National Academies Press: OpenBook
« Previous: 4 Major Cohort Studies
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

5
HEALTH OUTCOMES

In this chapter, the committee evaluates the evidence and draws conclusions about long-term health outcomes associated with serving in the Persian Gulf War. The studies reviewed in this chapter generally compare Gulf War veterans with veterans who, during the same period, were either deployed to the Gulf War or were deployed elsewhere. This chapter draws on information from many of the studies that were described in Chapter 4. The committee presents the health outcomes in order of their ICD-10 codes.1 The committee did not examine health outcomes related to or resulting from infectious and parasitic diseases because another IOM committee2 is examining those outcomes; its report will be released in fall 2006.


For each health outcome presented here, the committee first identifies the primary studies and then the secondary studies, as defined by the committee’s criteria (see Chapter 3). Because many cohort studies in this chapter examine multiple outcomes, a study might be referred to in more than one place. The same study might be deemed a primary study for several health outcomes or a primary study for one outcome and a secondary study for another outcome. The key determinant is how well the study’s health outcomes are defined and measured. For example, a study that was well designed for assessing a neurobehavioral effect might not be as well designed for assessing peripheral neuropathy. In general, only primary studies appear in the tables accompanying the discussions of health outcomes.

With rare exceptions, the chapter excludes studies of participants in Gulf War registries established by the Department of Veterans Affairs (VA) and the Department of Defense (DOD). Registry participants are not representative of all Gulf War veterans in that they are self-selected veterans who come to receive care. The VA and DOD registries were not intended to be representative of the entire group of Gulf War veterans.

CANCER (ICD-10 C00-D48)

Over a million people are diagnosed with cancer each year in the United States. About one of every two American men and one of every three American women will have cancer at

1

The International Statistical Classification of Diseases and Related Health Problems (ICD) provides a detailed description of known diseases and injuries. Every disease (or group of related diseases) is given a unique code. It is periodically revised and the tenth edition is known as the ICD-10.

2

The Committee on Gulf War and Health: Infectious Diseases.

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

some point during their lifetime. Cancer can develop at any age, but about 77% of all cancers are diagnosed in people 55 of age and older. Military personnel during the Gulf War were had a mean age of 28 years, thus it is likely too early for the development of most cancers in Gulf War veterans. Cancer strikes Americans of all racial and ethnic groups, and the rate at which new cancers occur (the incidence) varies from group to group (ACS 2006).

The mortality and hospitalization studies reviewed later in this chapter do not definitively identify overall increases in cancer among Gulf War veterans. However, there are other studies that have examined the possibility of brain and testicular cancer. Those cancer studies are evaluated in this section. Because few of the studies dealt with specific cancers, this section groups primary and secondary studies together.

Primary and Secondary Studies

Brain Cancer

Brain cancer is relatively rare. The annual age-adjusted incidence of brain cancer in the United States is 6.4 cases per 100,000 men and women. The median age at diagnosis of brain cancer is 55 years; about 22% of cases are diagnosed between the ages of 20-44 years old (Ries et al. 2005). There is one published study of brain cancer in Gulf War veterans.

Bullman and colleagues assessed cause-specific mortality among the 100,487 Gulf War veterans identified with the 2000 sarin plume model (described in Chapter 2) as having been subjected to nerve-agent exposure from the March 1991 demolition of weapons at Khamisiyah (Bullman et al. 2005). Compared with 224,980 Gulf War veterans similarly deployed but considered unexposed to the plume, there was an increased risk of brain cancer deaths with followup through December 31, 2000 (relative risk [RR] 1.94, 95% confidence interval [CI] 1.12-3.34). There was also a suggestion of a dose-response relationship, with the risk increasing from those who were unexposed to those exposed for 1 day to those exposed for 2 days. Specific subtypes of brain cancer were not considered. Because brain cancer is considered to have a latent period of 10-20 years, and the study included less than 9 years of followup, the results should be interpreted with caution. Further followup is necessary to draw any conclusions about the risk of brain cancer among Gulf War veterans.

Testicular Cancer

Two articles have focused specifically on testicular cancer. Testicular cancer is relatively uncommon in the United States. The annual age-adjusted incidence is 5.3 cases per 100,000 men. However, it is one of the few cancers whose usual age of onset is in the same range as the age of the Gulf War veterans, about 20 to 44 years old (Ries et al. 2005). In general, little is known about environmental risk factors for testicular cancer.

Knoke and colleagues (1998) examined testicular cancer among US servicemen on active duty during the time of the Gulf War (August 8, 1990, through July 31, 1991) and who remained on active duty at the end of the deployment period. Eligible servicemen included 517,223 people deployed to the gulf and 1,291,323 nondeployed. The authors identified cases of all first-hospital admissions, in US military hospitals worldwide, for a principal diagnosis of testicular cancer from the period of July 31, 1991 through April 1, 1996. Cases were identified by examining the DOD hospitalization database through April 1, 1997. A total of 505 cases were ascertained: 134 among the deployed and 371 among the nondeployed. In Cox proportional-hazards models adjusted for race and ethnicity, age, and occupation, no association with deployment status was

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

observed (RR 1.05, 95% CI 0.86-1.29). The deployed did have an increased risk in the early months after the end of the deployment period; it persisted, but did not increase, for about 3 years. The initial increased risk was originally reported in a study of all hospitalizations in the cohort (Gray et al. 1996). However, by the end of the followup period (1996), the cumulative probability of hospitalization of the two groups was the same (0.034% for deployed and 0.035% for nondeployed). There was no interaction between covariates and deployment status. The authors also assessed the association of testicular cancer with specific occupations. The highest RRs were observed for electronic-equipment repair (RR 1.56, 95% CI 1.23-2.00), construction-related trades (RR 1.42, 95% CI 0.93-2.17), and electrical or mechanical repair (RR 1.26, 95% CI 1.01-1.58).

