Methods Clinical Analysis
The most important contribution of the registries is that they provide objective clinical data about the health problems experienced by Gulf War veterans. Data from comprehensive clinical examinations can be used to evaluate:
For this study, military personnel and veterans were eligible if they had been deployed to the Gulf theater of operations between August 1990 and July 1991, and had completed a VA or DoD clinical registry examination by September 30, 1999.
A master file was obtained from the Defense Manpower Data Center, Monterey, CA, which contained a roster by social security number of 696,470 U.S. veterans who served in the Gulf during Operations Desert Shield and Desert Storm. This file was used to select eligible "conflict veterans" in the three registries. Also, this file was used to distinguish troops who served with active duty units and veterans who served in the Reserves or National Guard.
For VA, the first objective of this project was to combine the data from VA and DoD Gulf War registries in a manner such that the identity of the veterans who had participated in either clinical examination program would remain anonymous to both VA and DoD investigators. It was also necessary to develop a system in which additional databases -- such as military and demographic files, hospitalization files, vaccination files, and Gulf War exposure files -- could be linked to the clinical examination data. VA generated a randomized personal identification number (VA PIN) for each Social Security Number (SSN) of the 696,470 veterans who were deployed to the Gulf theater between August 1, 1990, and July 31, 1991. This study limited its analyses to those veterans who served during the first year of the Gulf deployment because of the potential for exposure to various wartime health hazards.
This roster of Gulf War veterans was matched by SSN to the VA and DoD registry databases to identify persons initially eligible for inclusion in this combined registry analysis project and to substitute the VA identification number for each individuals SSN. Participants of this combined registry analysis were further restricted to those veterans who had completed their registry examination by September 30, 1999.
Additional databases were matched to the roster of Gulf War veterans, SSNs were replaced with the corresponding VA PIN, and all other personal identifiers were deleted from each newly created database. The two registry databases, as well as the supplemental data files with only the VA PIN for identification, were distributed to the investigators of this combined registry project. There were 32,876 DoD registrants and 70,385 VA registrants who met all the eligibility requirements as described. This resulted in 100,339 unique Gulf War veterans for evaluation, including 2,922 veterans who were examined in both the VA and DoD registries.
Structural differences in the format of the CCEP and the initial and revised formats of the VA Gulf War clinical examination programs prevented the actual merging of the medical data contained in the various registry databases. The merging of the two VA registry formats and the CCEP with regard to symptom and diagnostic codes is technically possible. However, correct interpretation of the analyses of the merged data would be complicated by the variation in the structures of the three databases. In view of the lack of comparability of the structures of each database, data from the three registry programs have not been physically merged. For this report, data are reported separately by source (initial VA registry, revised VA registry, and CCEP).
In summary, after being reviewed for consistency, the data from the VA registries and CCEP, and related clinical and military databases, were linked by an anonymous VA PIN. VA and DoD then proceeded with a systematic evaluation of the registry data. Analysis of all data was therefore done anonymously.
Results of Clinical Analysis
There was a total of 100,339 individual Gulf War veterans eligible for evaluation in this study. Among the three clinical registries, there were 49,079 veterans evaluated in the initial VA clinical evaluation program, 21,306 evaluated in the revised VA registry, and 32,876 military personnel evaluated in DoDs CCEP (Figure 1). In this study population, 2,922 subjects had been evaluated in both the VA and DoD registries.
Figure 2 provides the number of veterans enrolling in the registries by calendar year. Most registry participants had enrolled for a clinical examination by 1996, within five years of the end of active hostilities in the Gulf War. There was an upsurge in registry enrollment in 1997 after over 100,000 Gulf War veterans who may have been exposed to chemical warfare agents shortly after the war were notified and encouraged to undergo a clinical examination in the VA or DoD clinical evaluation programs.
For the analysis of registry data, it is important to compare the demographic characteristics of registry participants with the overall population of 696,470 deployed U.S. troops (Table 2). As of 1991, the average age of registry participants was about two years older (29.8 years) than the total population of Gulf War veterans (28.0 years). A slightly higher percentage of registry participants were women than in the total population of Gulf War veterans (10% vs 7%, respectively). Similar percentages of racial/ethnic groups enrolled in the registries as were deployed to the Arabian Gulf. The level of education among registry participants was comparable to other Gulf War veterans. Reflecting their older age, registry participants were more often married (Table 2).
