hospitalizations for blood diseases (identified according to ICD-9-CM discharge codes) was similar in the three cohorts (Southwest Asia vs Gulf War veterans: HR 0.93, 95% CI 0.80-1.07; Bosnia vs Gulf War veterans: HR 0.93, 95% CI 0.75-1.15).
An earlier analysis by Smith et al. (2002) compared postwar hospitalizations among 405,142 active-duty Gulf War veterans who left the region after the war. Data for DoD hospitals were compared through July 1999 according to levels of exposure to oil-well fires in 1991 ranging from unexposed to an average daily exposure of greater than 260 μg/m3 for more than 50 days. The duration and length of exposure to particulate matter from the smoke were based on meteorological data, diffusion modeling and troop location data. This study did not observe a clear association between oil-well fire exposure and hospitalizations for blood disorders. The adjusted risk ratios for blood disorder hospitalization in those with the highest exposure to oil-well fires (> 260 μg/m3 for > 50 days) was 0.9 compared to those with no exposure. As with other studies of hospitalizations in DoD hospitals, the main limitations of this study include the lack of information on outpatient diagnosis as well as hospitalizations among those who left the service.
Smith et al. (2003) also examined hospitalizations for blood disorders according to potential exposure to nerve agents from the Khamisiyah demolition. Exposure to nerve agents was modeled following the 2000 Khamisiyah gaseous hazard area modeling done by the DoD. Rate of hospitalizations in DoD hospitals for these disorders were equivalent in those potentially exposed and nonexposed (risk ratio 0.96, 95% CI 0.89-1.03).
Two studies measured hematologic parameters in deployed and nondeployed veterans. A study conducted in 1997 among 686 Danish Gulf War veterans and 231 Danish nondeployed controls measured hemoglobin and blood cell counts in both groups (Ishoy et al., 1999b). In bivariate analyses, no differences were observed between deployed and nondeployed in blood hemoglobin (9.3 mmol/L in both groups), erythrocyte count (4.8 million/L in both groups), hematocrit (0.44 in both groups), mean corpuscular volume (91 10−15 L per cell in both groups), and leukocyte count (5.8 109/L in deployed and 5.9 109/L in nondeployed). Leukocyte fractions were also comparable. Platelet counts were slightly lower in deployed compared to nondeployed (205 109/L vs 211 109/L, p < 0.05). Differences in response rate (84% in deployed and 58% in nondeployed) and lack of adjustment for potential confounders reduces the usefulness of these results.
The Australian Gulf War Veterans’ Health Study (Sim et al., 2003) measured hemoglobin and other hematologic parameters in 1355 male and 30 female Gulf War veterans and in 1361 male and 32 female nondeployed veterans. In the males, hemoglobin (153.1 g/L vs 153.4 g/L), mean corpuscular volume (91.6 femtoliters [fl]2 vs 91.5 fl2), mean corpuscular hemoglobin (30.4 picograms[pg]3 vs 30.5 pg)3, platelets (227.8 109/L vs 231.3 109/L), and leukocytes (6.3 109/L vs 6.2 109/L) were similar in deployed and nondeployed veterans. No differences were observed for leukocyte fractions. In female veterans, deployed and nondeployed also presented similar hematologic parameters (hemoglobin: 131.8 g/L vs 134.3 g/L; mean corpuscular volume: 92.8 vs 93.4 fl; mean corpuscular hemoglobin: 29.8 pg vs 30.3 pg; platelets: 263.6 109/L vs 269.6 109/L; and lymphocytes: 2.0 109/L vs 2.1 109/L).
The committee did not identify any secondary studies of diseases of the blood or blood-forming organs in Gulf War veterans.