To assess whether the transient increase in risk among the deployed servicemen was related to factors associated with service (for example, the healthy warrior effect), the authors repeated the analysis beginning with followup on January 1, 1990. The rate of hospitalization among the deployed was lower during the prewar and deployment period, increased after the deployment period, and then decreased again. That suggests a healthy warrior effect. The overall relative risk for deployment in this analysis (which used the earlier followup date) was 0.89 (95% CI 0.75-1.06).

In conclusion, Knoke and colleagues did not observe an association of Gulf War service with a risk of testicular cancer (RR 1.05, 95% CI 0.86-1.29) during almost 5 years of followup. The followup period was short for a cancer assessment, but it did include the age-range (22-31 years) when the disease might appear. No specific Gulf War exposures were assessed although risk by occupational group was calculated.

There was some evidence of an association of testicular cancer with Gulf War deployment in a pilot cancer-registry-based study. Levine et al. (2005) matched a cohort of 697,000 Gulf War veterans (all personnel on active duty, in the reserves, and in the National Guard deployed to the Persian Gulf) and 746,248 non-Persian Gulf-region veterans (a stratified random sample of military personnel serving at the time of the conflict but not deployed) with the central cancer registries of New Jersey and the District of Columbia. Between 1991 and 1999, testicular cancer cases were identified in 17 deployed and 11 nondeployed for a crude proportional incidence rate (PIR) of 3.05 (95% CI 1.47-6.35). After adjustment for state of residence, deployment status, race, and age, the PIR was reduced to 2.33 (95% CI 0.95-5.70). The excess in the number of cases peaked 4-5 years after deployment, as opposed to the findings in the Knoke et al. study, where the excess was seen in the first few months after the soldiers returned home. The numbers of cancers included in this study are small, and no definitive conclusions can be made until additional registries are added to the overall study.

All Cancers

Macfarlane and colleagues (2003) assessed all first diagnoses of malignant cancer in a cohort of UK armed-services personnel. The deployed group consisted of all military personnel who served in the Persian Gulf in the period September 1990-June 1991 (n = 51,721). The comparison group was randomly selected from members of the armed services who were in service on January 1, 1991 but not deployed in the Persian Gulf, and was stratified to match the Persian Gulf cohort on age, sex, service branch, rank, and level of fitness for active service (n = 50,755). Followup was from April 1, 1991 until diagnosis of cancer, emigration, death, or July 31, 2002, whichever was earlier. Cancers were identified through the National Health Service Central Register. During followup, 270 incident cases of cancer among the Gulf War veterans

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

and 269 among the nondeployed group were identified; the RR—after adjustment for sex, age group, service branch, and rank—was 0.99 (95% CI 0.83-1.17). Thus, there was no evidence of an association of Gulf War Service with site-specific cancers. In subgroups of cohort members who participated in morbidity surveys that yielded more information on potential risk factors (28,518 Gulf War veterans and 20,829 nondeployed veterans), the RR was 1.11 (95% CI 0.86-1.44). That result did not change after adjustment for smoking or alcohol use, and there was no evidence of associations with exposure to pesticides; multiple vaccinations against anthrax, plague, and pertussis; or reported exposure to depleted uranium. See Table 5.1 for a summary of the cancer studies.

Summary and Conclusion

There is no consistent evidence of a higher overall incidence of cancer in Gulf War veterans than in nondeployed veterans. However, many veterans are young for cancer diagnosis and, for most cancers, the followup period after the Gulf War is probably too short to expect the onset of cancer.

The incidence of and mortality from cancer in general, and brain and testicular cancer in particular, have been assessed in cohort studies. An association of brain-cancer mortality with possible nerve-agent exposure (as modeled by DOD’s exposure model of 2000) was observed in one study (Bullman et al. 2005). As discussed in more detail in Chapter 2, there are many uncertainties in the exposure model. Further followup is warranted to see whether the association with brain cancer holds up with time. Results for testicular cancer were mixed: one study concluded that there was no evidence of an excess risk, and another, a small registry-based study, suggested that there may be an increased risk. Although the results are inconsistent, the committee believes that followup is warranted to see whether such an association exists when more time has passed, as it is still early for the development of most cancers in Gulf War veterans.

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

TABLE 5.1 Cancer Outcomes

Reference

Design

Population

Outcomes

Results

Adjustments

Comments or Limitations

Brain cancer

Bullman et al. 2005 (population from same source as Kang and Bullman 1996; Kang and Bullman 2001)

Cohort mortality study

US GWVs, grouped on basis of exposure to Khamisiyah chemical-munitions destruction (sarin gas) determined by 2000 plume model Exposed (n= 100,487) Unexposed (n=224,980)

Unknown (n =25,574)

Cause-specific mortality ascertained from BIRLS and NDI, followup from date left Gulf through 12/31/2000

Exposed (cases = 25) vs unexposed (cases = 27)

Adj RR 1.94(95% CI 1.12-3.34); GWVs:

Exposed 2+ days: RR 3.26 (95% CI 1.33-7.96)

Exposed 1day: RR 1.72 (95% CI 0.95-3.10)

12.2 deaths/100,000 for each day exposure (95% CI 4.8-19.7)

vs general population:

Exposed 2+ days: SMR 2.13 (95% CI 0.78-4.63)

Unexposed SMR 0.71 (95% CI 0.46-1.04)

