Presidential Advisory Committee on Gulf War Veterans' Illnesses
Final Report

Chapter Three

NATURE OF GULF WAR VETERANS' ILLNESS

Quantitative documentation and analysis of the extent of Gulf War veterans' illnesses remains elusive at this writing. Nevertheless, the clinical programs of VA and DOD, along with preliminary information from several federally funded epidemiologic studies, provide some data that can begin to place Gulf War veterans' illnesses in context. This chapter provides an overview of available data and the Committee's findings and recommendations about the nature of Gulf War veterans' illnesses.

DATA FROM CLINICAL PROGRAMS

As of October 1996, approximately 62,000 individuals had completed physical examinations in VA's Persian Gulf Health Registry, and VA has reviewed results for the first 52,216 veterans.109 More than 34,000 individuals had requested an examination in DOD's CCEP as of October 1996; information on the first 18,075 military personnel has been reported by DOD.277

Information derived from these data sets has clinical utility, and DOD and VA have used the information to address several concerns from a descriptive perspective. The data also have provided guidance in the formulation of certain epidemiologic research approaches. However, results from analyzing the Registry and the CCEP-two self-selected case series-cannot be generalized to the entire population of Gulf War veterans. As described later, scientifically valid, generalizable knowledge is the expected outcome of epidemiologic research currently underway.

Data from DOD's CCEP and VA's Registry

Gulf War veterans who have participated in the CCEP and Registry represent a broad cross-section of service members who deployed to the Gulf War (table 3-1). To provide a qualitative snapshot of the clinical status of the Gulf War veterans who have participated in the government's clinical evaluation programs, the following sections summarize information-based on material published by DOD and VA-for several key descriptive parameters.109,277

Reported symptoms. CCEP participants report a broad range of symptoms that span a variety of organ systems (table 3-2). The most common primary symptoms reported are joint pain, fatigue, headache, rash, and memory loss. Ten percent of participants are asymptomatic. The most frequent symptoms in the CCEP also are common in the general adult population in the United States.

The most common symptoms reported by VA Registry participants (table 3-2) nearly match the most common symptoms reported by CCEP participants. Asymptomatic individuals comprise about 12 percent of the Registry population.

Diagnoses. Approximately 10 percent of CCEP participants are found to be healthy. The other most common primary diagnostic categories of CCEP participants are psychological conditions; musculoskeletal system diseases (MSDs); and Symptoms, Signs, and Ill-defined Conditions (SSIDC) (table 3-3). Combined, the three categories account for more than 50 percent of primary diagnoses. Other primary diagnoses in the CCEP do not concentrate in any single organ system.

The most common diagnostic categories in VA's Registry are the same as in the CCEP: psychological conditions, MSDs, and SSIDC (table 3-3). As with CCEP participants, the diagnoses do not center in a single organ system beyond these three categories.

Morbidity/Disability. To approximate disability due to illness, DOD asks CCEP participants about the number of work days missed due to illness in the 90 days prior to the initial examination. Most individuals (80 percent) reported no missed days of work due to illness during this time period. Among those reporting one or more lost work days, the median number of lost days was five. This level of lost workdays in the past 90 days exceeds that found for the general U.S. population of civilian employees.292

CCEP data on lost workdays, however, cannot be viewed as an estimate of the overall prevalence of disability due to Gulf War service. Many individuals-some of whom could have disabilities-have left active service since the Gulf War and are not eligible for participation in the CCEP. No estimates of the degree of disability have been reported for individuals enrolled in VA's Registry.

Musculoskeletal system diseases. MSDs account for 18.3 percent of primary diagnoses in CCEP participants. DOD reports the occurrence of MSDs in the CCEP is about two times higher in male CCEP participants and three times higher in female CCEP participants than the rate of occurrence of MSDs in the general U.S. population aged 20 to 40 years.277 Whether the rate of MSDs for CCEP participants differs from that for the general military population is unknown, however, due to the paucity of baseline information on the health status of active duty personnel.

Military personnel must maintain certain levels of physical fitness, and many are required to participate in demanding physical training programs, placing considerable stress on joints and muscles. DOD reports that the majority of MSDs diagnosed in CCEP participants are wear-and-tear disorders-e.g., recurrent strains, sprains, and degenerative arthritis due to trauma on a joint-that could be expected in a physically active populace; occupational and recreational overuse injuries also frequently occur as a consequence of the physical activities of military training and operations.277

MSDs are the most prevalent diagnostic category among participants in VA's Registry.109 Additionally, as of September 1995 more than 15,000 Gulf War veterans had been admitted for inpatient treatment to a VA hospital; MSDs account for 21.3 percent of the total diagnoses received by these inpatients.

Infectious diseases. Infectious diseases are not a frequent cause of illness among CCEP participants. DOD reports 470 individuals have a primary diagnosis of an infectious disease, and about half of these are infections of the skin due to fungi that are common in the United States. VA's Registry reports similarly low occurrences of infectious diseases among its participants; 7.1 percent have a primary or secondary diagnosis of an infectious disease.109 For infectious diseases, athlete's foot, a fungal infection of the skin, was one of the most common-occurring in 1.4 percent of the Registry population.

Infectious diseases have affected a variety of organ systems in Gulf War veterans without any observable patterns. Most infectious diseases identified through the clinical programs are minor conditions common among the general population; such conditions do not explain serious, persistent, systemic complaints. To date, few individuals have demonstrated characteristic physical signs and laboratory abnormalities that indicate a chronic infectious process.65,90,293 Moreover, it is unlikely that Gulf War veterans have infections that have evaded the systematic diagnostic efforts mandated by the standardized protocol.

Cancer. Cancer is rare among CCEP enrollees. A primary diagnosis of cancer has been made in 52 individuals (0.3 percent), and the types and frequencies are shown in table 3-4. Lymphomas are the most frequent cancer diagnosed in CCEP participants; lymphomas also are the most common types of cancer among 20- to 40-year olds in the general U.S. population. The second most frequent cancer diagnosis for CCEP participants is skin cancer-again, one of the most common malignancies in the age-matched U.S. general population. Clinical evaluation through the CCEP identified four individuals with testicular cancer, which is a common type of cancer among young men in the U.S. general population.

Cancer also is rare among individuals in VA's Registry. There does not appear to be an unusual incidence of any specific type of cancer in this population. A primary diagnosis of cancer was made in 226 individuals (0.4 percent; table 3-4).109 The same three types of cancer most common among CCEP participants are the most frequently diagnosed in VA's Registry population: lymphomas, skin cancer, and testicular cancer.

Data from Great Britain and Canada. Great Britain implemented a systematic medical evaluation program for Gulf War veterans in late 1993. To date, about 500 individuals have enrolled in the U.K. medical program, and the clinical results for 284 individuals are available.34 Canada began its medical evaluation program for Gulf War veterans in early 1995, and the clinical information for 49 veterans is available.227 Both programs are thorough, and the procedures resemble Phase II of the U.S. protocol (see chapter 2). Because the proportion of eligible Gulf War veterans who have enrolled in these programs is small, physicians involved in both programs view results as preliminary. The Committee draws no conclusions based on the available data from these two countries.

Great Britain. Approximately 45,000 British troops were deployed to the Gulf War. In late 1993, the Ministry of Defense set up a medical evaluation program for these individuals, and in September 1994 a modified version of the U.S. protocol was adopted. About 20 percent of participants remain on active duty.

