IJE Advance Access originally published online on October 26, 2005
International Journal of Epidemiology 2005 34(6):1403-1408; doi:10.1093/ije/dyi205
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Articles |
Long-term mortality amongst Gulf War Veterans: is there a relationship with experiences during deployment and subsequent morbidity?
1 Epidemiology Group, Department of Public Health, School of Medicine, University of Aberdeen, Aberdeen AB25 2ZD, UK.
2 Gulf War Illness Research Unit, Department of Psychological Medicine, Guy's, King's and St Thomas' School of Medicine, London SE5 8AZ, UK.
3 London School of Hygiene and Tropical Medicine, University of London, London WC1E 7HT, UK.
4 Health Statistics Branch, Defence Analytical Services Agency, Ensleigh, Bath BA1 5AB, UK.
5 Arthritis Research Campaign Epidemiology Unit, Division of Epidemiology and Health Sciences, The University of Manchester School of Medicine, Manchester M13 9PT, UK.
* Corresponding author. Epidemiology Group, Department of Public Health, School of Medicine, Polwarth Building, Foresterhill, Aberdeen AB25 2ZD, Scotland, UK. E-mail: g.j.macfarlane{at}abdn.ac.uk
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
Background Gulf War Veterans have previously been shown to have, in the short-term, an excess risk of death from ‘external’ (i.e. non-disease) causes of death. This study aims to determine whether there remains an excess of non-disease-related deaths in Gulf Veterans, 13 years after deployment, and, for the first time, to determine whether there is a relationship between experiences reported in the Gulf, post-war symptoms, and subsequent mortality experience.
Methods We conducted a cohort study with follow-up from April 1, 1991 (the end of the Gulf War) to June 30, 2004. Participants were 53 462 Gulf War Veterans and a cohort of military personnel, matched for age-group, sex, rank, service and level of fitness, who were not deployed to the Gulf. The outcome measure used was mortality as recorded on the NHS central register.
Results There is no difference, 13 years after the end of the Gulf War, in the overall mortality experience of Gulf War Veterans. The excess in non-disease-related deaths previously reported is confined to the initial 7 years of follow-up [mortality rate ratio (MRR) 1.31, 95% confidence interval (CI) 1.06–1.63] rather than the more recent period (MRR 1.05, 95% CI 0.83–1.33). Overall experiences reported during Gulf deployment did not influence subsequent risk of dying, but there was non-significant increased risk of dying from a disease-related death (MRR 1.99, 95% CI 0.98–4.04) associated with reported exposure to depleted uranium and of a non-disease-related death associated with reporting handling of pesticides (MRR 2.05, 95% CI 0.91–4.61). Reporting of morbidity in the health surveys conducted was not related to future risk of death.
Conclusion The higher rates of non-disease-related deaths in Gulf War Veterans is not evident in the period of follow-up since 1997. Neither the excess morbidity reported in health surveys nor the experiences during deployment significantly influenced future mortality. The two non-significant associations found (reported depleted uranium exposure and disease death, reporting handling pesticides and non-disease deaths) need to be considered in the context of the number of possible associations examined and potential biases—although they are biologically plausible.