[similar results when limited to medical-record confirmed cases]

Age at entry, race, sex, unit component and rank

Latent period too soon (risk increases with time since exposure); multiple comparisons; death certificate diagnosis

Testicular cancer

Levine et al. 2005

Population-based survey—pilot study

US, all personnel (including reserves) deployed to Gulf War (GWVs) and random sample of NDVs;

GWVs (n=697,000)

NDVs (n = 746,248)

Cancers diagnosed 1991-1999 and registered by DC Cancer Registry or NJ State Cancer Registry

GWVs (cases = 17) vs NDVs (cases = 11) (358 males with cancer)

Crude PIR 3.05 (95% CI 1.47-6.35)

Adj PIR 2.33 (95% CI 0.95-5.70)

SIR (compared with SEER) for GWVs 1.42, NDVs 0.94; GWVs peaked 1995-1996, NDVs constant

Age, state of residence, deployment status, race

 

Knoke et al. 1998 (followup of

Cohort study

US, all regular, active-duty male service members

First diagnosis of testicular cancer at US military

GWVs (cases = 134) vs NDVs (cases = 371)

RR 1.05 (95% CI 0.86-1.29)

Race or ethnicity, age, occupation

Short followup time, but right age range;

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

Reference

Design

Population

Outcomes

Results

Adjustments

Comments or Limitations

Gray et al. 1996)

 

GWVs (n = 517,223)

NDVs (n = 1,291,323)

hospitals worldwide 7/31/1991-4/1/1996

 

 

no specific exposures evaluated; military hospitals only

Gray et al. 1996

Cohort study

US, regular, active-duty military personnel deployed in gulf and randomly selected nondeployed personnel

GWVs (n = 579,931)

NDVs (n = 700,000)

Cause-specific hospitalizations in military hospitals 10/1/1988-7/31/1990, 8/1/1991-12/31/1991, 1992, 1/1/1993-9/30/1993

1991 followup period

GWVs (cases = 29) vs NDVs (cases = 14)

RR 2.12 (95% CI 1.11-4.02)

Later followup periods

No excess risk

Age, sex

Active duty personnel only; short followup

All cancers

Macfarlane et al. 2003 (followup of Macfarlane et al. 2000)

Cohort study

UK armed service personnel

GWVs (n = 51,721)

NDVs (n = 50,755);

Subgroup participated in morbidity study with information on risk factors and exposures (GWVs = 28,518, NDVs = 20,829)

First diagnosis of cancer 4/1/1991-7/31/2002 identified through NHSCR

GWVs (cases = 270) vs NDVs (cases = 269)

Main study: RR 0.99 (95% CI 0.83-1.17)

Morbidity study subgroup: RR 1.12 (95% CI 0.86-1.45)

Main analysis:. sex, age group, service branch, rank;

Morbidity study: smoking, alcohol

Short followup; low age; grouped all cancer sites because of small numbers

Kang and Bullman 2001 (also in Gray et al. 1996—followup thru 1993)

Cohort mortality study

US, all military personnel in Gulf before 3/1/1991 and random sample of nondeployed

GWV (n = 621,902)

NDV (n = 746,248)

Cause specific mortality ascertained from BIRLS, death certificates and NDI; followup from date left gulf through 12/31/1997

Males:

GWVs (cases = 477) vs controls (cases = 860): adjusted RR 0.90 (95% CI 0.81-1.01)

Females:

GWVs (cases= 49) vs controls (cases = 103): adjusted RR 1.11 (95% CI 0.78-1.57)

Age, race, branch of service, unit component, marital status

Short latency; low age range; Death certificates

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

NOTE: BIRLS = Beneficiary Identification Records Locator System; GWV = Gulf War veteran; NDI =National Death Index; NDV = nondeployed veteran; NHSCR = National Health Service Central Register; PIR = proportional incidence ratio; SEER = Surveillance Epidemiology and End Results; SIR = standardized incidence ratio; SMR = standardized mortality ratio.

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

MENTAL AND BEHAVIORAL DISORDERS (ICD-10 F00-F99)

War is a known health risk for psychiatric conditions (Pizarro et al. 2006; Wessely 2005). The description of the extent and type of psychiatric affliction and its course has depended on the development of modern psychiatric diagnostic systems and epidemiologic methods. The development of a structured diagnostic system and diagnostic instruments has facilitated the diagnosis of behavioral disorders. Moreover, the prevalence of psychiatric disorders in epidemiologic samples drawn from the general population has become available (Kessler et al. 2005) and provides baseline data with which to compare data from specific inquiries. Thus, after the Persian Gulf War, many methodologic and scientific details were in place to support a sound assessment of the psychologic consequences of war. The Persian Gulf War was highly unusual in that the air war lasted 40 days and the ground war concluded in 5 days, so there was a limited theater and set of conditions amenable in many respects to scientific study. In fact, each of the large cohort studies of Gulf War veterans, described in Chapter 4, included items pertaining to mental health. Nested within them was analysis of mental health characteristics based on direct interview techniques or validated symptom scales.

Types of psychiatric ill health that could be associated with the Gulf War, particularly posttraumatic stress disorder (PTSD), were predicted on the basis of their descriptions from previous wars (O'Toole et al. 1996; Roy-Byrne et al. 2004). As background, psychiatric disorders in the general population are common, as well as disabling and chronic (Department of Health and Human Services 1999). Diagnosable psychiatric disorders are found in about 20% of the US population, but their prevalence in military populations is lower, largely as a result of the healthy-warrior effect. Psychiatric disorders can be grouped into several large classes, for example, mood disorders (that is, depression and bipolar disorder); anxiety disorders (that is, generalized anxiety disorder, obsessive-compulsive disorder, panic disorder, post-traumatic stress disorder, and social phobia); and substance use disorders (for example, abuse of drugs and/or alcohol).