Information about the first 284 participants (0.6 percent of British troops deployed) was presented publicly in August 1996.34 Table 3-5 presents the most frequently reported symptoms, which are similar to the symptoms reported by U.S. Gulf War veterans. The most common primary diagnoses in tthe British participants are also common among U.S. Gulf War veterans (table 3-6). Fifteen percent of diagnoses for British participants are coded under the SSIDC category. A small percentage of these cases meet the Oxford criteria for CFS, which are slightly less stringent than CDC's criteria.34 There was a clear relationship between service in the Gulf War and the onset of psychiatric illness in a number of participants in the British evaluation.214

Canada. Canada established its medical evaluation program for Gulf War veterans in early 1995 and had enrolled about 60 veterans as of Summer 1996. A compilation of the evaluations for the first 49 participants (about 1.0 percent of the 4,500 Canadian troops deployed to the Gulf) were presented publicly in August 1996.227

Table 3-7 presents the ten most frequent symptoms among Canadian participants; these symptoms are nearly identical to the most frequent symptoms reported by U.S. Gulf War veterans. Table 3-8 reports the frequencies of the major diagnostic categories in the 49 Canadian participants. Most of the MSD cases in the Canadian population are mechanical low back pain, osteoarthritis, and degenerative disc disease. Eight percent of the diagnoses are SSIDC.227

DATA FROM EPIDEMIOLOGIC STUDIES

As noted in the Committee's Interim Report, epidemiologic studies are crucial for better understanding of the extent and nature of Gulf War veterans' illnesses. Any large population will include people who are experiencing a variety of different illnesses. While clinical programs provide valuable information, they cannot provide answers to whether and how rates of illnesses or death in the whole Gulf War veteran population differ from those that would be expected in any similar large population. Answers to these questions could focus attention on the most useful and relevant interventions or treatments.

Carefully designed epidemiology studies are time-consuming, and many were not initiated until several years after the Gulf War. Thus, many important studies addressing the general characteristics of Gulf War veterans' illnesses are still underway, and preliminary results are not yet publicly available. These include such studies as VA's National Health Survey and the Health Assessment of Persian Gulf War Veterans from Iowa, as well as several studies at VA's Environmental Hazards Centers. The following sections review data from completed epidemiology studies for which results have been published in the peer-reviewed literature or government reports, or for which preliminary data have been released publicly by the investigators (table 3-9).

Mortality Studies

Mortality studies examine deaths among the selected population. Research in this area has focused on deaths that occurred during and after the Gulf War.

Mortality during the Gulf War. Two epidemiologic studies have been completed on service member deaths that occurred during the Gulf War. One study reported on mortality during the six weeks from January 17, 1991 through February 28, 1991. It found that battle and nonbattle casualty rates were the lowest experienced by the United States in any major 20th century conflict.85

A more detailed mortality study covered the time period of August 1, 1990 through July 31, 1991, spanning the entire Operations Desert Shield/Desert Storm campaign and the post-war recovery.342,343 This study examined the cause of each death and compared cause-specific mortality rates of troops in the Gulf region with those of U.S. troops serving elsewhere during this time period. Of the 372 deaths among active duty service members stationed in the region during this 1-year period, 147 (40 percent) were a direct result of combat, 194 (52 percent) resulted from nonbattle injuries, and 30 (8 percent) were the result of illness.* An excess of unintentional injury (accident) deaths-e.g., from motor vehicle and aircraft accidents-was identified in the Gulf campaign participants compared to the nondeployed population. No excess mortality from illness or unexpected/undefined causes was observed, nor were there clusters of the deaths in timing or location.

Mortality since the Gulf War. A study of mortality among Gulf War veterans and a comparison population of era veterans since the Gulf War has been conducted by VA's Environmental Epidemiology Service.107,108 Mortality among all 695,516 military personnel who served in Operations Desert Shield/Desert Storm between August 1990 and April 1991 was compared to 746,291 era veterans, adjusting statistically for branch of service, type of unit, age, sex, and race. During the study's timeframe (Gulf War through September 1993), 1,765 deaths occurred among Gulf War veterans, and 1,729 deaths occurred among the era veterans sampled.

Results of this study indicate that for the 2.4 years following the war, Gulf War veterans had significantly elevated mortality (nine percent higher) compared to era veterans. Excess fatalities were entirely attributable to external causes, including all types of accidents and, in particular, motor vehicle accidents. No excess of suicides, homicides, or deaths from disease-related causes was observed. The risk of death from infectious and parasitic diseases was significantly lower in Gulf War veterans.

When examined separately by sex, both men and women Gulf War veterans had a significant risk of mortality from external causes compared to era veterans. Mortality risks to both male and female Gulf War veterans from disease-related causes were similar and not significantly different from those of era veterans. When compared to the general population, both Gulf War and era veterans had significantly lower mortality rates when adjusted for age, sex, race, and year of death. This finding is consistent with findings in other military populations. Physical screening at entrance, continued standards for physical fitness, and access to medical care contributes to better survival rates among military personnel than observed in the age-matched population as a whole.

Summary of mortality studies. The completed mortality studies reveal no excess of deaths from natural causes during either Operations Desert Shield/Desert Storm or in the two years that followed. Death rates from all illnesses, including infectious diseases and cancers, have been the same or lower in the population deployed to the Gulf than those deployed elsewhere. Death rates from external causes have been elevated among Gulf War veterans.

Elevated mortality from external causes, particularly from motor vehicle accidents, is consistent with trends observed in populations of combat veterans from other wars. Studies of mortality in Vietnam veterans document an increased mortality rate from external causes such as injuries.21,257,290,326 The mechanism underlying the excess deaths due to external causes among combat veterans is not well understood.

Morbidity Studies

To answer fundamental questions about disease prevalence, epidemiologic studies must be population-based, meaning they must draw information from samples representative of the entire population of interest. Little information is available from population-based studies on Gulf War veterans, although several studies are ongoing. For this part of the Final Report, we have reviewed findings from epidemiologic studies on Gulf War veterans' illnesses for which investigators have published results or publicly presented preliminary results and provided written abstracts. The studies undertaken most quickly after reports of illnesses in Gulf War veterans surfaced were investigations of clusters of reported illnesses or analyses of health databases of subgroups of the Gulf War veteran population.

123d Army Reserve Command investigation. In Spring 1992, an interdisciplinary team interviewed and examined 79 members of the 123d Army Reserve Command at Fort Benjamin Harrison, Indiana, after a number of members reported a variety of symptoms.41 After physical and psychiatric examinations, the team found no evidence of an outbreak or cluster of any unique disease process. The most frequently reported symptom was fatigue (71 percent), which most commonly had its onset within several weeks of return from the Gulf. Physical examinations and laboratory screenings yielded limited positive objective findings similar to those seen in nondeployed soldiers. The group was self-selected, and therefore the results cannot be generalized to the larger Reserve or Gulf War veteran population.

Seabee study. A cluster study to investigate health complaints in Gulf War veteran reservists was carried out in the 24th Naval Mobile Construction Battalion between 1993 and 1994.10 Two detachments, in which media reports had indicated a large degree of symptomatic illness, were evaluated with a standard questionnaire and a review of medical records to verify diagnoses and obtain additional information. No physical examinations or laboratory tests were performed. One year later, these detachments and two additional detachments were visited and surveyed again. Types and frequencies of diagnosed illnesses did not seem unusual for this age group, but no control group was analyzed. The symptoms did not suggest a pattern or particular illness to the investigators. As in the previously described study, the study group was self-selected, and therefore the results cannot be generalized to the larger Reserve or Gulf War veteran population.

Study of women veterans. A study on health symptoms in women Air Force Gulf War veterans was carried out with a survey questionnaire from 1991 to 1993; preliminary results have been presented publicly and await review for publication.199-201 A randomized sample of Air Force women from the active duty, Reserve, and Guard were asked to report any conditions or symptoms for which they sought medical care since beginning service in the armed forces. The sample of 525 women included those deployed to the Gulf and those deployed elsewhere during the same time period.