For military characteristics, there were a number of important differences among the registries and the overall population of deployed U.S. Troops (Table 3). Because eligibility for DoD health care is generally restricted to active duty troops, a lower percentage (9%) of CCEP participants were Reserve or National Guard personnel compared to the initial VA registry (43%) and the revised VA registry (24%). Overall, active duty troops were less likely to undergo a registry evaluation than Reserve and National Guard members, who represented ~ 16% of the deployed force. This finding may be due to the fact that Reserve and National Guard personnel are not routinely eligible for either VA or DoD health care when they are not currently on active duty. The VA registry program therefore offered ready access to high quality health care, which many veterans availed themselves of after the war.
Among the military service branches, Army personnel were most likely to participate in the registries. A slightly lower percentage of deployed officers enrolled in the registries (Table 3).
Among veterans evaluated during the first few years after the war in the initial VA registry, 29% reported that their health was poor or very poor (Table 4). A lower percentage (26%) of veterans evaluated after 1994 in the revised VA reported their health to be poor or very poor. Likewise, a lower percentage of CCEP participants (6%) reported their health to be poor. These data suggest that Gulf War veterans self-perception of overall health did not decline over time.
Although registry participants did not indicate a decline in general health over time, they perceived themselves to be less healthy than the general population of Gulf War veterans. Fewer than 6% of VA and DoD registry participants reported their health to be very good or excellent (Table 4). In contrast, in the VA National Survey of Gulf War veterans, which was a random sample of the overall population of U.S. Veterans, 16% reported that their health was excellent and 28% reported that it was very good.121 In an investigation of Gulf War veterans from Iowa, 21% reported that their health was excellent.170 Another study of British veterans produced similar data,167 all of which indicates that clinical registry participants are more ill than the overall population of Gulf War veterans.
A wide variety of symptoms were reported by Gulf War veterans (Table 5). Among over 100,000 registry participants, the most common symptoms were muscle and joint pain, fatigue, headache, memory problems, and sleep disturbances. Higher percentages of veterans reported these symptoms in the revised VA registry and the CCEP than in the initial VA registry because a greater number of symptoms could be coded in the latter two registries. These symptoms can be severe and debilitating but are not specific for any particular medical or psychological illness. They are the type of symptoms commonly reported by adults in outpatient clinic populations.
Frequency Distribution of Major Diagnoses
Table 6 provides the frequency distribution for broad ICD-9 diagnostic categories. The most common diagnoses were:
Tables 7 through 20 lists the 20 most frequent, specific diagnoses within broad ICD-9 classifications among the three registries. Data were presented separately for the three clinical programs because there were consequential differences among the registries in terms of the populations evaluated and how the data were collected and coded. The number of veterans with each diagnosis is provided. For tables 7 through 20, the percentages listed relate to the particular registry and the major ICD-9 diagnostic code; these are not percentage estimates for a particular diagnosis among all 100,339 registry participants.
The original VA registry provides data from 1992 to 1994, fairly soon after the war, and before the more complex "revised" VA registry and the CCEP were instituted. Additionally, the VA registries more often evaluated veterans who had left active duty or were members of the Reserves and National Guard; whereas, the DoD registry predominately enrolled active duty military personnel. Lastly, the frequency distribution of diagnoses was affected by the fact that the three registries coded different numbers of diagnoses: three in the initial VA registry, 10 in the revised VA registry, and seven in DoDs CCEP.
Despite these differences, the six most frequent diagnoses among broad ICD-9 classifications were similar among veterans evaluated early in the initial VA registry and veterans evaluated later in the revised registry (Table 6). Also, the most common diagnoses among active duty troops in the CCEP were similar to the VA registries, which evaluated a higher percentage of Reservists and National Guard personnel.
For specific diagnoses within broad ICD-9 categories (Tables 7-20), a wide range of common health problems were diagnosed, as noted in the following discussion. When reviewing the data from these registries, the specific percentages of individual diagnoses within categories cannot be compared, but an assessment of the rank order of these diagnoses provides important information. Again, although there are substantial differences in the databases, the most frequently identified diagnoses were similar in all three registries.
Infectious and Parasitic Diseases (ICD-9 001-139)
Like civilian populations of adults evaluated in outpatient clinics, the most common infectious diseases among Gulf War veterans were: athletes foot, common skin discoloration, and superficial infections of the groin and nails (Table 7). There was no cluster of unusual infectious diseases among the 100,339 evaluated Gulf War veterans, and infectious diseases associated with immunosuppression were infrequently diagnosed.