Keywords Gulf War, mortality, cohort study, depleted uranium, pesticide
Accepted 8 August 2005
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Epidemiology studies have consistently demonstrated that UK Gulf War Veterans experience an excess of self-reported morbidity in comparison with those who were not deployed.1–3 Mortality experience (up to 8 years after the conflict) has been found to be, overall, almost identical between the deployed and non-deployed military personnel: although this masks a slight excess of external causes of death and deficit of disease-related causes amongst veterans.4,5 In some studies there is little specificity in the relationship between symptoms and reported experiences and exposures,1 but in others particular self-reported experiences in the Gulf have been associated with biologically plausible symptoms. For example, in the Manchester study of UK Gulf War Veterans those who reported exposure to smoke from burning oil-well fires were more likely to report a cluster of respiratory symptoms.6
In this report we first examine mortality experience in the longer term (13 years after the end of the Gulf War), in particular, to evaluate whether there is still an excess of non-disease-related deaths amongst Gulf War Veterans. Second we examine whether, amongst veterans who participated in at least one of the three morbidity surveys 
6 years after deployment, there is a relationship between the self-reported experiences during deployment or morbidity reported subsequently and future mortality.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
We carried out a cohort study of Gulf War Veterans and a comparison group of non-deployed service personnel. Details of the study have previously been described in detail.5,7 In brief, the Gulf cohort consisted of all UK armed forces personnel who served in the Gulf region sometime between September 1990 and June 1991, while the comparison (Era) group was of the same size randomly selected from members of the armed forces who were in service on January 1, 1991 but who were not deployed to the Gulf area and group-matched for age-band, sex, service branch, rank and (for the army and Royal Air Force) fitness for active service [see Figure 1; Ref. (5) for details]. Details of the cohort members were sent to the Office for National Statistics (ONS), and in 96% of the cases the study subject was identified on the National Health Service (NHS) central register allowing information to be obtained on vital status and (if relevant) details of the date and cause of death [coded to the International Statistical Classification of Diseases and Health-Related Conditions (ICD) version 10].
Data were analysed using a Cox proportional hazards model and the comparison between the Gulf and Era cohorts expressed as a mortality rate ratio (MRR) with a 95% confidence interval (CI). Person years at risk of death were calculated from April 1, 1991 (the nominal end of the Gulf War) to the earliest of either date of emigration from the UK, date of death, or June 30, 2004. Analyses were adjusted for age (years). Supplementary analyses were conducted dichotomizing the follow-up period up to and including January 1, 1997, and the period beyond.
The Gulf and Era cohorts also acted as a sampling frame for three morbidity surveys conducted between 1997 and 2001; two general health surveys conducted by teams at the University of Manchester2 and King's College London1 and one of reproductive health (which surveyed the whole cohort) conducted by the London School of Hygiene and Tropical Medicine8 (Table 1). Subjects in both cohorts were invited to participate in at most two surveys (one study of general health and the reproductive health study). These studies collected information on self-reported experiences during the Gulf War and lifestyle (possible confounding factors of the relationship between deployment exposures and death, such as smoking and alcohol consumption). We used information from the largest study, on reproductive health, where available (Study 3) and, if not, from the relevant general health study (Studies 1 and 2). Amongst the survey participants we compared the mortality experience and then, within the Gulf cohort, we examined the relationship between self-reported experiences in the Gulf and future mortality using the same analytical techniques as described above. Person years at risk of death were calculated from the date of completing the relevant questionnaire.
|
Finally, using information from persons who participated in one of the two general health surveys we examined whether reporting high levels of morbidity in these two surveys was associated with an increased risk of dying. Details of the methods of collecting information on morbidity in these two studies have been described in detail previously.1,5 In this analysis of the two studies we divided persons into tertiles of reported symptoms based on, in Study 1, the number of troublesome symptoms they reported and, in Study 2, on the number of symptoms they scored as 50 mm or more on a 100 mm visual analogue scale on the degree to which they were troublesome.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Both the Gulf and Era cohorts initially consisted of 53 462 members. We excluded from follow-up 44 Gulf Veterans and 10 from the Era cohort who, prior to April 1, 1991, had died and two from each cohort who had emigrated. A further 3990 people (1655 Gulf and 2335 Era) could not be identified on the NHS central register. Further information from the Ministry of Defence excluded 8 and 307 Gulf and Era cohort members, respectively, who were subsequently found not to meet the entry criteria leaving 51 753 and 50 808 Gulf and Era cohort members, respectively, who were followed-up. During follow-up 637 Gulf cohort members and 629 Era cohort members died (MRR 1.03, 95% CI 0.92–1.15). Despite no overall difference in the death rates between cohorts, the Gulf cohort experienced slightly lower mortality from disease-related causes (MRR 0.91, 95% CI 0.77–1.07), pre-1997 (MRR 1.08, 95% CI 0.78–1.49) post-1997 period of follow-up (MRR 0.86, 95% CI 0.71–1.03) (Table 2). In contrast there was significantly higher mortality from external causes (MRR 1.19, 95% CI 1.02–1.39), an excess, which occurred pre-1997 (MRR 1.31, 95% CI 1.06–1.63) rather than post-1997 (MRR 1.05, 95% CI 0.83–1.33). Amongst the external causes of death, the strongest relationship was with transport accident deaths (MRR 1.44, 95% CI 1.13–1.84).