The rates of baseline symptoms (i.e., those experienced prior to the Gulf War) did not differ between the two populations. However, results indicated a higher prevalence after the Gulf War of skin rashes, depression, unintentional weight loss, and headaches reported by those deployed to the Gulf compared with those deployed elsewhere. Reported health problems in general were higher in those deployed to the Gulf, and highest among those reporting they were no longer in the military.

A follow-up survey carried out between 1994 and 1995 on the same group also found higher levels of reported skin rashes and headaches among those deployed to the Gulf region. It also found increased reports of cough, memory problems, lumps or cysts in the breast, and abnormal Pap test results in Gulf-deployed women veterans. Differences between the two groups in reported depression, unintentional weight loss, and insomnia no longer were statistically significant. An additional follow-up survey is planned to see if reported differences persist.

Increases in self-reported health symptoms in this small, but representative, sample of Air Force women are consistent with the increased reports of health symptoms seen in cluster investigations. Since this study did not involve physical examinations or evaluate medical records, it cannot assess the extent to which increased concern or vigilance in Gulf-deployed service members could contribute to higher levels of reported signs and symptoms.

Pennsylvania Air National Guard study. CDC's National Center for Infectious Disease has carried out a study of illnesses reported among Gulf War veterans in a Pennsylvania Air National Guard unit. This three-stage study began in late 1994 as a rapid response to reports of an outbreak of illnesses in the unit. At this time, findings from the first two stages have been published or presented publicly.211,291

The first stage involved standardized interviews and physical examinations of 59 Gulf War veterans reported to be symptomatic. Most frequently reported symptoms were fatigue, joint pain or stiffness, nasal or sinus congestion, diarrhea, gas, difficulty remembering, muscle pains, headaches, abdominal pain, general weakness, and impaired concentration. All study participants reported several symptoms that had persisted at least six months. No consistent abnormalities were identified through standardized physical examination or by review of medical records and laboratory tests.

In the study's second stage, unit members and three comparison units (a total of 3,927 individuals) were surveyed to determine the prevalence of selected symptoms identified in the first stage. In all units, the prevalence of 13 chronic (lasting six months or more) symptoms was significantly higher among individuals who had deployed to the Gulf.

The operational case definition developed for the illness in this population is similar to the definition for CFS recently developed by CDC, but lacks the requirement of severity of symptoms. Criteria for the case-as defined for the purposes of this study-were met in 45 percent of the veterans surveyed who had been deployed to the Gulf, but were also met in 15 percent of the nondeployed veteran respondents and in 12 percent of a San Francisco civilian population surveyed-suggesting the causes of the problems are not unique to Gulf War service. Symptoms were not associated with the place of service in the Gulf, number of deployments to the Gulf, or timing of deployment to the Gulf.

The study's final stage explored associations between having the symptoms defined in the study as being a "case" and selected infectious, behavioral, and environmental risk factors for developing the illness.212 No physical or laboratory abnormalities were associated with being defined as a case. Despite the absence of physical findings, veterans who fit the definition of a severe case had measurable deficits in reported functioning. Veterans in this group also were more likely to meet screening levels for PTSD on the Mississippi Scale for PTSD. Data from this study stage are still preliminary and undergoing additional analyses.

Because the Pennsylvania Air National Guard study relied on volunteers who were a minority of the target population, the potential for bias exists in the research findings. Furthermore, the generalizability of the results from this study to the entire Gulf War veteran population is limited. Subjects were all members of the Air National Guard or Air Force and were not chosen to reflect the makeup of the larger Gulf War service member population.

DOD hospitalization study. The Naval Health Research Center in San Diego carried out a study of hospitalizations in military hospitals in the two years following the Gulf War.75 Hospitalization frequency for 547,076 active duty Gulf War veterans was compared to that for 618,335 military personnel who did not go to the Gulf region. In the two years prior to the Gulf War, Gulf-deployed personnel were at lower risk of hospitalization than those not deployed. After the war, overall rates of hospitalizations in the two populations were similar. When examined by specific discharge diagnosis, Gulf War veterans were at elevated risk of hospitalization for some diagnoses, including testicular cancer and genitourinary disorders in 1991, diseases of the blood and blood-forming organs in 1992, and mental disorders during 1992 and 1993. The investigators found the differences were not consistent over time and concluded their results do not suggest an emerging illness associated with Gulf War service.

This study's results cannot be generalized to the Gulf War population at large because it only includes hospitalizations from active duty troops while they remained in the service. Members of the Reserves or National Guard, or those who left service, are not represented. Additionally, the study cannot provide information about types of illnesses that do not result in hospitalization. This research does, however, indicate the absence of a large increase in illnesses requiring hospitalization among active duty Gulf War veterans.

Cognitive testing studies. Some small epidemiologic studies have been carried out in Gulf War veterans to assess complaints of cognitive difficulties such as memory problems. Comprehensive, structured neuropsychological testing is used clinically to evaluate subtle cognitive difficulties. Typical dimensions that are evaluated by these tests include: intellectual functioning (i.e., estimated premorbid IQ), attention, concentration, language, visuospatial processing, learning/memory, and motor skills.316

Results of cognitive testing of four populations of Gulf War veterans have been published or presented at national medical conferences.73,120,316,318 Although these four studies were small-groups of Gulf War veterans ranging in size from 19 to 149 people-several consistent findings emerge. On objective testing, memory and concentration performances were the same or only slightly decreased in groups of Gulf War veterans compared to control participants. Self-perceptions of memory dysfunction, however, were greater among the groups of Gulf War veterans. A small minority of Gulf War veterans who were significantly distressed due to PTSD or other psychiatric diseases did have objective memory and concentration impairment. These data are preliminary and require replication in additional studies.

Summary of morbidity studies. Completed morbidity studies show an increase in reporting of symptoms-such as fatigue, joint pain, memory problems, and headaches-in individuals who were deployed to the Gulf. The study results, however, do not indicate consistent abnormal laboratory or physical findings in these groups. Until results from some of the larger, population-based epidemiologic studies become available, conclusions cannot be generalized from these studies about the extent of illnesses in the Gulf War veteran population as a whole.

DATA ON STRESS-RELATED DISORDERS

Physicians have observed in many previous wars that physical and psychological stress can lead to the development of higher rates of psychiatric illnesses than are observed in the general population. PTSD and depression are particularly prevalent problems in combat veterans. Stress is also known to affect the endocrine, cardiovascular, immune, and central nervous systems-i.e., to cause serious biological problems that are in no way trivial. As expected from experiences in previous wars, some Gulf War veterans report physical and psychological symptoms that frequently can be manifestations of stress, including fatigue, headaches, loss of appetite, sleep problems, and cognitive difficulties (such as memory problems and difficulty in concentration).

Psychological Diagnoses in Clinical Programs

Psychological conditions are either the primary or secondary diagnosis in 36.0 percent of CCEP participants.277 The most common conditions are: major depressive disorder, neurotic depression (also called dysthymia), depression (not otherwise specified), PTSD, anxiety disorders, adjustment disorders, alcohol-related disorders, and substance-related disorders (table 3-10).