The most widespread infectious disease problems during Operations Desert Shield and Storm were those associated with crowding (acute upper respiratory infections) and reduced levels of sanitation (travelers-type diarrhea). However, respiratory and gastrointestinal infections were not prominent diagnoses after the war among registry participants. Other than the 12 known U.S. Gulf War veterans with visceral Leishmaniasis, no new case of this unique infectious disease was identified among over 100,000 VA and DoD registry participants.
Since the Gulf War, a pattern of infectious diseases has not emerged among patients with chronic unexplained symptoms, who have been hypothesized to have a unique "Gulf War syndrome." Nor has a characteristic physical sign or laboratory test abnormality been observed that would indicate a chronic infectious process, including a unique skin rash, lymphadenopathy, hepatosplenomegaly, transaminase elevations, or hematological abnormalities.171
Malignant Neoplasms (ICD-9 140-208)
Malignant neoplasms were a rare diagnosis among registry participants. Of the 16 major diagnostic categories, malignant neoplasms ranked last (Table 6). Moreover, diverse kinds of neoplasms were found in this population and not a particular type of tumor (Table 8). The most frequent malignancies observed among registry participants were skin cancers. This observation is consistent with the general civilian population. In the U.S., skin cancers are the leading cause of cancer among adults.172
There were 42 cases of testicular cancer, which is one of the most common cancers among young men. Because testicular cancer occurs most frequently in men 20 to 40 years of age, young military populations are at higher risk.173 Knoke et al., examined Defense Department postwar hospitalization records for testicular cancer and found no association with Gulf War deployment.21
Endocrine, Nutritional, and Metabolic Diseases, and Immunity Disorders (ICD-9 240-279)
In the three clinical programs, endocrine and nutritional disorders ranked 9th and 10th in frequency among broad ICD-9 categories (Table 6). Neither morbidity or mortality studies of Gulf War veterans has to date shown an increased risk of these health problems.123,129 No abnormalities were found among symptomatic Gulf War veterans with undefined illnesses in a study of in vitro immunological responses.174
The most common diagnoses within this ICD-9 classification were overweight and laboratory findings of elevated cholesterol (Table 9). As in the general population, diabetes and hypothyroidism also were relatively common diagnoses. A recent study has shown that 20% of Americans are severely overweight, and 6% to 7% of the U.S. population has diabetes.175,176 Being overweight, lack of exercise, a poor diet, and cigarette smoking have been associated with an increased risk of diabetes.177
Diseases of the Blood and Blood-Forming Organs (ICD-9 280-289)
There were relatively few diagnoses in this disease category (Table 6). Anemia and low white blood cell counts (agranulocytosis or neutropenia) were the most common observation on laboratory analysis (Table 10). These are frequent findings in routine laboratory testing of clinic patients in the general population.
The most common cause of significant neutropenia is iatrogenic, the result of drug therapy, particularly antibiotics and immunosuppressive drugs used to treat malignancies and autoimmune diseases.178 In a study of post-war hospitalizations, anemia among Gulf War veteran women was associated with pregnancy, which is often accompanied by anemia.123
Mental Disorders (ICD-9 290-319)
Mental disorders were common diagnoses in all three registries, ranking 2nd and 3rd most prevalent (Table 6). The most frequently diagnosed specific conditions were the same for both VA and DoD registries: tension headache, prolonged post-traumatic stress disorder, anxiety, and depression (Table 11). A wide variety of other mental disorders also were identified, which is expected in general outpatient populations.