|
Amongst the Gulf cohort, 29 363 subjects completed a questionnaire in at least one of the morbidity surveys, and there were 138 deaths in such persons (0.47% of the cohort or 93 per 100 000 person years). We next examined the relationship, within the Gulf cohort, of self-reported experiences in the Gulf to subsequent mortality (Table 3). For anti-biological warfare prophylaxis (inoculations such as pertussis, plague, anthrax, or the combination) the mortality rates were very similar between those who did and did not report having them. There were no increased mortality risks associated with reporting any type of pesticide use or handling. Neither smoke from oil well fires nor reported experiences of small arms fire or Scud missiles passing nearby influenced subsequent mortality. Reported exposure to depleted uranium was uncommon amongst those deployed and, overall, there was no significant increased risk of death. Those who reported coming under fire had a significantly lower risk of subsequent death (MRR 0.5). However, after adjustment for potential confounding factors the lower risk was attenuated and not statistically significant.
|
For some of the exposures it is more plausible that they could be associated with either disease or non-disease-related deaths but not both. We, therefore, examined the previous self-reported exposures in relation to disease and external causes of death separately (Table 4). No self-reported experience in the Gulf was related to external causes of death. Those who reported handling pesticides (7% of those deployed) were at twice the risk of dying from an external cause, but this was not statistically significant (MRR 2.05, 95% CI 0.91–4.61). The causes of (non-disease) deaths amongst those reporting exposure to handling pesticides were one transport accident and six deaths from intentional self-harm or ‘accident of undetermined intent’. With respect to disease related causes, the small proportion of Gulf Veterans who reported exposure to depleted uranium (7%) experienced a doubling in the risk of dying, although again the result was not statistically significant (MRR 1.99, 95% CI 0.98–4.04). Of the nine people who reported exposure to depleted uranium and who died from a disease-related cause, seven were from cancer: three malignant cancers of the oesophagus, three malignant cancers of the brain, and one cancer of the brain of uncertain behaviour.
|
The final analysis considered whether morbidity reported in the health surveys affected subsequent mortality. For this analysis we used the sub-group of 5954 subjects responding to one of the two general health (rather than specifically reproductive health) surveys. Mortality did not differ between the tertiles of symptom count (Table 5) nor was there any evidence of a difference between the two studies in the effect of symptom count on mortality (data not shown).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
This study has demonstrated that 13 years after the end of the Gulf War there remains no difference, overall, in the mortality experience of Gulf War Veterans in comparison with other veterans who were not deployed. However, over the period as a whole, there were small differences in rates with respect to disease (decreased mortality) and non-disease (increased mortality) causes. For the first time, this study has examined whether the risk of dying is related to reported experiences during the Gulf. Overall it was not, but there was a 2-fold risk of dying from a disease-related cause associated with reported exposure to depleted uranium and of a non-disease cause associated with reporting handling of pesticides, although neither was statistically significant and the number of cases was low. High levels of morbidity reported in the previously conducted general health surveys were not related to an increased risk of death in the follow-up period.