Among participants in VA's Registry, 15.1 percent of the top three diagnoses for each patient were psychological conditions, with the most common being depression (not otherwise specified), PTSD, and anxiety disorders (table 3-10). Additionally, 15,486 Gulf War veterans had been admitted for treatment to a VA hospital as of September 1995, and psychological conditions were the most common diagnosis for these inpatients (43 percent of total diagnoses). Specific psychological conditions included PTSD and adjustment disorders, alcohol dependence, and drug dependence.109

The psychological conditions diagnosed among Gulf War veterans also are common in the general population. The best estimates of the prevalence of psychiatric disorders in the general population are based on the National Comorbidity Survey (NCS), a comprehensive, highly structured, population-based survey of 8,098 adults, aged 15 to 54 years.**113,114 The diagnostic criteria used in this national study are basically the same as that used by DOD and VA.

The percentage of individuals who met diagnostic criteria for several disorders in the 12 months preceding NCS interviews was: major depressive disorder, 10.3 percent; dysthymia (neurotic depression), 2.5 percent; generalized anxiety disorder, 3.1 percent; alcohol-related disorders, 9.7 percent; and substance-related disorders, 3.6 percent.113 The lifetime prevalence of PTSD was 7.8 percent.114 Among the age groups encompassing 15 to 54 years, these serious psychiatric diseases peaked during 25 to 34 years; there was a significant decline in lifetime prevalence with increasing age.113

Treatment for Psychological Disorders

Stress-related illnesses are real, often debilitating illnesses for which treatment interventions are available. Treatment forstress-related disorders is, by necessity, case-specific and symptom-oriented. For instance, no one treatment regimen is appropriate for the overlapping range of problems for tension headaches, chronic fatigue, fibromyalgia (FM), depression, and anxiety disorders. Despite some variability in therapeutic approaches to depression, anxiety disorders, and other psychological conditions, there is a relatively narrow range of treatment options.

Testimony before the Committee and interviews during site visits clearly indicate a significant stigma remains associated with psychological diagnoses. This perception often interferes with veterans receiving or accepting adequate care. In many instances, individuals report meeting command resistance to granting the necessary time off to maintain an adequate treatment program. This is true of all chronic illnesses, but especially so for veterans with psychological diagnoses-despite the fact that since 1986, service members with certain chronic illnesses that require medical monitoring have been allowed to remain on active duty.

Frank PTSD is particularly difficult to treat because alcoholism or other comorbidities frequently are present. Nevertheless, there has been some agreement on the basic approaches to treating PTSD. The Director of VA's National Center for PTSD describes three phases of treatment: stabilization/establishing trust and safety; trauma-focused therapy (therapy targeting the traumatic event), i.e., what happened and how one deals with and makes sense of what happened; and moving from the past to present reintegration into society by disconnecting from the trauma and reconnecting with the present.62

In randomized clinical trials, cognitive-behavioral therapy (CBT) has been the most successful treatment for PTSD. CBT centers on two psychological theories-learning theory and how one appraises a situation and develops a more adequate coping response. There are a variety of CBTs, including exposure therapies that include systematic desensitization, imaginal and in vivo exposure, and anxiety management training therapies (e.g., stress inoculation training, biofeedback, cognitive therapy, and relapse prevention). Often, exposure therapy and anxiety management are combined. CBTs can be used either individually or within group settings.

Several additional types of psychotherapy ranging from peer counseling to marital counseling to long-term dynamic therapy exist. Some are designed to be short-term and problem focused, while others are long-term and ongoing. Other disorders should be treated concurrently, although substance abuse usually must be treated first.

Some pharmacological treatments with drugs developed for depression (e.g., Prozac and Zoloft) also can be successful for PTSD. A fruitful area for research and development will be drugs that act on the neurobiological systems most implicated in PTSD and other stress-related disorders-e.g., corticotropin releasing factor antagonists, N-methyl-D-aspartate antagonists, and neuropeptide antagonists.

Stress-related Symptoms Reported in the Clinical Programs

Headaches are a frequent symptom reported by Gulf War veterans who have received clinical evaluations through DOD and VA (39 percent of the top seven symptoms for CCEP and 18 percent of the top three symptoms in the Registry).109,277 Tension headaches are coded under the diagnostic category "Psychological Conditions" and were diagnosed in 11.3 percent of CCEP participants and 2.3 percent of Registry participants. Other types of headaches are coded under "Nervous System Diseases" (migraine headaches) and "SSIDC" (nonspecific headaches). Migraine headaches were the primary diagnosis in 2.7 percent of CCEP participants, and nonspecific headaches were the primary diagnosis in 2.7 percent of this group. The frequency of headaches among Gulf War veterans does not appear to be unusual, since headaches are one of the most common reasons for seeking medical care.124,137

Gulf War veterans commonly report cognitive difficulties. These symptoms can be associated with diseases such as major depression and PTSD. For the CCEP, 34 percent of participants report memory loss and 27 percent report difficulty concentrating.277 Fourteen percent of Registry participants report memory loss.109

To date, DOD reports that only a few CCEP participants have demonstrated cognitive deficits following neuropsychological testing. That is, such testing generally has eliminated an underlying neurologic etiology for the reported memory problems.277 Organic brain syndrome (OBS) is a generic medical term for brain damage due to several diseases, such as head trauma or Alzheimer's disease. OBS is the primary diagnosis in 0.6 percent of the CCEP participants. The extent of OBS in the VA Registry population has not been reported.

The symptoms (i.e., diagnostic criteria) of common psychological conditions overlap with some of the symptoms that frequently are reported by Gulf War veterans. As noted earlier in this section, such conditions include major depression, PTSD, and anxiety disorder. As an example of the extent of overlap, diagnostic criteria for major depression are provided in table 3-11.3Comparison to table 3-2, which lists some of the symptoms frequently reported by the first 18,075 CCEP participants, reveals that symptoms relevant to major depression are: Criteria 1-depression (23 percent of CCEP participants); Criteria 4-sleep disturbance (32 percent); Criteria 6-fatigue (47 percent); Criteria 8-memory loss (34 percent); and Criteria 8-difficulty concentrating (27 percent). Criteria 3-weight loss was a frequently reported symptom for 7 percent of the first 18,075 CCEP participants.

Epidemiologic Studies of Stress-related Conditions

The need to understand the effects of stress and related disorders in Gulf War veterans was recognized in 1991, and several epidemiologic studies focused on these issues were launched. Among the completed studies, the results primarily address the prevalence of stress-related conditions and the role of risk factors and protective factors.

Large epidemiologic investigations on effects of stress. Research targeted to increase knowledge about how stress could contribute to Gulf War veterans' illnesses involves several large-scale efforts. Studies with generalizable results are described briefly in the following sections. Other studies with similar, though not generalizable, results include the West Haven, CT, VA study of the 142nd medical unit and the 143rd military police unit of the Connecticut National Guard,239,240 the Little Rock, AR, VA study of U.S. Army and Air National Guard and Reserve personnel,300,329 and the Mountain Home, Johnson City, TN, VA study of the 24th Marines, Third Battalion, Company H.232-234 Only one significant study has included large numbers of active duty troops-the Walter Reed Army Institute of Research (WRAIR) study of units from Pennsylvania and Hawaii.245-247,272 Regrettably, the response rate in the WRAIR study was too low to extrapolate results to the overall Gulf War veteran population, but the study did find elevated rates of physical and mental distress among survey respondents who were deployed to the Gulf compared to those who were not.

Fort Devens, Massachusetts VAMC. Still in progress, the Fort Devens study involves 2,344 Gulf War veterans who have been followed by a research team at the Boston VAMC since 1991.334,337 Comparisons between the study sample and the overall Fort Devens population indicated that the study subjects were representative of the military population that was processed through this base during that time. The study population included 46 units with a wide range of military occupational specialties from several regions in the United States. Hence, its results are relevant to the health status of the larger population of Gulf War veterans who were in the U.S. Army Reserve and National Guard.