The specification of characteristics of mental diagnoses has made research on their incidence and prevalence possible (Tasman et al. 2003). Depression, a type of mood disorder, is characterized by lifelong vulnerability to episodes of depressed mood and loss of interest and pleasure in daily activities. Some symptoms of clinical depression include sleeping too little or too much, reduced appetite and weight loss or increased appetite and weight gain, restlessness, irritability, difficulty concentrating, feeling guilty, hopeless or worthless, and thoughts of suicide or death. Depression is categorized as major depressive disorder (MDD) or, when it accompanies mania, as bipolar disorder. PTSD is a subtype of anxiety disorder; it occurs after exposure to a traumatic event and is diagnosed when a person manifests severe distress on recollection of the event, avoids the situation, and suffers symptoms of anxiety in daily life. Substance abuse is defined as a maladaptive pattern of substance use (there are many types of abused substances) that results in a failure to fulfill major social roles (such as work or family-care performance), that involves use of the substance despite physical hazards and in association with legal consequences, and that involves use despite deleterious social and interpersonal consequences.

The prevalence of those disorders in the general population is addressed in the US National Comorbidity Survey Replication, a nationally representative face-to-face household

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

survey conducted in the period February 2001-April 2003 (Kessler et al. 2005). The data show that the prevalence estimates for all anxiety disorders were 28.8% (lifetime) and 18.1% (in the last 2 months); for all mood disorders, 20.8% (lifetime) and 9.5% (in the last 12 months); and for substance use disorders, 14.6% (lifetime) and 3.8% (in the last 12 months). Those prevalence estimates are generally higher than those in deployed veterans exposed to war and much higher than in the control veteran populations. It is expected, however, that the military population is healthier than a broad epidemiologic population-based sample, given their age, general health, and military screening.

Primary Studies

In general, each of the large epidemiologic studies of Gulf War veterans’ health included items pertaining to mental health. Moreover, there was often a nested case-control study of mental health characteristics in the primary epidemiologic cohort studies that used direct interview techniques. The studies reviewed here drew their data from the direct-interview studies from the large Gulf War cohort studies reviewed in Chapter 4, and often used validated instruments to complement the interview.

Black et al. (2004b) reanalyzed the population-based, telephone interviews from the Iowa cohort of 4,886 randomly selected veterans (military and reserve), deployed and nondeployed (Iowa Persian Gulf Study Group 1997). The initial cohort study had uncovered higher than anticipated levels of anxiety; therefore, this analysis of the interview data looked more carefully into the features of anxiety in that population. The original cohort was interviewed by telephone using the Primary Care Evaluation of Mental Disorders (PRIME-MD). The Post Traumatic Stress Disorder Checklist-Military (PCL-M) was used for the diagnosis of PTSD (Blanchard et al. 1996; Dobie et al. 2002; Forbes et al. 2001; Weathers and Ford 1996); the CAGE3 was used to estimate alcoholism, and structured questions identified medical conditions and military preparedness. The study identified over a twofold increase in the prevalence of generalized anxiety disorder, panic disorder, PTSD, and any anxiety disorder in veterans deployed to the Gulf War vs the nondeployed. Nearly 6% of deployed veterans met criteria for having any anxiety disorder, compared with nearly 3% of nondeployed veterans (odds ratio [OR] 2.3, 95% CI 1.5-3.5). Participation in combat was an important independent risk factor for the development of anxiety disorders, particularly PTSD (OR 2.1, 95% CI 1.7-4.2). Anxious Gulf War veterans were more likely to have had a pre-existing psychiatric condition, to have taken medications associated with a psychiatric diagnosis, or to have had a previous psychiatric hospitalization. Anxiety conditions occurred with several psychiatric and some medical conditions.

In a separate analysis of the same population based on the telephone survey data, Barrett et al. (2002) used the PCL-M to examine PTSD in the Gulf War veterans. Each of 17 items was rated on a 1-5 severity scale and the scores were summed. A score of 50 or more defined PTSD. Symptoms of physical health were solicited and recognition assessment of medical problems was obtained. PTSD-positive veterans had a mean score of 58.7, whereas those without PTSD had a mean score of 19.7. The PTSD score was significantly associated with the magnitude of other physical illness and negative emotional characteristics.

3

CAGE = acronym for four questions: Have you ever thought you should Cut down on your drinking? Have you ever felt Annoyed by others’ criticism of your drinking? Have you ever felt Guilty about your drinking? Do you have a morning Eye-opener?

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

In a nested case-comparison study, Black et al. (2004a) conducted face-to-face interviews with 602 veterans in 1999-2002. They used the Structured Clinical Interview for DSM-IV (SCID) with a random group of veterans drawn from strata of the PRIME-MD-interviewed group who reported one or more of the following symptom-based conditions during their previous interview: depression (major or minor depression), widespread chronic pain (established criteria for generalized, severe, and chronic pain), and cognitive dysfunction (amnesia or cognitive impairment of a moderate and prolonged intensity). Veterans were stratified by each symptom combination (one, two, or all) and by deployed or non-deployed status. Controls did not have any of those conditions and might have been deployed or not deployed. The veterans were selected randomly for interview from each stratum to optimize the match between cases and controls; interviewers were trained in the use of the diagnostic instruments.