High rates of mental disorders should be anticipated among adult outpatients because this is one of the most common clinical problems in the general population. Based on U.S. Population estimates, about 15% of the population 18 years of age and over fulfill criteria for at least one alcohol, drug abuse, or other mental disorder during a one month period.179 A higher prevalence of most mental disorders has been found among younger adults, except for severe cognitive impairments. Men have been found to have higher levels of substance abuse and antisocial personality, whereas women have had higher rates of affective, anxiety, and somatization disorders.179
Gulf War service was associated with acute mental reactions to stress after the war among U.S. Veterans in a controlled study that utilized hospital records.130 Personnel who served in ground war support occupations (men and women) were at greater risk for postwar drug-related disorders. Men who served in ground war combat occupations were at higher risk for alcohol related disorders. In a clinical case-control study of British Gulf War veterans, the findings suggested that ill Gulf War veterans do not have greater levels of psychiatric disorders typically seen in general psychiatry clinic but do have elevated rates of somatization.180
Diseases of the Nervous System and Sense Organs (ICD-9 320-389)
The most common diagnoses in this category were similar among the three registries (Table 12). They included: migraine headaches, non-specific hearing loss, buzzing in the ears, and carpal tunnel syndrome. Minor visual problems also were commonly found among registry participants. Exposure to pesticides and chemical warfare agents during the Gulf War have been hypothesized to have caused both central and peripheral neurologic damage among Gulf War veterans.181,182 However, diagnoses of these neurological disorders were relatively uncommon among over 100,000 registry participants: This diagnostic category ranked 5th and 6th most frequent among broad ICD-9 classifications in the three clinical programs (Table 6).
Intensive study of a small group of Gulf War veterans have found indications of possible neurological pathology.80,183-187 Other studies of Gulf War veterans have found no evidence of a characteristic disease of either the peripheral or central nervous system.188-195 Currently, the study of diseases of the nervous system among Gulf War veterans is one of the most important areas of Federally- funded research.31
Diseases of the Circulatory System (ICD-9 390-459)
In the three clinical programs, this diagnostic category ranked 8th to 11th among broad ICD-9 classifications (Table 6). The two most common diagnoses within this ICD-9 category were high blood pressure and hemorrhoids (Table 13). Greater than 20% of the adult population may have hypertension, which makes it a highly prevalent disorder in outpatient populations.196 There was a wide variety of other common heart and vessel disorders among registry participants.
Diseases of the Respiratory System (ICD-9 460-519)
A large number of respiratory illnesses were found among registry participants, which were not directly related to infectious diseases (Table 6). The most common diagnoses were allergic rhinitis (hay fever), asthma, and inflammation of the nose and sinuses (Table 14). These conditions are also prevalent in the general population of adults. Asthma has been estimated to occur in 4% to 5% of the U.S. population.197 The small number of diagnoses of bronchitis among DoDs CCEP participants may reflect the relative health of young active duty troops.
The initial VA registry was established because of concerns about the health effects of smoke exposure from burning oil well fires. These fires caused a major ecological disaster, which was characterized by the burning of more than 4 million barrels of oil per day and the production of 3400 metric tons of soot.86 During the war, deployment to Kuwait near the oil well fires was associated with an increased incidence of eye and upper respiratory tract irritation, shortness of breath, cough, rashes, and fatigue.86 However, monitoring of air pollution levels after the war did not indicate that troops were heavily exposed to chemicals from smoke or oil spills.84,85 A large epidemiological study of postwar hospitalizations of Gulf War veterans diagnosed with conditions thought most likely to be sequelae from smoke exposure failed to demonstrate a strong statistical association with serious illness.198
Diseases of the Digestive System (ICD-9 520-579)
Irritable bowel syndrome was the most frequent diagnosis observed in this category (Table 15). Also commonly diagnosed were esophageal reflux, dyspepsia, and noninfectious gastroenteritis. These conditions are all extremely common in the general population.
Irritable bowel syndrome can be defined as the presence of unexplained abdominal discomfort or pain for 12 weeks or longer during the previous 12 months. These symptoms have to be accompanied by two or more of three characteristics: 1) there is relief of symptoms with defecation, 2) the onset of symptoms was associated with new onset of either diarrhea or constipation, or 3) the onset of symptoms was associated with a change in stool consistency.199
Irritable bowel syndrome may affect as much as 15% of adults, with women more prone to report symptoms of this syndrome.199 In the U.S., gastroenterologists diagnose irritable bowel syndrome more frequently than any other condition. Twelve percent of visits to primary care providers may be due to symptoms of irritable bowel syndrome but only a minority of individuals with symptoms visit a doctor for evaluation. Patients seeking health care for this syndrome are more likely to have accompanying psychological problems than individuals who tolerate their symptoms.199
The cause of irritable bowel syndrome is unknown. A consistent pattern of organic disease has not been found among patients reporting symptoms of this illness. Generally, no abnormalities are detected in the physical examination or laboratory testing of patients. Therefore, the diagnoses of irritable bowel syndrome is based on exclusion of other causes of abnormal bowel symptoms.