Men and women in the Fort Devens sample had equivalent levels of combat exposure.336 Fifty-six percent of both genders reported little or no direct combat exposure. However, if combat exposure was held constant, certain types of stressors appeared to affect people differently. For women, witnessing death and serious accidents significantly predicted poorer psychological adjustment. For men, marital strife or being placed on chemical/biological alert or SCUD alert were more strongly associated with the development of psychological symptoms. Sexual assaults and harassment were the most important factors that explained the different rates of PTSD symptoms between women and men.334 This group of veterans has been evaluated at three time points, starting with five days from their return.

Using the Mississippi Scale for Combat-Related PTSD, investigators found that approximately nine percent of women and four percent of men had PTSD-like symptoms that likely would qualify them for a positive diagnosis at Time 1 (five days after return).336 At Time 1 and Time 2 (18 to 20 months after return), women reported more PTSD symptoms than men.337 At Time 2, both men and women reported higher levels of PTSD symptoms than at Time 1 (11 percent for men and 21 percent for women). At Time 3 (approximately three years after return), however, rates of PTSD symptoms declined to approximate levels for Time 1, so there appears to be some recovery.334

To provide context for the rates of PTSD in Gulf War veterans, the best estimates of the rates of PTSD in the general U.S. population are based on the NCS, as described earlier.113,114 Based on NCS data, lifetime prevalence of PTSD in men was five percent, and it was most commonly associated with combat experience during a war or witnessing someone being badly injured or killed. The lifetime prevalence of PTSD in women was 10.4 percent, and it was most commonly associated with a history of rape or sexual molestation.

During the follow-up examination at Time 4 (begun late 1996), actual functional status will be examined, such as days of work lost and quality of life. Continued funding for the Fort Devens study is being provided as part of the Boston VA Environmental Research Center. DOD recently funded a parallel study assessing psychological status in a group of Gulf War-era military personnel who did not deploy.

New Orleans, Louisiana VAMC. In 1991, the New Orleans VAMC developed a psychological assessment program as part of a series of programs set up at VAMCs nationwide.300 The New Orleans research team evaluated 1,520 Reserve and National Guard troops who were mobilized for Gulf War duty. Because this larger study sample included Reserve and National Guard troops in the U.S. Army, Navy, Air Force, and Marines, study results are relevant to the health status of the larger population of Gulf War veterans who were members of the Reserve or National Guard.

The initial assessment took place within a few months of the end of the war.254,317 Compared to nondeployed troops, individuals from deployed units reported more physical symptoms and had more negative mood states, including depression, anger, and anxiety. The two groups differed in prevalence of reported headaches, general aches and pains, lack of energy, and sleep disturbance. Twenty-three percent of war-zone-deployed troops reported at least mild levels of clinical depression, while 14 percent reported clinically significant levels of PTSD. Individuals diagnosed with PTSD also displayed less proficient cognitive performances in neuropsychological functioning, pertaining mostly to attention and new learning.317 Deployed troops who reported higher levels of war-zone stress exposure were characterized by more depression, anxiety, anger, hostility, physical symptoms, and PTSD symptoms.

Women reported more physical symptoms than men, regardless of war-zone assignment.255 Ethnic minorities reported more depression than nonminorities, regardless of war-zone assignment. No gender differences existed for measures of PTSD or psychological distress among deployed troops. Minorities among deployed troops were at greater risk for developing symptoms of PTSD than nonminorities among deployed troops.

Highland Drive (Pittsburgh) Pennsylvania VAMC. Starting in July 1991, the PTSD team at the Highland Drive VAMC in Pittsburgh conducted mental health screening and outreach programs for about 620 Reserve personnel from deployed and nondeployed units in the U.S. Army, Navy and Marines in western Pennsylvania, eastern Ohio, and West Virginia.197,300 Because the study included individuals in the U.S. Army, Navy, and Marines, results can be viewed in context of the larger population of Reserve personnel who served in the Gulf War. In addition, a wide variety of stressors were encountered by the units evaluated, ranging from simple unit activation for groups that stayed in the United States to the deaths and injuries suffered by the 14th Quartermasters (QM) Unit when its barracks were destroyed by a SCUD missile.

The 439 reservists who were deployed to the Gulf region demonstrated significantly higher rates of psychological symptoms than individuals sent to Europe or who stayed in the United States.197 Gulf War veterans reported higher rates of PTSD, depression, and global psychological distress.

Focused, small-scale epidemiologic studies on stress. Researchers also have investigated stress responses in certain veterans who performed specific duties (e.g., graves registration) or experienced significant combat trauma (e.g., a SCUD missile attack). Because these research subjects experienced more extreme levels of stress than the average Gulf War veteran, data from these studies are not generalizable.

U.S. Army unit that experienced SCUD missile attack. The PTSD Clinical Team at the Highland Drive VAMC (Pittsburgh) developed an early treatment intervention program for the members of the 14th QM Detachment, whose barracks were destroyed by an Iraqi SCUD missile on February 25, 1991.198 When the SCUD missile struck their barracks, 28 soldiers were killed, and 99 were wounded. Blast effects on survivors included extensive shrapnel wounds and ruptured eardrums.

The PTSD Clinical Team initially contacted members of the 14th QM during the week of March 18, 1991, and treatment continued until April 24, 1991. Five of the 20 soldiers who were onsite at the time of the missile attack were judged to have met the criteria for PTSD during the initial assessments. Testing revealed these soldiers reported distress from nearly twice as many symptoms related to war stress as the four soldiers who had been on guard duty three miles away at the time of the attack. Nine of these 24 soldiers reported an increase in their alcohol consumption since their return home.

At the end of treatment, the 20 soldiers who had been onsite showed significant decreases of symptoms related to PTSD and depression. One patient was judged to continue to meet full criteria for the diagnosis of PTSD, while the other four patients who previously had met the PTSD criteria were still showing significant, though decreased, stress symptoms. Alcohol consumption was reported as decreased for most of those interviewed.

A follow-up study of the 14th QM in 1993 revealed about half of the treated patients were continuing to show improvement, and about half of the patients were returning to pretreatment levels of symptoms.204 The PTSD team at the Highland Drive VAMC continues to follow and treat the surviving members of the 14th QM.

U.S. Army units that performed graves registration duties, New Orleans VAMC. Among the sample of 1,520 military personnel studied by the New Orleans VA, were 194 members of QM units assigned graves registration duties that encompassed handling, identification, and processing of bodies and body parts. In one unit (24 people), investigators found the prevalence of PTSD was 46 percent. They also reported a high incidence of psychiatric diagnoses concurrent with PTSD, including depression (33 percent) and alcohol abuse/dependence (13 percent).252

A second study compared 40 service members who performed graves registration duties with 20 individuals from the same units who were not deployed to the Gulf War and who did not perform graves registration duties.253 Current diagnoses of PTSD were made in 48 percent of the deployed troops, compared to none for nondeployed service members. Diagnoses concurrent with PTSD included depressive disorder (18 percent) and alcohol dependence (10 percent). After one year, 42 percent of the service members who had performed graves registration continued to meet criteria for PTSD.317 The New Orleans research team continues its followup of both groups who performed graves registration duties.

U.S. Army units that performed graves registration duties, Walter Reed Army Institute of Research. Researchers at WRAIR's Department of Military Psychiatry also studied units with graves registration duties. They found consistent, but milder, symptoms compared to the New Orleans groups.152,153

Summary of epidemiologic data on stress. Epidemiologic studies to assess the effects of stress invariably have found higher rates of PTSD in Gulf War veterans than among individuals in nondeployed units or in the general U.S. population of the same age. It also appears groups with the most severe stress, such as the group injured by the missile attack, have a greater risk of PTSD than other Gulf War veterans.