The results of this study are similar to the first report on mortality amongst UK troops 8 years post-conflict.5 It is also consistent with the early mortality experiences of US Gulf War Veterans.9 Overall, however, mortality amongst Gulf War Veterans is considerably lower than would be expected based on age-sex mortality rates in the UK general population (www.dasa.mod.uk/natstats/gulf/intro.html). With respect to the excess of non-disease-related deaths, it has been suggested10 that it may be owing to suicide. However, in the analysis, deaths specifically attributed to suicide are not in excess; instead the strongest relationship was with transport accidents. One hypothesis is that service in the Gulf has altered deployed personnel's perception of danger and that they may be willing to take greater risks putting themselves at greater risk of an accidental death (e.g. in exceeding speed limit). The observation that the effect is only in the first period of follow-up (pre-1997) is consistent with this hypothesis. This could be tested in future, prospective, studies of deployed personnel. Importantly such an excess of non-disease-related deaths has been reported amongst veterans in the first few years after the Second World War, the Korean War, and the Vietnam War—and so there is unlikely to be a unique cause related to the Gulf War.11,12
There were two exposures, the reporting of which was related to an approximate doubling in risk of death. Handling pesticides was related to non-disease deaths and depleted uranium was linked to a disease-related death, although both results were non-significant. First, it is important to acknowledge that this study had limited power to examine the effects of exposures that were uncommonly reported, e.g. for an exposure reported by 7% of subjects there was 30% power to detect a 50% increase in the risk of death and 70% power to detect a doubling of risk. Second, even though the proportions of persons reporting these exposures is low, this equates to over 2000 persons—many more than are believed to have had these experiences. Exposure to pesticides has been associated with mood changes,13,14 but the long-term consequences of exposure are unclear. The non-disease-related deaths amongst persons reporting handling of pesticides were related to suicide or possible suicide. Nevertheless, amongst persons experiencing low mood at the time of completing the survey questionnaires there exists the possibility of differential recall of past experiences, which may, artefactually, have resulted in the apparent link.
Depleted uranium has been suggested as a cause of so-called ‘Gulf War Syndrome’—but as previously found, this study has confirmed that reports of exposure to depleted uranium are uncommon amongst Gulf War Veterans. In this study it was associated with an increased risk of death, albeit non-statistically significant. The distribution of cause of death was also unusual—seven from nine deaths were cancers of the oesophagus and brain. A recent review of the evidence surrounding the health effects of exposure to depleted uranium concluded that ‘Exposure to sufficiently high levels might be expected to increase the incidence of some cancers, notably lung cancer and possibly leukaemia, and may damage the kidneys’. Using a worst-case scenario, they estimated an extra 1.2 deaths per 1000 from lung cancer amongst those with the highest exposure (e.g. personnel in a vehicle struck by a depleted uranium penetrator).15 We found no deaths from lung cancer. Our previous study did not find an association between reported exposure to depleted uranium and registered incident cancers.7 If persons with these cancers had completed the questionnaire after diagnosis, there is again the possibility of recall bias—particularly given the amount of media interest surrounding depleted uranium. Each of the seven study subjects with cancer who reported exposure to depleted uranium participated in the study of reproductive health, and it is from this study questionnaire that their ‘exposure’ information has been taken. At the time of questionnaire completion subjects were asked ‘Since 1990, have you experienced any new medical problems or changes in your general health?’ Two subjects reported having been diagnosed with a brain tumour (one without dates provided, and one with dates of diagnoses in 1993 and 1997). Both these subjects stated that they considered themselves to have Gulf War Syndrome. The remaining five subjects did not make any mention of a cancer diagnosis and none stated that they considered themselves to have Gulf War Syndrome.
The collection of information in a similar, and in most cases the same, way by different studies allowed us additionally to examine the consistency in the reporting of ‘exposures’ between studies. This was done using the Kappa statistic that measures agreement over agreement expected by chance. It varies between 0 (chance agreement only) and 1 (perfect agreement). There is no agreed definition of values, but the values 0.4–0.6 are often interpreted as indicating moderate agreement and above 0.6 as good agreement. The data in Table 6 shows at least moderate agreement between studies for most exposure reports and good agreement for some.
|
In summary, this manuscript has demonstrated that Gulf Veterans, considering the entire time since the war, have experienced significantly higher rates of external deaths and slightly lower rates of disease deaths. The higher rate of deaths from external causes is no longer evident in the second half of the follow-up period. Neither the excess morbidity reported in the health surveys nor the experiences during deployment significantly influenced subsequent mortality. However, there were associations observed between the reporting of handling pesticides and possible suicide, and reporting exposure to depleted uranium and cancer death. These two associations, however, do need to be considered in the context of the number of exposure/death associations we have examined, and the fact that neither was statistically significant. Although, our study has found no long-term increased risk of mortality, identifying risk factors for specific outcomes (e.g. accidental death), designing and evaluating interventions within deployed populations should remain a high research priority.