Treatment of irritable bowel syndrome is primarily directed at alleviating symptoms. The avoidance of certain foods and the addition of fiber to the diet may be helpful. Finally, drug therapy analgesics, tricyclic compounds, and antispasmotics may help relieve symptoms.199
Diseases of the Genitourinary System (ICD-9 580-629)
Relatively few registry participants had a disease of the genitourinary system. This diagnostic category ranked 11th and 12th most frequent among the three clinical programs (Table 6). The low number of cases probably reflects the fact that women most often developed these types of health problems, but women represented only one out of ten registry participants. A common laboratory finding, hematuria (blood in urine), was the most common diagnosis in all three registries (Table 16).
Diseases of the kidney were very rare among registry participants. This finding is important because exposure to depleted uranium (DU) munitions during the Gulf War has been hypothesized to cause long-term health problems. Since DU is only slightly radioactive, its potential toxicity as a heavy metal is of greater concern. Because heavy metals are nephrotoxic, exposure can cause kidney disease.95
Diseases of the Skin and Subcutaneous Tissue (ICD-9 680-709)
Gulf War veterans often were diagnosed with a condition that fell within this diagnostic category, which ranked 3rd through 5th most common among broad ICD-9 classifications in the three registries (Table 6). The most common diagnosis within this category was skin rash due to contact with an irritant -- a condition that usually results from recent skin irritation (Table 17). Typical, male-pattern baldness and acne were also frequent diagnoses in this category.
Skin rashes are one of the most common complaints of all adult populations presenting for nonemergency health care. In a specific study of skin conditions among Gulf War veterans, a unique or unusual skin rash was not identified.200 In addition, military dermatologists did not observe unusual skin diseases during the Gulf War deployment.201
Diseases of the Musculoskeletal System and Connective Tissue (ICD-9 710-739)
In all three registries, this was the leading diagnosis (Table 6). Within this ICD-9 classification, back pain and nonspecific joint pain were the most common specific diagnoses (Table 18). These are also the most frequent musculoskeletal problems reported by outpatient populations of civilian adults.202 It is estimated that there are 315 million outpatient visits each year in the U.S. for musculoskeletal complaints.203 Also, 17% of adults report back pain.202 In contrast to musculoskeletal problems, connective tissue diseases were infrequently diagnosed among Gulf war registry participants.
Among military populations, musculoskeletal problems are prevalent because of vigorous training and physical fitness requirements. In particular, combat operations like those in the Gulf are associated with injuries to the musculoskeletal system.204,205 Specific studies of Gulf War veterans have demonstrated that veterans are suffering from the same types of musculoskeletal health problems as the general adult population. In a 1998 case-series, Escalante and Fischback tabulated the rheumatic manifestations of 145 VA registry participants who were referred for rheumatological evaluation.206 The investigators noted that the types of conditions Gulf War veterans experienced after the war were not unusual (fibromyalgia, nonspecific arthralgias, osteoarthritis, etc.) and they inferred no suggestion of a new Gulf War-related syndrome. However, these clinicians noted that " pain is common and widespread in these patients and their health related quality of life is poor."206
Three other teams of researchers studied a total of 698 Gulf War veterans evaluated for possible rheumatological conditions and noted a high prevalence of common health problems, but no indication of a unique Gulf War diagnosis or condition.207-209
Symptoms, Signs and Ill-Defined Conditions (ICD-9 780-799)
The diagnostic category, "Symptoms, Signs and Ill-defined Conditions," includes symptoms, signs, illnesses, and abnormal laboratory results or other investigative abnormalities, which are not coded elsewhere in the IDC-9-CM. It is important to note that the "Symptoms, Signs, and Ill-defined Conditions" category is not a diagnosis of an unusual or mystery illness. This diagnostic classification encompasses more than 160 sub-categories and primarily consists of common symptoms that do not have a definite cause or isolated laboratory abnormalities.
Among Gulf War veterans evaluated in the initial VA registry, the diagnostic category, "Symptoms, Signs and Ill-defined Conditions," ranked 13th most common (Table 6). Inpatients seen later, when a greater number of diagnoses were coded, this diagnostic category was more common, ranking 7th in the VAs revised registry and 2nd in the DoD CCEP. The reason for the different frequency of diagnosis between the revised VA registry and the CCEP was not clear but may be the result of coding preferences in the two clinical programs, because many of the diagnoses in the "Symptoms, Signs and Ill- Defined Conditions" classification can be coded adequately using ICD codes in other diagnostic categories.