In the large epidemiologic studies performed in Boston and New Orleans, the rates of PTSD and other psychological conditions had increased at the one-year follow-up evaluation, rather than ameliorating over time. Long-term followup to determine the effectiveness of treatment and outreach efforts is indicated in these study groups. The long-term effects of stressors of the Gulf War on active duty troops remain largely unexplored.

DATA ON UNDIAGNOSED ILLNESSES

A significant number of Gulf War veterans who have participated in the government's clinical programs report symptoms that do not fall into standard diagnostic categories. The epidemiologic studies also have identified many veterans who report symptoms of illness, but who do not show abnormalities on physical examinations or standard diagnostic tests. Congress has authorized VA to provide disability compensation to Gulf War veterans with undiagnosed illness (Public Law 103-446; 38 CFR 3.317), but the impetus to determine the underlying cause of the problem remains.

Data from Clinical Programs

More than 40 percent of CCEP participants have a primary or secondary diagnosis of SSIDC. This diagnostic category includes an extremely heterogeneous group of miscellaneous symptoms that do not fit elsewhere in the diagnostic coding system. As shown in table 3-12, the category encompasses generalized symptoms, such as malaise and fatigue; isolated abnormal laboratory results (i.e., a nonspecific reaction to the tuberculin test or an elevated sedimentation rate); and symptoms that prove to be transient (e.g., an episode of seizures or a rash, by history only).277 No significant anatomical, physiological, biochemical, or pathological abnormalities are detectable in individuals whose symptoms are coded in the SSIDC category. DOD has reported that the frequency of symptoms coded under SSIDC for CCEP participants is about five times higher than the frequency of coding of SSIDC in the general U.S. population aged 20 to 40 years.277

Of VA's Registry participants, 10,391 individuals (19.9 percent) reported some symptoms, but they did not have a characteristic set of signs and laboratory test abnormalities that allowed a medical diagnosis to be made.109 This group of Registry participants is comparable to the group of CCEP participants who were coded with a primary diagnosis of SSIDC, and their symptoms are similar. Table 3-13 presents the most common symptoms among these VA participants.

Symptom-based Diagnoses

The Committee noted the interest of many veterans in possible links between unexplained illnesses and recognized diagnoses-such as CFS and FM-that are based on symptoms reported by the patient rather than on physical abnormalities evident to a physician or on laboratory findings. Veterans also expressed a need to know more about Multiple Chemical Sensitivity (MCS), which is not currently a recognized diagnosis in U.S. medical practice.

Chronic Fatigue Syndrome. The CDC's 1994 consensus case definition for CFS requires both:

The 1994 CDC case definition lists many medical and psychiatric conditions that exclude the diagnosis of CFS. CFS is strictly a diagnosis of exclusion, and no confirmatory lab test exists.63

The prevalence of CFS in Gulf War veterans is unknown. VA diagnoses patients with CFS, but it has not reported the proportion of veterans in the Registry with this diagnosis. DOD has reported that 42 of the first 10,020 participants in the CCEP (0.42 percent) met the 1994 CDC case definition for CFS;278 its report on 18,075 participants did not provide this information.

Fibromyalgia. The 1990 American College of Rheumatology consensus case definition of FM requires both:

Other symptoms FM patients frequently report include sleep disturbance, fatigue, morning stiffness, anxiety, headache, and depression. No exclusions are made for the presence of concomitant X-ray or laboratory abnormalities. Therefore, a patient may be diagnosed with FM and another disorder simultaneously, such as rheumatoid arthritis, osteoarthritis, or major depression. There is no confirmatory lab test.

The prevalence of FM in Gulf War veterans is unknown. VA diagnoses patients with FM, but it has not reported the proportion of veterans in the Registry who have FM. DOD has reported that approximately 1.5 percent of CCEP participants have received a primary or secondary diagnosis of FM.277

Multiple Chemical Sensitivity. There is no consensus case definition for MCS, although two recent government-sponsored conferences have attempted to develop one. MCS patients report many symptoms, including tiredness or lethargy, fatigue, memory difficulties, difficulties concentrating, dizziness or lightheadedness, and depressed feelings when exposed to low levels of common, everyday substances. Symptoms relevant to many different organ systems have been linked to MCS in the clinical ecology literature; symptoms related to the central nervous system are the most common.

The majority of patients diagnosed with MCS have no objective abnormalities on physical examination or on routine laboratory testing. The physicians who use this diagnosis use a variety of nontraditional diagnostic and treatment techniques, none of which have been validated in a controlled trial.

One physician who specializes in MCS has consulted on 75 patients at the Houston VA Persian Gulf Referral Center. She reported that among her first 59 consultations, 46 patients (78 percent) reported a variety of symptoms, referred to as intolerances, when exposed to various chemical inhalants such as traffic exhaust, perfume, or tobacco smoke.160 No other data on chemical intolerances in Gulf War veterans exist.

Overlap of symptom-based diagnoses. Several studies have demonstrated that symptoms of CFS, FM, and MCS overlap. A 1994 Seattle study evaluated three groups with 30 patients each who had been diagnosed with CFS, FM, or MCS. Researchers for this study concluded that symptoms typical of each disorder were prevalent in the other two conditions and that, "with the exception of tender points, other physical examination findings appear to be prominent by their absence in CFS, FM, and MCS."19 The symptoms of CFS, FM, and MCS also are common among CCEP and Registry participants with undiagnosed illness. The government is sponsoring research on each of these conditions.

Deficits in peripheral nerve function. At least one researcher has suggested a link between unexplained illnesses among Gulf War veterans and measurable deficits in peripheral nerve function attributable to one or more Gulf exposures.99 The publication, however, reports on a small population that was not randomly selected, so data are not generalizable to the entire Gulf War service population. Reported results also revealed no objective differences between Gulf War veterans and civilian participants, and there was no evidence for a clinically demonstrable peripheral neuropathy in any of the Gulf War veterans who participated in the study.

ILLNESS AMONG FAMILY MEMBERS

Some veterans and their family members, scientists, and physicians have voiced concern that Gulf War veterans' illnesses could or do affect the health of their families. The potential for adverse reproductive outcomes-infertility and birth defects-and new hypotheses regarding communicable diseases have generated the most anxiety.

Data from the CCEP

Since CCEP's inception in 1994, spouses and children of active duty personnel have been eligible for enrollment. DOD's April 1996 report discussed its evaluation of 332 spouses and 191 children.277 VA began a similar program for the spouses and children of veterans in April 1996. As of November 1996, about 1,500 individuals had enrolled in the VA program, but the clinical results had not been reported.

Table 3-14 reports the frequency in CCEP of the primary and all diagnoses in the 332 spouses.277Overall, the distribution of diagnoses in spouses is similar to the distribution of the diagnoses in enlisted individuals. The most prevalent major diagnostic categories are psychological conditions, MSDs, and SSIDC. The genitourinary system is one organ system that has substantially higher rates of diseases in spouses, who largely are women-not surprising as this is a finding also typical in women in the general U.S. population aged 20 to 40 years, when compared to men the same age.222

Primary diagnoses for the 191 children in DOD's CCEP are shown in table 3-15.277 Seventy-two children were healthy. Thirty-five children were born with various congenital anomalies that were not concentrated in a single organ system. Seventeen children had skin problems of the types that are common in the general U.S. pediatric population. The remaining 67 children were diagnosed with a range of diseases in several organ systems. As with the adult CCEP population, results from analyzing data collected from this population-a self-reported case series-cannot be generalized to the entire Gulf War population.