Personality disorders were assessed with the SNAP (Schedule for Nonadaptive and Adaptive Personality). Level of functioning was assessed using the SF-36 (36-Item Short-Form Health Survey). The Whiteley Index was used to determine hypochondriasis. The study found a 32% rate of lifetime depression diagnosis (all types) and that was the same in deployed and nondeployed veterans. There were few diagnostic differences between the depressed deployed and the depressed nondeployed veterans, except for PTSD (27.3% in deployed, 5.0% in non-deployed), anxiety disorders (51.5% deployed and 25.0% non-deployed) and any disorder (68.2% deployed and 51.7% non-deployed). The deployed depressed veterans were more likely to have a diagnosis of any substance-use disorder (69.7% in deployed vs 51.7% in non-deployed). What was most surprising about the direct interview analysis was that there was so little difference between the deployed and the nondeployed veterans in aspects of depression (36.6% vs 30.3%). Most differences were found in anxiety disorder (51.5% vs 25.0%) as noted above.

Kang et al. (2003) conducted a population-based stratified random sample of 15,000 US Gulf War troops compared to a similar sample of troops deployed elsewhere. Phase 1 was a mail survey and phase 2 was a telephone-based survey of PTSD symptoms and chronic fatigue (CFS) symptoms. In the interview cohort, 12.1% of Gulf War veterans and 4.3% of other veterans had symptoms of PTSD, with an adjusted OR of 3.1 (95% CI 2.7-3.4) for PTSD in the Gulf War group; and 6% of the Gulf War veterans and 1.2% of the other veterans (OR 4.8) had CFS symptoms. It was interesting to note that PTSD symptoms showed a monotonic relationship to intensity of war stress, whereas the CFS symptoms did not show any relationship to war stress. The rates of PTSD tracked rates of stressors closely. Deployment, but not war stress, was associated with CFS symptoms.

Wolfe et al. (1999a; 1999b), and Proctor et al. (1998) examined cohorts of veterans randomly sampled and stratified from the Fort Devens and New Orleans-deployed Gulf War veterans, as well as a cohort deployed to Germany (a noncombat area). The Gulf War-deployed veterans from Fort Devens were followed longitudinally from the day of their arrival home from the gulf (time 1) to about 2 years later (time 2) with a 78% participation rate. The Fort Devens cohort was mainly male, Caucasian, and National Guard; rates of PTSD measured at time 1 were 3%. From those cohorts, stratified random samples were selected for closer study with direct interview (220 of the Fort Devens cohort, 73 of the New Orleans cohort, and 48 of the Germany-deployed). The researchers used questionnaires (the 52-item expanded Health Symptoms Checklist [HSC] and the Expanded Combat Exposure Scale), a neuropsychologic test battery, an environmental interview, and psychiatric diagnostic instruments (the Clinician-Administered

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

PTSD Scale [CAPS] or the Mississippi Scale for Combat-Related PTSD) (Proctor et al. 1998). Current PTSD (time 2) was diagnosed in 8.1% of the Fort Devens group, 7.6% of the New Orleans group, and none of the Germany group. Health status and function were lower in the Fort Devens cohort. The three most prevalent symptoms in the Fort Devens group were “forgetfulness,” “fatigue,” and “unsatisfactory sleep.”

Wolfe et al. (1999b) also recruited cases from the Fort Devens and Germany cohorts with a stratified random-sampling strategy (148 from the Fort Devens group, 73 from the New Orleans group, and 48 from the Germany group). They used the Laufer Combat Scale to assess exposure to combat situations and the Mississippi Scale for Combat-Related PTSD to assess PTSD. The deployed Fort Devens group had higher levels of current and lifetime PTSD and current and lifetime MDD than the Germany group; little else regarding psychiatric function was different between the groups. Compared with the PTSD prevalence in the general population (7.8%) (Kessler et al. 1995), the Germany group (controls) had much lower rates of PTSD. However, the low prevalence estimates in the controls increases from zero to 5-8% when the veterans are deployed to active war situations. A strength of this study is that it is characterized by direct interview.

A different study of the cohort examined above (Wolfe et al. 1999a) looked at the course and predictors of PTSD and found that there was a higher rate of PTSD at time 2 (8%) than at time 1 (3%), indicating the development of new cases. Responders at time 2 were more likely to be younger, belong to racial minorities, and be deployed; however, the absence of differences in PTSD rates due to those characteristics indicates a lack of selection bias at time 2. Women were significantly more likely to have PTSD (OR 3.2 at time 1; OR 2.3 at time 2), although their numbers were very low at each assessment.

Brailey et al. (1998) studied Gulf War veterans on their return from service (an average of 9 months after their return) with a face-to-face debriefing and psychologic assessment, comparing Gulf War-deployed (n = 876) with nondeployed veterans (n = 396 mobilized but not deployed), including National Guard and reserve troops. A subset of 349 received a followup assessment an average of 16 months later. They used standard psychiatric rating scales for their assessments including: the Beck Depression Inventory (BDI), the State Anger, State Anxiety, the Brief Symptom Inventory (BSI) Depression, BSI Anxiety, BSI Hostility, and the Health Symptom Checklist (HSC). The deployed veterans had higher scores than the nondeployed on the BDI, the State Anger, the BSI Anxiety, and the HSC. When the Gulf War-deployed veterans were reassessed on average of 16 months later, they showed increases on all scales, including the BDI, the State Anger, the BSI Anxiety, the BSI Hostility, HSC, and on both PTSD scales (the 17-item DSM-III R PTSD Checklist and the Mississippi Scale for Desert Storm War Zone Personnel). They showed increased rates of depression (6.9% to 13.8%), PTSD (2.3% to 10.6%), and hostility (4.9% to 13.8%). The authors correlated war stress with those symptoms and found that the higher the war-zone stress, the more severe the depressive and anxiety symptoms. Troops who were assigned to high-risk activities, such as grave registration, showed a high prevalence of PTSD (46%), depression (25%), and substance abuse (13%).