Compared to other diagnostic groups, there was more diversity in ranking among the specific diagnoses within this category. Nevertheless, the most common diagnoses were widely varied and not clearly related to a particular disease (Table 19). Most of the diagnoses represented isolated laboratory test abnormalities or symptoms without accompanying pathology. For the twenty most common primary and secondary diagnosis in this category among the 49,079 patients examined in the initial VA Registry protocol, eight of the top ten were abnormal laboratory results or other investigative abnormalities. Abnormal laboratory results are routinely encountered when performing multiple tests on outpatients. For patients evaluated in the revised VA registry and the DoD CCEP, which coded more diagnoses, non-specific symptoms were tabulated more often than abnormal laboratory results. Five of the top ten diagnoses were the same in the revised VA registry and CCEP, and included the following common symptoms: headache, malaise and fatigue, other general symptoms, rash and nonspecific skin eruption, and other insomnia (Table 19).
These findings are comparable to other outpatient populations and the general civilian population. Physical symptoms, like those classified in "Symptoms, Signs and Ill-defined Conditions," are one of the most common reasons for outpatient visits. Published data from ambulatory care surveys found the following self-reported prevalence of symptoms: 210-212
The high prevalence of these particular symptoms in the general clinic population is also supported by data from the National Ambulatory Medical Care Survey (NAMCS).213 This national sample of clinics in the United States found that in 1989 the number of outpatient visits in the United States for fatigue was estimated to be 7 million; for headaches, 10 million; for joint pains, 17 million; and for skin rash, 14 million. Many patients report more than one of these symptoms at the same time.211
In adult populations, common symptoms often remain unexplained, despite diagnostic testing. In approximately one-third of patients presenting with a physical symptom, a precise physical cause of the symptom cannot be identified even after medical evaluation.210 Symptom-based diagnoses are increasingly recognized as prevalent and persistent problems among civilian populations, in which they are associated with high levels of subjective distress, functional impairment, and extensive use of health care services.214 In military populations, unexplained symptom-based conditions have been observed after military conflicts dating back to the U.S. Civil War 115,215 and have been observed in recent military deployments.143-146
Studies of patients with multiple, medically unexplained symptoms have found increased rates of major depression and panic disorder.212,216-221 Several mechanisms might account for this association. Physical illness may cause psychosocial distress through a direct biological link or by overwhelming a patients ability to cope.222 Distress also may increase unhealthy behaviors that increase the risk of such symptoms.216
Many studies also demonstrate that patients with primary anxiety and depressive disorders present with physical symptoms.223-225 For example, in a study of more than 1,000 health maintenance organization enrollees, increasing numbers of pain complaints were associated with elevated levels of anxiety, depression, and physical symptoms.226 Importantly, depression and anxiety are consistently associated with medically unexplained symptoms across many studies that have employed different research methodologies, including cross-sectional,227 casecontrol, 228-231 and longitudinal designs.232,233 In several studies, the percentages of patients with anxiety and depressive disorders increased with increasing numbers of physical symptoms.212,225 Katon and colleagues found that the relationship of physical symptoms to common anxiety and depressive disorders is linear: as the number of anxiety and depressive symptoms or lifetime episodes of psychological disorders increases, so does the prevalence and number of physical symptoms.223,228
In a study of 18,495 Gulf War veterans, physical symptoms often did not begin until after the war, with 40% of symptoms having a latency period of greater than one year. 234 Moreover, there was no association between symptoms and self-reported exposures. As in the civilian population, the study found a "dose-response" relationship between physical symptoms and coexisting psychological diagnoses. Like other studies, the more symptoms that Gulf War veterans reported, the more likely they were to have a psychological disorder.212,223,225,228
In another study, an intensive examination was conducted of 21,579 Gulf War veterans who had been evaluated in the CCEP and received a diagnoses of "Symptoms, Signs, and Ill-defined Conditions."161 More definitive, often psychological, diagnoses could be made by increasing the comprehensiveness of the evaluation and by multidisciplinary input. In this study, no evidence of a new or unique syndrome was found, and there was no association between a diagnosis of "Symptoms, Signs, and Ill-Defined Conditions" and wartime exposures. In a third study, more than one-half of U.S. Gulf War veterans with unexplained symptoms had a treatable anxiety or depressive disorder.235 In a fourth study, PTSD was diagnosed in 50% of veterans with complaints of chronic fatigue.236 Taken together, these data suggest that multiple, unexplained symptoms are often a marker of psychosocial distress.
