Epidemiological Analysis

Methods – Epidemiological Analysis

A second important area of research is to compare diagnostic information contained in these clinical databases with several major sources of exposure data. Although registry participants were self-selected and not chosen randomly from the entire Gulf War veteran population, limited epidemiologic analysis can be conducted to generate hypotheses about the causes of veterans' health problems. Because innumerable possible comparisons can be conducted using the registry databases and exposure data, all interpretation of statistical analyses has to consider: 1) the occurrence of chance findings; 2) biologic plausibility; and, 3) likely bias and confounding from selection and reporting bias. Available sources of exposure data included the following:

  1. One type of exposure data is the self-reports of veterans. Registry participants were asked about potentially hazardous exposures during the Gulf War. For the combined database of these three clinical programs, the following self-reported exposures are available for analysis:
      1. Diesel and other petrochemical fumes
      2. Smoke from tent heaters
      3. Burning trash and feces
      4. Smoke from oil fires
      5. Passive smoking
      6. Skin exposure to fuels
      7. Paints and solvents
      8. CARC (chemical agent-resistant coating)
      9. Personal use of pesticides and insect repellents
      10. Pyridostigmine bromide tablets
      11. Microwaves
      12. Depleted uranium (DU)
      13. Anthrax vaccination
      14. Eating food not provided by the military
      15. Eating or drinking contaminated food
      16. Bathing in water not provided by the military

Exposure data can be compared with diagnostic information to determine whether there are significant associations. However, this analysis is limited by the fact that self- reported exposures rarely can be verified or quantitated and are subject to recall and reporting biases.237

  1. In addition to self-reported information, investigations can be conducted using other sources of exposure data derived from extensive studies conducted to evaluate Gulf War health questions using the Deployment Environmental Surveillance Program Geographic Information System (GIS). A GIs has been established that provides information on the daily location of U.S. Gulf War troops -- both at the unit level and for individuals -- from the beginning of Operation Desert Shield (August 8, 1990) to 10 months after the ground war (December 31, 1991). This database was constructed by examining all existing Gulf War records, such as troop unit logbooks and situation reports, that contained daily troop-unit location data by latitude and longitude. Over 5 million records were examined. The individual personnel in each troop unit were determined from the Defense Manpower Data Center's Gulf War Registry.

    This GIs database can be used to conduct a number of assessments, which compare battlefield geographic locations with participation in the three registries and with diagnostic findings from the registries.198

Results – Epidemiological Analysis

Because of the massive amount of complex data, it was possible to provide only basic analysis and preliminary results in this monograph. More detailed analyses can be conducted with the linked database of registry data. The proportional distribution for the time of arrival in the Gulf Theater of operations was very similar in the three registry programs (Table 29). Persons examined in the CCEP were more likely than VA registry personnel to have served in Iraq or Kuwait, but the percentage having combat MOS codes were similar across all registry programs. It is important to control for these characteristics in epidemiological assessments of registry data because of potential differences in exposures among the various Gulf War veteran populations.

Table 30 provides information on self-reported exposures during the Gulf War. A high proportion of registry participants reported a wide variety of potential exposures. The most commonly reported exposures were to: 1) petroleum fumes and fuel, 2) passive cigarette smoking, 3) burning trash or feces, 4) oil well fire smoke, and 5) local food.

In table 31, participants in the three registries were compared with all Gulf War veterans to determine whether demographic characteristics were associated with enrollment and completion of a VA or DoD registry examination. Women Gulf War veterans were significantly more likely to participate in a registry examination than men. This finding is consistent with civilian health care practices: Women are more likely than men to visit a health care provider for a broad range of health complaints.

There was a linear relationship between age and enrollment in a registry. The older the veteran, the more likely they were to participant in one of the registries. Married veterans also were slightly more likely to enroll for a registry examination, but this finding may reflect the fact that older (and more often married) veterans had a registry examination. For unknown reasons, veterans whose home of record was the Southwest were the least likely to enroll in one of the registries. There was no clear association between the race/ethnicity of Gulf War veterans and enrollment for a registry examination.

In table 32, participants in the three registries were compared with all Gulf War veterans to determine whether military characteristics were associated with completing a registry examination. Reservists were much more likely to have participated in one of the clinical evaluation programs. This fact has been pointed to as possible evidence that Reservists were not in as good general health as active duty troops. However, higher rates of participation in the registries by Reservists may indicate a problem with access to health care, rather than more severe health problems. Because active duty personnel had ready access to free DoD health care after the war, they had the opportunity to have their questions and concerns answered by a knowledgeable health care provider during routine health care. Active duty troops did not need to undergo a registry examination to address most health complaints.

In contrast to active duty troops, Reservists and National Guard personnel lost access to the military health care system as soon as they re-deployed to the United States and returned to an inactive military status. For no-cost VA health care, Reservists and National Guard personnel generally have to demonstrate a service-connected health problem or meet a financial means test. As a result, the VA Gulf War registry became a primary source of high-quality health care for many veterans from the Reserves and National Guard, particularly veterans of limited means or with insufficient private health insurance. Higher numbers of non-active duty veterans therefore could be expected to take advantage of the VA program during the first few years after the war. The finding that enlisted personnel were significantly more likely to enroll for a registry examination may also reflect a relative lack of private health care insurance compared to officers with greater access to civilian health care (Table 32).

Army and Marine Corps troops were more likely to participate in the registries than Air Force or Navy personnel (Table 32). This finding may be due to the more arduous duties of ground troops, who often sustain musculoskeletal injuries – one of the most common health problems of registry participants. The higher rate of participation of ground troops in the registries could also result from adverse exposures during the ground war, which were not shared by Air Force and Navy personnel. For example, ground troops exposed to oil well fire smoke as determined by the GIs were more likely to participate in the registries (Table 33). However, it is noteworthy that the veterans who had stayed in the theater of operations for the longest period of time were only slightly more likely to obtain a registry examination (Table 33). Additionally, in one epidemiologic study of Gulf War veterans, the period of deployment to the Arabian Gulf was not associated with unexplained physical symptoms.238

In table 33, participants in the three registries were compared with all Gulf War veterans to determine whether certain Gulf War exposures were associated with completing a registry examination. As already noted, exposure to oil well fire smoke was associated with registry participation. In addition, troops near the destruction of chemical warfare agents at Khamisiyah were more likely to participate in one of the registries. However, these veterans had been contacted individually by letter after the extent of possible CW exposure was determined, and the letter recommended that they obtain a registry evaluation.

Two studies have been conducted that focused on veterans who may have been exposed to CW agents after the demolitions at Khamisiyah: No association was found between having been in proximity to the demolitions and hospitalization after the war in a DoD hospital.239 And, no increase in mortality was observed among veterans who may have been exposed to sarin and cyclosarin as determined by two plume models of the Khamisiyah demoltions.240

Table 33 also shows that receiving the anthrax or botulinum vaccine was not associated with registry participation. In contrast, a history of hospitalization before the war was associated with later participation in the registries. This finding indicates that veterans who had health problems before deployment had greater health problems after the war compared to other veterans.


| First Page | Prev Page | Next Page |