Goss Gilroy et al. (1998) assessed all 3,113 Canadian Gulf War veterans deployed to the war zone and a comparison group of nondeployed veterans with a mail questionnaire; the methodologic details are in Chapter 4. Using the PCL-M, the investigators found that symptoms of PTSD were 2.5 times more prevalent in the deployed than in the nondeployed veterans (OR 2.69, 95% CI 1.7-4.2). Using the PRIME-MD, the investigators found that the deployed had

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

higher prevalences of major depression (OR 3.67, 95% CI 3.0-4.4), chronic dysphoria, and anxiety. Anxiety and depression were more severe in lower income veterans.

The studies of psychologic outcomes in Australian Gulf War veterans were distinguished by inclusion of the entire deployed population and the use of direct assessments. McKenzie et al. (2004) used the SF-12, the PCL-M, and the GHQ-12 (12-item version of the General Health Questionnaire) to assess 1,424 male Gulf War veterans (86.5% Navy) and 1,548 male Australian Defence Force members who were not deployed to the Gulf War. On those self-rating instruments, the Gulf War-deployed had overall poorer psychologic health and more PTSD-like symptoms than control veterans. The psychologic distress increased with age in the comparison group but decreased with age in the Gulf War veterans (that is, the young Gulf War veterans had the worst psychologic ill-health). Moreover, the perceived level of exposure to war stress was associated with both psychologic ill-health and PTSD-like symptoms, although very few experienced direct combat.

Ikin et al. (2004) conducted a comprehensive health assessment of the same (Australian) cohort, including an interview-administered psychologic health assessment with the Composite International Diagnostic Interview (CIDI), a structured interview of demonstrated reliability and validity. The CIDI data allowed them to make an estimate of pre-Gulf War disorder, post-Gulf War disorder, and current (last 12 months) disorder. Those interview data were used with postal questionnaire data to form a complete workup of 1,381 Gulf War veterans, and 1,377 comparison veterans. Both the veterans and the controls completed both the health assessment and the postal questionnaire. The two groups were demographically similar, although the Gulf War veterans were significantly younger, more likely to have been in the Navy, and less highly ranked than the comparison veterans. The two veteran groups were similar in prevalence of prewar psychiatric disorders. However, the Gulf War veterans were more likely than the comparison group to have developed any disorder after the war (31% vs 21%). The greatest risks were for the anxiety disorders (for example, PTSD: OR 3.9, 95% CI 2.3-6.5), major depression (OR 1.6, 95% CI 1.3-2.0), and alcohol dependence/abuse (OR 1.5, 95% CI 1.2-2.0); the rates of somatic disorders were low in both groups. In addition, the Gulf War group was 2-5 times more likely to have anxiety, PTSD, obsessive-compulsive disorder (OCD), social phobia, or panic, than the comparison group in the preceding 12 months. On average, the Gulf War veterans had twice as many current psychiatric disorders as the comparison veterans. The strengths of this study were the large sample, the comparable control group, the use of well-validated psychologic interviews, and the analyzed participation bias, which was estimated to be low.

A study of DOD postwar hospitalizations for mental disorders (June, 1991-September 30, 1993) using 10 categories from the International Classification of Diseases, 9th Revision, Clinical Modification, 6th Edition (ICD-9-CM) was conducted by Dlugosz and colleagues (1999). It compared all active-duty personnel during the Gulf War era (n = 1,984,996) with those who did not serve. It also sought to identify risk factors for hospitalization. Nearly half the postwar hospitalizations were for alcohol-related disorders. Gulf War veterans were at greater risk for hospitalizations than nondeployed veterans due to drug-related disorders (RR 1.29, 95% CI 1.10-1.52) and acute reactions to stress (RR 1.45, 95% CI 1.08-1.94). Adjustments were made for age, sex, and military service branch. Although the database of ICD-9 codes does not allow determination of whether stress reactions expressly included PTSD, the authors noted that if posttraumatic stress was diagnosed, it would probably have been coded as an unspecified acute reaction to stress (ICD-9 code 308.9). Alcohol-related diagnoses were not increased. Exposure to the ground war in Iraq was associated with a greater risk of alcohol-related hospitalizations in

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

men (RR 1.13, 95% CI 1.04-1.23). Serving as support for the ground war without being in direct combat was associated with a greater risk of drug-related hospitalizations in men (RR 1.42, 95% CI 1.03-1.96) and women (RR 3.61, 95% CI 1.70-7.66). The limitation of this study is that it examined only hospitalizations and thus was not representative of most psychiatric disorders which require outpatient treatment rather than hospitalization. It also did not include veterans who left the military after the Gulf War.

Table 5.2 summarizes the results of the primary studies on psychiatric outcomes.

Secondary Studies

Findings on many other major cohorts of Gulf War veterans support what has been found in primary studies (Gray et al. 2002; McCauley et al. 2002). The most important limitation was their reliance on self-reports of “physician-diagnosed disorders” rather than measurement of symptoms with validated questionnaires or face-to-face interviews. In the UK cohort studied by Unwin et al. (1999), investigators asked some questions taken from the Mississippi Scale for Combat-Related PTSD but did not administer the entire questionnaire. They found that “post-traumatic stress reaction” was about 2-3 times more likely in deployed than in two nondeployed groups. The magnitude of the increase is consistent with that seen in the primary studies. Several other secondary studies have found an association between serving in the Gulf War and psychiatric disorders (Holmes et al. 1998; Magruder et al. 2005; Simmons et al. 2004; Steele 2000; Stretch et al. 1996a; Stretch et al. 1996b; Sutker et al. 1995).