It is important to note that the VA and DoD Gulf War health examination registries found that about 18 to 20% of participants have undiagnosed symptoms. In comparison, about 17 percent of Vietnam veterans participating in VAs Agent Orange registry have had undiagnosed symptoms (unpublished data, Dr. Han Kang, Department of Veterans Affairs, Washington, DC). These findings indicate that veterans from prior wars, regardless of the types of exposures, develop similar, difficult-to-explain health problems.
There were three symptom-based conditions that received a unique diagnostic code in the revised VA registry: chronic fatigue syndrome (990.02), sleep apnea (990.01), and fibromyalgia (990.03). Because these postulated conditions are not specifically coded in ICD-9, they were not tabulated in the initial VA registry or the CCEP. In the revised VA registry, chronic fatigue syndrome was diagnosed in 793 veterans, which is 3.7% of this registrys participants. Two hundred and sixty veterans (1.2%) were diagnosed with sleep apnea and 78 (0.4%) were diagnosed with fibromyalgia.
In table 21, a secondary diagnosis of "Symptoms, Signs and Ill-defined Conditions" was compared to primary diagnoses grouped by major ICD-9 classifications. As shown for the revised VA registry and CCEP, a secondary diagnosis of "Symptoms, Signs and Ill-defined Conditions" was found most frequently among veterans who also had this set of conditions as a primary diagnosis. In the revised VA registry, a secondary diagnosis of "Symptoms, Signs and Ill-defined Conditions" was also associated with a primary diagnosis of endocrine and blood diseases; and in the CCEP, this secondary diagnosis was associated with a primary diagnosis of musculoskeletal and digestive disorders. There is no indication that a secondary diagnosis of "Symptoms, Signs and Ill-defined Conditions" was a manifestation of an underlying mental or psychiatric disorder.
Injury and Poisoning (ICD-9 800-999)
This diagnostic classification also contains diverse conditions. Among registry participants, various sprains and strains were the most common diagnoses, which is consistent with the fact that musculoskeletal injuries are frequent among military personnel (Table 20).
Time Series Analysis
Table 22 presents the percent distribution of the symptoms recorded in the initial and revised VA registry databases for the calendar years 1992 to 1999. As reflected in the layout of the table, the initial VA registry codesheet was use to record examinations in the years 1992 through 1995, whereas the revised VA codesheet was used to record examinations from 1996 forward. The initial codesheet was different structurally than the revised codesheet, with the former recording up to three symptoms compared to at most 10 in the revised VA registry.
For any particular symptom listed in Table 22, the distribution by year of examination for either type of codesheet was reasonably stable. There is a marked increase in the percent of veterans reporting symptoms by year of examination when comparing the data recorded on the revised codesheet with the initial codesheet. This is a reflection of both the limited coding range employed in the initial codesheet and the fact that the revised codesheet could record up to 10 symptom codes compared to only three on the initial codesheet.
The most frequently reported symptom for each year in the initial VA registry was fatigue, but the most frequently reported symptom across all years in the revised VA registry was joint and muscle pain as observed in the CCEP (Table 23). The revised VA registry format is structurally closer to the CCEP than the initial codesheet by virtue of the potential number of symptoms that can be recorded. There did not appear to be any trailing off in the frequency of reported symptoms in the later years of the revised VA registry as was observed for the CCEP.
Table 23 presents the percent distribution of ten common symptoms in the DoDs CCEP by calendar year of examination for the years 1994 to 1999. The symptoms tabulated in the CCEP were captured using a checklist of symptoms administered to veterans during the evaluation. These are often referred to as physician-elicited symptoms. Within all of the symptom categories there is considerable variation from year to year in the percent of persons examined who report having the symptom in question. A common pattern for each of the reported symptoms in the CCEP is that compared to all other examination years, 1997 had the highest proportion of persons who report a particular symptom. Also, there is a dramatic drop off in the reporting of symptoms, with fewer symptoms reported in 1998 and 1999.
