Smith et al. (2006) compared cause-specific postdeployment hospitalization in DoD military treatment facilities during the period October 1, 1988, through December 31, 2000, among US active-duty servicemembers who served in the Gulf War (n = 455,465), Southwest Asia following the Gulf War (n = 249,047), or Bosnia, also following the Gulf War (n = 44,341). After adjusting for age, sex, marital status, pay grade, race/ethnicity, service branch, occupation, and predeployment hospitalization, the hazard ratio (HR) for nondefined neoplasms was 1.03 (95% CI 0.93-1.15) for the Gulf War deployed veterans compared to the Southwest Asia cohort and 0.61 (95% CI 0.50-0.76) compared to the Bosnia group. The adjusted HR for testicular cancer in Gulf War veterans was 0.64 (95% CI 0.32-1.28) and 0.80 (95% CI 0.27-2.39) compared to the Southwest Asia and Bosnia groups, respectively.
The UK Defence Analytical Service Agency (DASA, 2009) published summary statistics comparing mortality rates of 53,409 UK Gulf War veterans with those of 53,143 UK armed forces personnel of similar age, sex, service status, and rank who were in service at the same time, but not deployed to the gulf (era cohort). It reported 209 and 228 malignant neoplasms among the gulf and era cohorts, respectively (age-adjusted mortality RR 0.97, 95% CI 0.81-1.18). They did not observe any significant associations for specific neoplasms.
The Canadian Department of National Defense used the national mortality database and the national cancer registry to examine mortality rates and cancer incidence among Canadian Gulf War veterans from 1991 through 1999 (Statistics Canada, 2005). Two cohorts were established—the deployed cohort consisting of 5117 servicemembers sent to the gulf between August 1990 and October 1991, and the nondeployed cohort of 6093 servicemembers who were eligible for deployment but were not deployed. During the follow-up period, 10 deaths from cancer were identified in the deployed cohort and 15 in the nondeployed. The age-adjusted HR was 0.85 (95% CI 0.38-1.90). Among the deployed and nondeployed cohorts 29 and 42 incident cancers, respectively, were identified (age-adjusted HR 0.86, 95% CI 0.54-1.39). The largest number of the cases were cancers of the digestive tract (n = 15). There were also 8 testicular cancers and 4 brain cancers (all among the nondeployed). There was no evidence of an association between deployment and these specific cancers.
In continued follow-up of the study by Macfarlane et al. (2003) (discussed above), there was still no excess risk of mortality from malignant neoplasms with 2 more years of data (RR 1.01, 95% CI 0.79-1.30) (Macfarlane et al., 2005).
The Update committee identified eight secondary studies of multiple outcomes that had been included in Volume 4, but had not been considered in that review of malignant neoplasms (Goss Gilroy, 1998; Iowa Persian Gulf Study Group, 1997; Ishoy et al., 1999a; Kang et al., 2000; Kelsall et al., 2004a; McCauley et al., 2002; Simmons et al., 2004; Steele, 2000). Because there is specific mention of cancer in these studies, they are described here in chronological order.
From September 1995 through May 1996, the Iowa Persian Gulf Study Group (1997) performed a cross-sectional telephone survey to solicit self-reported illness in Iowan military personnel active during the Gulf War (n = 4886). Members of the National Guard who had been deployed (n = 911) were more likely to report any cancer than nondeployed National Guards members (n = 831) (prevalence difference [PD] 1.3, 95% CI 0.6-2.0), but the prevalence was similar compared with the deployed (n = 985) and nondeployed regular military (n = 968) (PD 0.3, 95% CI −0.6-1.2). Specific reports of skin cancer followed a similar pattern.