Summary and Conclusion

Two well-designed studies using interview-based assessments have found that several psychiatric disorders, notably PTSD and depression, are 2-3 times more likely in Gulf War-deployed than in nondeployed veterans (Black et al. 2004b; Wolfe et al. 1999b). Direct interviews are labor-intensive; so many other studies administered validated symptom questionnaires. The findings were remarkably similar, that is, an overall increase in magnitude, by a factor of 2-3, of psychiatric disorders. When war exposure was assessed with symptoms, studies characteristically showed higher rates, particularly of PTSD, in veterans who had more traumatic war experiences than in those with lower levels of exposure. In other words, studies found a dose-response relationship between the degree of traumatic war exposure and PTSD. Nevertheless, deployment to a war zone without direct combat exposure is a traumatic war exposure, considering that one well-designed study found deployment without combat to increase the risk of psychiatric disorders by about 60% (Ikin et al. 2004).

Other risk factors were war preparedness, enlisted status (possibly correlated with war preparedness), smoking, and previous psychiatric diagnosis. Two studies indicated that severe PTSD symptoms would worsen over time, and so suggested that careful assessment of the longitudinal course of postwar psychiatric conditions is needed.

It is confirmatory to see that the primary studies, regardless of their techniques of ascertainment or their target population, reported almost identical conclusions regarding the psychiatric outcomes of Gulf War deployment for veterans, that is, that depression, substance abuse or dependence, and anxiety disorders, especially PTSD, were increased in Gulf War veterans after deployment, and that symptom severity was associated with the level of war stress.

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

TABLE 5.2 Psychiatric Disorders

Study

Design

Population

Outcomes

Results

Adjustments

Comments

Black et al. 2004b, Iowa Persian Gulf Study Group 1997

Population-based interview study, by telephone; stratified random sample with proportional allocation

1,896 deployed vs 1,799 nondeployed veterans listing Iowa as home state at time of enlistment

PRIME-MD (major depression, panic disorder, GAD)

PCL-M, combat exposure assessed in basic demographic questionnaire.

CAGE questionnaire (alcohol abuse)

Panic disorder (OR 2.2, 95% CI 1.2-3.8);

GAD (OR 2.5, 95% CI 1.5-4.1); PTSD (OR 2.5, 95% CI 1.2-5.0); any anxiety disorder (OR 2.3, 95% CI 1.5-3.5)

Age, sex, race, branch of military, rank, military status, prior mental-health condition

Large, population-based sample

Barrett et al. 2002

Population-based survey; completed telephone survey about their health status

3,682 Gulf War veterans and control subjects

PCL-M

Gulf War-deployed were twice as likely to screen positive for PTSD than those not deployed; current smoking stautus was associated with PTSD status; PTSD associated with higher number of medical symptoms than those without PTSD and lower levels of functioning and quality of life

 

Brief PTSD screen used; used 50 as the cut-off score with the PCL-M; low number of subjects who screened positive for PTSD; the sample from Iowa might not be representative of all US military personnel

Black et al. 2004a

Nested case-comparison; face-to-face interviews

602 veterans and controls

SCID (face-to-face interviews); SNAP; SF-36; Whiteley Index

PTSD (27% vs 5% in deployed vs controls); anxiety disorders (52% vs 25%); any disorder (68% vs 52%)

Validated PTSD checklist against SCID (70.4% sensitivity and 86.2% specificity of questionnaire)

 

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

Study

Design

Population

Outcomes

Results

Adjustments

Comments

Kang et al. 2003

Cross-sectional; population-based stratified random sample of Gulf War deployed compared with deployed elsewhere

11,441 deployed vs 9,476 nondeployed

Mail survey and telephone-based survey of PTSD symptoms

PTSD (OR 3.1, 95% CI 2.7-3.4)

Sex, age, marital status, rank, and unit component

Nationally representative sample, questionnaire only

Wolfe et al. 1999a

Wolfe et al. 1999b

Proctor et al. 1998

Cross-sectional survey and interviews from larger cohorts followed longitudinally

220 Ft. Devensvs 73 New Orleans vs 48 Germany; New Orleans and Germany cohorts only studied at time 2

Health Symptom Checklist, Mississippi PTSD Scale (times 1 and 2), SCID, CAPS (clinician diagnostic interviews, time 2 only)

Risk factors for PTSD were being female; having high combat exposure; current PTSD, depression, dysthymia; more prevalent in Ft. Devens vs Germany; prevalence of PTSD increased from time 1 (3%) to time 2 (8%) in Ft. Devens

Sex, reported health symptoms

Small sample deployed to Germany

Brailey et al. 1998

Longitudinal; psychologic interviews 9 months after war, and subgroup followup at 16 months; Louisiana National Guard and Reserve troops (Marine, Army, Air Force, Navy)

876 deployed (349 at time 2) vs 396 nondeployed

BDI, State Anger; State Anxiety; the BSI Depression; BSI Anxiety; BSI Hostility, the HSC, PTSD Checklist and the Mississippi Scale

Depression increased over time in deployed veterans from time 1 (6.9%) to time 2 (13.8%); PTSD increased from time 1 (2.3%) to time 2 (10.6%)

Age, education

Large attrition by time 2

Goss Gilroy Inc. 1998

Cross-sectional survey; all deployed

3,113 deployed vs 3,439 nondeployed

PCL-M and the PRIME-MD

PTSD (OR 2.69, 95% CI 1.7-4.2);

major depression (OR 3.67, 95% CI 3.0-4.4);

alcohol abuse (NS)