Adverse Reproductive Outcomes

In the years after the Gulf War, media reports based primarily on anecdotal evidence asserted increased rates of birth defects in children born to Gulf War veterans. Reports of high levels of infertility and pregnancy loss also appeared in the national press. Among the difficulties in assessing whether Gulf War veterans were experiencing these problems because of their service, is the fact that reproductive failures are common occurrences in the United States.

Physicians diagnose major birth defects in three to four percent of infants in the first year of life. A birth defect is a structural abnormality present at birth, including malformations, which involve poor tissue formation; deformations, which involve abnormal changes in developing tissue; and disruptions, which involve the breakdown of normal tissue. Major birth defects are those that affect survival, require substantial medical care, or result in marked physiological or mental impairment.

Birth defects are the leading cause of infant mortality in the United States, accounting for more than 21 percent of all infant deaths. Of the approximately 100,000 to 150,000 infants born with a major birth defect each year, approximately 6,000 die during the first 28 days of life, and an additional 2,000 babies die before reaching their first birthday. Annually, then, 92,000 to 142,000 children affected to various degrees by birth defects live beyond the first year.16,179

More than one in eight couples in the United States is classified as infertile, defined as the inability to conceive after trying for 12 months. In addition, approximately 20 percent of pregnancies end in spontaneous abortion (miscarriage) between the 4th and 28th week of pregnancy. However, estimates of total pregnancy loss that include figures derived from studies prior to the fourth week of pregnancy conclude that as many as 75 percent of all conceptions are lost-often without the woman ever knowing she was pregnant.28

The causes of birth defects, in general, can be determined in just over half of all cases.143 In the vast majority of cases, birth defects occur in families where there is no history of the disorder. The reasons for infertility and pregnancy loss also are often difficult to impossible to elucidate.

To evaluate potential associations between Gulf War service and adverse reproductive outcomes, the Committee undertook two primary tasks: an appraisal of the biological plausibility of such an association and an assessment of government studies in this area. (In chapter 2, the Committee evaluated government services that are relevant to addressing reproductive health-related clinical needs and concerns of veterans. We analyzed reproductive health risks of specific Gulf War exposures in chapter 4 as part of our analyses of health effects on all organ systems.)

Biological plausibility. Many things can go wrong in a pregnancy. In fact, many scientists posit that it is miraculous that most often children are born healthy. Determining the causes of adverse outcomes is a complex process. To establish an environmental exposure as the cause of a birth defect, there must be a valid, even if hypothetical, explanation as to how a particular agent or agents could have acted biologically to produce a particular effect.

Teratogenicity. Teratogens are environmental agents that adversely affect the fetus in the uterus. Exposure to teratogens accounts for three percent of all birth defects. Over the past 30 years, a significant amount of data have been collected on the reproductive risks of exposing pregnant women to infectious agents, drugs, chemicals, and physical environmental agents (i.e., ionizing radiation or heat). The use during the 1960s of thalidomide by pregnant women and the resulting limb reductions in their children serves as a prime example.

It is likely that some teratogens were present in the KTO. In part to protect fetuses from potential exposure to teratogens, pregnancy was cause for either nondeployment or evacuation from the theater.309 There are no known reports of birth defects in children born to women who were pregnant with them while they were in-theater.

Teratogens to which males are exposed throughout conception and pregnancy, such as those reported in certain occupational settings, potentially could affect the developing fetus. An embryo or fetus could be exposed either by transfer of the teratogen via the semen during intercourse or by exposure of the pregnant woman to the toxic agent via the clothing of the male.194 However, exposures unique to the Gulf War environment are highly unlikely to have affected children via this male-mediated mechanism.

Mutagenicity. Toxic agents that cause mutations are called mutagens. Mutagens act on either somatic cells (the cells of the body) or germ cells (eggs and sperm), or both. Their actions on germ cells are distinct from the effects of teratogens because they alter the genetic constitution of the germ cells prior to conception. Ovarian exposure to some mutagens can disrupt the ovarian cycle and damage the oocyte, resulting in potential infertility, fetal loss, or birth defects. Because females receive a fixed amount of oocytes before birth, chronic or long-term exposures to ovarian mutagens can have a cumulative and permanent effect on the health and vitality of ova.

In contrast, the effects of mutagens on male reproductive biology most often are transient, although a few agents are known to be stem cell mutagens or remain for long periods in the body. Males manufacture millions of sperm daily in a cyclical, constantly renewing process of cell division. The adverse effects of mutagens are likely to dissipate in 90 days-the time required for complete turnover of sperm-after the exposure ends. If conception were attempted within that 90-day period and the exposure had a mutagenic effect, infertility or pregnancy loss would be the most likely outcome. Either the sperm would be too damaged by the mutagen to fertilize the egg, or the sperm would contribute to the creation of an embryo that carried too many mutations to survive beyond a few days. In a few documented exposures (such as cancer patients exposed to high doses of radiation and chemotherapy), sterility can be permanent because of damage to the stem cell pool.225 It is possible that some types of nonlethal mutations to stem cells may result in birth defects.56

Some veterans and advocates have hypothesized a connection between exposure to mutagenic agents in the Gulf and development of Goldenhar syndrome in veterans' offspring. Decades of research have revealed that mutagenic agents are not specific.16,78 Current evidence is such that a mutagen would be expected to cause a random increase in the incidence of genetic disease, not the increase of particular genetic syndromes to the exclusion of others.

Data available concerning the types and levels of exposures that occurred during the Gulf War do not indicate the presence of potent mutagens, particularly agents that would affect stem cells.17,134 It is known that mustard agent, as a carcinogen, has mutagenic properties (i.e., it affects the somatic cells of the exposed individual), but it is unlikely the effects of low-level exposure would manifest as birth defects. Infertility would be the more likely reproductive outcome, cancer in the individual exposed the most likely long-term outcome.

The mechanisms of male reproductive biology make it unlikely that acute exposures to environmental agents present in Southwest Asia would result in adverse reproductive outcomes. Under the circumstances surrounding possible exposures to males in the KTO, an increase in a single type of birth defect is biologically implausible. If there were demonstrable adverse effects related to exposure, infertility or reduced fecundity are the most expected outcomes.

Studies of reproductive health of Gulf War veterans. In addition to establishing biological plausibility, epidemiologic studies are necessary to determine whether a connection exists between exposure experiences while in the Gulf and subsequent adverse reproductive outcomes. If such an association exists, two outcomes of epidemiologic studies would be expected.

First, a higher overall prevalence of adverse outcomes-including birth defects, infertility, pregnancy loss, stillbirth and prematurity-than would be expected in the normal population should be demonstrated. If approximately 200,000 children are born to Gulf War veterans, approximately 6,000 to 8,000 congenital malformations would be expected based on general population risks.

Second, if an association with a teratogen (not a mutagen) exists, one might see an unusual cluster of a specific type of birth defect appearing across multiple independent studies. Should either of these results occur, investigators could then work backward to evaluate whether there is some commonality among those who are affected-e.g., living in the same geographic location, similar ethnicity, or common exposures of the fetus to drugs, illness, or environmental agents.

Even with sophisticated epidemiologic studies, however, caution must be exercised. Any epidemiologic study dealing with all birth defects invariably will find some birth defect occurring at an increased incidence-i.e., a cluster-in the study population.

Researchers can assess and monitor approximately 60 major birth defect outcomes. In a properly designed protocol that encompasses an appropriately large sample size, one would expect that three birth defects would appear to be substantially increased, just by chance alone. In other words, any single epidemiologic study of birth defects in children of any cohort is likely to reveal-based on chance alone-a cluster of defects unrelated to exposure. In fact, it has been argued that if a properly designed epidemiologic study of birth defects does not result in a statistically significant cluster, this finding in and of itself would be particularly noteworthy.

