumatic epilepsy. Archives of Neurology 40(1):7â10. Williams, D., J. Annegers, E. Kokmen, P. O'Brien, and L. T. Kurland. 1991. Brain injury and neurologic sequelae: A cohort study of dementia, Parkinsonism, and amyotrophic lateral sclerosis. Neurology 41(10):1554â1557.