To evaluate these patterns further, the distribution of the number of symptoms reported by calendar year was assessed for the CCEP, which primarily evaluated active duty troops. The percent reporting no symptoms was 20%, 21%, 15%, and 16% from 1994 through 1997, but jumped to 92% and 74% in 1998 and 1999. These data suggest that just as the interest in enrolling for a CCEP examination trailed off in 1998 and 1999, the number of symptoms reported by each veteran and noted by the examining physician also trailed off. Despite the variation from year to year, within any given year, the top three most frequently reported symptoms were joint and muscle pain, fatigue, and headaches.
Table 24 presents the percent distributions of broad diagnostic categories by year of examination for diagnoses recorded either in the initial VA registry (years 1992-1995) or in the revised VA registry (years 1996-1999). Within either the initial or revised registry, the percent having a diagnosis in any one category appears to be fairly uniform across the applicable years. This was true except for the diagnosis of "Symptoms, Signs, and Ill-defined Conditions," which increased in frequency each year in the revised VA registry. As with the symptom table for the VA registry breakdown (Table 22), an increase in frequency for a particular diagnostic group is seen between the years covered by the initial VA codesheet and the time period covered by the revised codesheet. This increase is a reflection of the increased coding from a maximum of 3 diagnoses in the initial registry to a maximum of 10 diagnoses in the revised registry.
Diagnoses involving the musculoskeletal system and mental disorders are among the most frequently reported diagnoses in the VA registries, as with the CCEP diagnostic data (Table 25). While the category "Symptoms, Signs and Ill-defined Conditions" is among the top three diagnostic categories in the CCEP, it was almost never used on the initial VA codesheet and was not frequently used on the revised codesheet.
Table 25 presents the percent distribution of diagnoses by year of examination among the 32,876 veterans evaluated in the CCEP. Diagnoses again are grouped by broad diagnostic categories, and individuals with multiple diagnoses in a single category are counted only once in that category. Unlike the distribution of symptoms by year of examination (Table 23), the frequency of diagnoses reported in the CCEP is fairly uniform from year to year. Unlike self- reported symptoms, there was no unusual drop in percentages after 1997 for any of the diagnostic categories that would suggest a trailing off in the problems affecting Gulf War veterans. The three most frequently recorded diagnoses in the CCEP were those involving the musculoskeletal system; symptoms, signs, and ill-defined conditions; and, mental disorders.
Veterans Evaluated in Both the VA and DoD Registries
There were 2,922 veterans who were examined in both the VA and DoD clinical evaluation programs. This is approximately 3% of all registry participants. Comparing these veterans demographic and military characteristics to that of the entire group of registry participants shows that the overlap group more closely resembled CCEP participants, except for the fact that a greater percentage of the overlap group (36%) served with the National Guard and Reserves than did the entire group of CCEP participants (9%). This is consistent with the fact that National Guard and Reservists generally would have been eligible for a VA registry examination earlier than veterans whose unit component designation was "Active Duty."
The other area where the overlap group was different from the overall CCEP group was the fact that a greater proportion of the overlap group served only in Saudi Arabia (43% vs 32%), away from direct combat. This is consistent with the fact that a larger percentage of the overlap group belonged to the National Guard and Reserves, a military component less likely to be sent to Kuwait or Iraq as ground troops.
Table 26 presents the distribution of symptoms reported by veterans in the overlap group. Compared to the symptoms recorded in the CCEP and the symptoms recorded in the VA registries (Table 5), every symptom category in the overlap group was reported at a higher rate. These data indicate that veterans in the overlap group report greater ill health on average than the registry participants who underwent only one examination.
For diagnoses in the overlap group (Table 27), musculoskeletal diseases and mental disorders were the most frequently recorded diagnoses, as was true for overall participants in the three registries (Table 6). Also, the frequency of many diagnostic categories was similar between the overlap group and the general registry populations.
In summary, the group of 2,922 Gulf War veterans seeking multiple registry examinations report more ill health than other registry participants but have not received a greater number of diagnoses. VA plans further research of this unique group of veterans (see Future Research).
Analysis of Spouses and Children of Veterans
As of October 31, 2001, a total of 1,121 individuals had participated in the clinical program that evaluated spouses and children of veterans. Included in the program were 415 adult female spouses, 19 adult male spouses, 355 female children of veterans, and 332 male children of veterans. The following table provides the most common diagnoses among spouses and children. No unusual health problems were observed in this population.
| First Page | Prev Page | Next Page |