Thus, before a cluster can be validly linked to an exposure, additional investigations of similar, but distinct, populations must be conducted to determine if a similar cluster is again observed. Only by surveying the total relevant population (which is usually prohibitive for practical and economic reasons) could one be absolutely certain of determining the true prevalence rate. Moreover, as noted earlier, biological plausibility also must be assessed.

Well-designed, scientifically valid epidemiologic studies comparing events among a random sample of Gulf War veterans to an appropriate group are required to determine whether an association exists between Gulf War service and the risk of adverse reproductive outcomes. The government's initial attempts to study the prevalence of birth defects in the children of Gulf War veterans showed mixed, nongeneralizable results. For example, slightly elevated rates of birth defects were found in a self-reported population surveyed by VA, but no increased prevalence of birth defects was found in a study of the children of National Guard members from Mississippi.195 To date, the government has sponsored three studies that should yield some generalizable, though limited, results.

Record-based evaluation of the risk of birth defects and military service in the Gulf War. In 1994, DOD funded a record-based evaluation of the risk of birth defects and military service in the Gulf War (known as Study 3). Although the study has a number of design deficiencies, it still is notable that researchers found that risks of birth defects-whether broadly or narrowly categorized-were not different among the deployed and nondeployed groups (using an adjusted odds ratio).37

Survey of reproductive outcomes. Another DOD-sponsored project (known as Study 4) aims to determine whether Gulf War veterans are experiencing partial or total infertility or other adverse reproductive outcomes at rates greater than would be expected. Experiences of a total of 16,000 couples will be compared for different reproductive experiences: married couples in which the woman served in the Gulf during the war; married couples in which the man served in the Gulf during the war; married couples in which the woman served in the military during the war but not in the Gulf; and married couples in which the man served in the military during the war but not in the Gulf. The survey, which is still underway, has been plagued by a low response rate.

Prevalence of congenital anomalies among children born to Gulf War veterans. A third study (often referred to as Study 7) has been designed to determine whether the prevalence and types of major congenital anomalies among children born between January 1, 1989 and December 1993 differ among Gulf War veterans, nondeployed military personnel, and civilians; among active duty and separated veterans; and among pre- and post-deployment conceptions. It will describe any patterns of major birth defects revealed and compare rates of fetal death, low birth weight, and pre-term deliveries. This study has considerable design advantages over previous studies assessing the reproductive health of Gulf War veterans.

A substudy of this investigation involves an evaluation of the prevalence of oculoauricularvertebral spectrum (Goldenhar syndrome) among children born or hospitalized in DOD medical treatment facilities. To date, five cases of Goldenhar have been identified among offspring of 34,067 Gulf War veterans, and three cases were identified among offspring of 41,220 nondeployed veterans. Although the rate per 100,000 of the Gulf War veteran group is seemingly twice that of the nondeployed group, the difference is not statistically significant.189

Absence of baseline data. One problem that plagues all studies of reproductive outcomes among Gulf War veterans is the absence of baseline data on military populations. The birth defects surveillance programs operated and coordinated by CDC have helped to diminish this problem in the civilian sector. The primary purpose of the National Survey of Family Growth, performed by the National Center for Health Statistics (NCHS), is to collect national data on factors affecting pregnancy and birth rates in the United States. NCHS has conducted the survey five times since 1973 and uses widely accepted sampling and survey methods to estimate rates of infertility, pregnancy loss, and birth defects. The samples are representative of the civilian noninstitutionalized population of women aged 15 to 44. Selected, proximate risk factors that might affect infertility and pregnancy loss also are collected.28

In addition to being used to search for increases in the incidence of specific malformations, surveillance systems can be used to develop baseline data, provide timely rates, identify geographic areas of concern for cluster investigations, and provide the basis for ecological investigations and follow-up studies to identify causes or risk factors such as drugs, nutritional factors, environmental exposures, maternal illnesses, and genetic factors.143 No such baseline data collection system specific to the reproductive health of military personnel currently exists.

VA program for spouses and children. In April 1996, VA established the Examination Program for Spouses and Children of Persian Gulf Veterans to fulfill a congressional mandate (Public Law 103-446). This program could eventually yield some data about reproductive outcomes among the families of Gulf War veterans. Under this authority, VA may provide examinations to any individual who: is the spouse or child of a veteran who is listed in VA's Registry and is suffering from illnesses or disorders; is suffering from, or could have suffered from, an illness or disorder (including a birth defect, miscarriage, or stillbirth) that cannot be disassociated from the veteran's service in the Southwest Asia theater of operations; or has granted VA permission to include in the Registry relevant medical data from the evaluation.

The program initially was funded at $2 million and was to be open to the first 4,500 individuals who called VA's Helpline. It has been extended through 1998. Examinations are provided by university-affiliated physicians at 32 VAMC coordinating centers, and participants are examined at one of these contractor sites. The program does not pay for travel or reimburse for incurred expenses. Examinations of spouses are similar to VA's Phase I Registry examination, including a standardized history and physical examination. The protocol for the children of veterans involves a detailed medical history, including symptoms and a developmental history. VA is not authorized to provide medical followup or treatment of conditions diagnosed by the medical examination. The program currently is not designed to provide useful research results, but could be used to identify areas needing further evaluation.

Infectious Diseases

Infectious diseases are a special concern to Gulf War veterans and their family members. Based on its CCEP, DOD reports no clinical evidence that Gulf War veterans have transmitted an infectious disease endemic to the Gulf to their spouses or children. Among 332 spouses of veterans who have been evaluated, 23 individuals (7.0 percent) have a primary or secondary diagnosis of an infectious disease (table 3-14).115 These include 14 cases of fungal skin infections, six cases of vaginal yeast infections, two cases of warts, and one case of tuberculosis-all of which are common infectious diseases in the general U.S. population.

Among 191 children of Gulf War veterans who have been evaluated in the CCEP, 17 children (8.9 percent) have a primary diagnosis of an infectious disease (table 3-15).277 Nine children have an upper respiratory infection, six children have otitis media (ear infection), one child has tinea capitis (fungal skin infection), and one child has chronic pneumonia. All of these diagnoses are common childhood infectious diseases in the general U.S. population, with the exception of chronic pneumonia.

Three other microorganisms have been hypothesized as possible etiologies for illnesses in some Gulf War veterans and their families: Mycoplasma infections, microsporidia infections, and occult, systemic streptococcal infections. Each hypothesis awaits systematic, controlled research to confirm it as a potential cause of morbidity in Gulf War veterans.

Based on available evidence, the Committee believes it is unlikely these microorganisms are responsible for widespread disease among Gulf War veterans or their families.

SUMMARY

In the absence of generalizable, quantitative information about the extent of Gulf War veterans' illnesses, only a qualitative range of symptoms and illnesses being reported by Gulf War veterans can be described. This general picture derives from information about participants in DOD's and VA's clinical evaluation program and preliminary data from several epidemiologic studies.

The Committee believes the most significant findings about the nature and extent of Gulf War veterans' illnesses-and recommendations for followup-must await the conclusion of the population-based epidemiologic studies. These results will come in well after the Committee disbands. Still, extensive information about many aspects of Gulf War illnesses exists. Based on in-house expert consultations, literature reviews, briefings, and testimony, the Committee makes the following findings and recommendations.

FINDINGS

RECOMMENDATIONS


*The cause of one death was unknown. Since this paper was published, some of these casualties have been reclassified. The current official count of deaths from hostile action in Operation Desert Storm is 148 person.

**Participants were not selected from a treatment-seeking population, but were asked whether they had recently sought medical care.