|
Acknowledgments |
|---|
G.J.M., M.H., N.M., and M.L. were responsible for the design of the study. All authors participated in its conduct. The statistical analysis was conducted and the manuscript was written by G.J.M. and M.L., and then all authors substantially contributed to revising and finalizing the document. All authors are guarantors for the manuscript. We acknowledge the contribution of other researchers to the conduct of the UK morbidity surveys and mortality study: David Baxter, Nicola Cherry, Bill Coker, Francis Creed, Anthony David, Graham Davies, Graham Dunn, Susan Ferry, Lisa Hull, Khalida Ismail, Samantha Lewis, Ian Palmer, Margo Pelerin, Sue Prior, Patrick Sampson, Alan J Silman, Joanne Smedley, Stewart Taylor, Catherine Unwin, and Simon Wessley. We thank Scott Williamson and Anne Fedrick (the Defence Analytical Services Agency (DASA) who analyse statistics on behalf of the Ministry of Defence) who identified the study cohorts and provided considerable administrative help in ensuring the provision of the necessary information. DASA is continuing to publish information on mortality amongst Gulf War Veterans biannually. The study was funded by the Ministry of Defence, but the views expressed are solely those of the authors. Ethical approval for the conduct of the study was obtained from The University of Manchester Committee for the Ethics of Studies on Human Beings.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Unwin C, Blatchley N, Coker W et al. Health of UK servicemen who served in Persian Gulf War. Lancet 1999;353:169–78.[CrossRef][Web of Science][Medline]
Cherry N, Creed F, Silman F et al. Health and exposures amongst UK Gulf War veterans. Part I: The pattern and extent of ill health. Occup Environ Med 2001;58:291–98.
Simmons R, Maconochie N, Doyle P. Self-reported ill health in male UK Gulf War veterans: a retrospective cohort study. BMC Public Health 2004;4:27.[CrossRef][Medline]
Kang HK, Bullman TA, Macfarlane GJ et al. Mortality among US and UK veterans of the Persian Gulf War: a review. Occup Environ Med 2002;59:794–99.
Macfarlane GJ, Thomas E, Cherry N. Mortality among UK Gulf War Veterans. Lancet 2000;356:17–21.[CrossRef][Web of Science][Medline]
Cherry N, Creed F, Silman F et al. Health and exposures amongst UK Gulf War veterans. Part II: The relation of ill health to exposure. Occup Environ Med 2001;58:299–30.
Macfarlane GJ, Biggs A-M, Machonochie N, Hotopf M, Doyle P, Lunt M. Incidence of cancer among UK Gulf War veterans: cohort study. BMJ 2003;327:1373–75.
Maconochie N, Doyle P, Davies G et al. The study of reproductive outcome and the health of offspring of UK veterans of the Gulf War: methods and description of study population. BMC Public Health 2003;327:1373.
Kang HK, Bullman, TA. Mortality among U.S. veterans of the Persian Gulf War. N Engl J Med 1996;335:1498–504.
Bell NS, Amoroso PJ, Wegman DH et al. Proposed explanations for excess injury among veterans of the Gulf War: a call for greater attention from policymakers and researchers. Inj Prev 2001;7:4–9.
Nefgzer MD. Follow-up studies of World War II and Korean War prisoners: I. Study plan and mortality findings. Am J Epidemiol 1970;91:123–28.
Boyle CA, Decoufle P. Postservice mortality among Vietnam veterans. JAMA 1987;257:790–95.
Stallones L, Beseler C. Pesticide poisoning and depressive symptoms among farm residents. Ann Epidemiol 2002;12:389–94.[CrossRef][Web of Science][Medline]
Mearns J, Dunn J, Lees-Haley PR. Psychological effects of organophosphate pesticides: a review and call for research by psychologists. J Clin Psychol 1994;50:286–94.[Web of Science][Medline]
Royal Society. The health hazards of depleted uranium munitions Part I (2001).
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  M. S Kaplan, N. Huguet, B. H McFarland, and J. T Newsom Suicide among male veterans: a prospective population-based study J Epidemiol Community Health, July 1, 2007; 61(7): 619 - 624. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 In brief BMJ, November 5, 2005; 331(7524): 1040 - 1040. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||