PTSD: income, service;

Major depression: rank, income

Entire cohort surveyed, actual alcohol findings not reported

McKenzie et al. 2004

Cross sectional survey of all Australian deployed

1,424 male Australian Gulf War veterans vs 1,548 male Australian

SF-12; PCL-M, and GHQ-12

PTSD: OR 2.0, 95% C I1.5-2.9

Service, branch, rank, age category, education, marital

86.5% Navy in Gulf War veterans

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

Study

Design

Population

Outcomes

Results

Adjustments

Comments

 

veterans

Defence Force nondeployed veterans

 

 

status

 

Ikin et al. 2004

Cross sectional survey of all Australian deployed veterans

1,381 GWVs vs 1,377 comparison veterans

CIDI

Prevalence of any

Disorder: 31% in GWVs vs 21% in comparison group;

PTSD: OR 3.9, 95% CI 2.3-6.5);

major depression: OR 1.6, 95% CI 1.3-2.0

Service type, rank, age, education, marital status

GWVs younger, more likely in the Navy, and lower ranked than comparison group

Dlugosz et al. 1999

Post-war hospitalizations June 1991-September 1993

Active-duty men (1,775,236) and women (209,760)

June 1991-September 1993;

Gulf War veterans vs non-Gulf War veterans

ICD-9 CM categories for 10 mental disorders

Gulf War-deployed had increased risk of hospitalizations due to: acute reactions to stress (RR 1.45, 95% CI 1.08-1.94); drug-related disorders (RR 1.29, 95% CI 1.10-1.52)

Age, sex, service-branch adjusted rates; modeling adjusted for age, race, length and branch of service, medical catchment area, prewar mental-health hospitalizations, dates of service in Gulf War, length of deployment in Gulf War, duty occupation, sex, marital status, rank

Active duty only; no assessment of outpatient treatment

NOTE: BDI = Beck Depression Inventory; BSI = Brief Symptom Inventory; CAGE = acronym for four questions: Have you ever thought you should Cut down on your drinking? Have you ever felt Annoyed by others’criticism of your drinking? Have you ever felt Guilty about your drinking? Do you have a morning Eye opener?; CAPS = Clinician Administered PTSD Scale; CIDI = Composite International Diagnositic Interview; ICD = International Classification of Diseases; GAD = generalized anxiety disorder; GHQ-12 =12-Item General Health Questionnaire; HSC = 52-Item Expanded Health Symptoms Checklist; PCL-M = Post Traumatic Stress Disorder Checklist-Military Version; PRIME-MD = Primary Care Evaluation of Mental Disorders-MD; SCID =Structured Clinical Interview for DSM-IV; SNAP =Schedule for Nonadaptive and Adaptive Personality; SF-12 =12-Item Short Form Health Survey; SF-36 = 36-Item Short Form Health Survey.

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

NEUROBEHAVIORAL AND NEUROCOGNITIVE OUTCOMES (ICD-10 F00-F99)

After the Persian Gulf War, veterans often reported memory and concentration difficulty (for example, CDC 1995). The nature of the symptoms, coupled with concern about potential exposure to chemical warfare agents that affect the nervous system, led to studies that used objective neurobehavioral tests. This section focuses on neurobehavioral performance as measured by tests of cognition and, in some cases, sensory integrity or motor speed and coordination. It begins with background on the tests used to evaluate neurocognitive and neurobehavioral capabilities. The confounding factors that can affect performance on these tests also are described. The evaluative portions of the section are divided according to the two major questions addressed by this research in Gulf War veterans: (1) Do veterans who were deployed to the Persian Gulf War differ in neurobehavioral performance tests from veterans who were not so deployed? (2) Do deployed veterans reporting symptoms that meet various case definitions of possible “Gulf War Syndrome” differ in neurobehavioral performance tests from veterans who report having no symptoms?

For the purposes of this section, the committee defines primary studies as high-quality studies that used neurobehavioral tests that had previously been used to detect adverse effects in population-based research on occupational groups. Most secondary studies also used those tests but had methodologic limitations. The secondary studies were reviewed and are included in the discussion because they evaluated the same functional domains, such as attention and memory, and in some cases used the same neurobehavioral tests as did primary studies; they therefore provide valuable supplementary information that helps to increase or decrease confidence in the conclusions drawn from the primary studies. Confidence in a secondary study is substantially reduced if its statistical analysis did not adjust for confounders or if individually administered neurobehavioral tests were given by examiners not blinded to the status of cases and controls. Nevertheless, secondary studies are included in the tables and text whether or not they support the results of primary studies; this departure from most other sections of this report is necessary because of the ill-defined nature of nervous system symptomatology and the need to consider the basis of veterans' most prevalent complaints exhaustively.

Neurobehavioral Tests and Confounding Factors

Neurobehavioral tests have been used for over a century to measure human performance, initially to evaluate performance in schoolchildren (for example, Matarazzo 1972). In the 1960s, neurobehavioral tests came into use to study populations occupationally exposed to neurotoxic agents, and this type of research continues and is still evolving (Lucchini et al. 2005). Neurobehavioral testing is conducted specifically to determine whether cognitive, sensory, or motor performance has been affected by chemical exposure. Most, if not all, of the tests identified in this section have been used for that purpose (Anger 2003), and their demonstrated validity in research increases our confidence in their ability to provide data that will assist researchers in answering questions about illnesses in Gulf War veterans.

The neurobehavioral or neurocognitive tests used in these studies are affected by a variety of factors that can confound their results: age, education, sex, and native intellectual ability (Lezak et al. 2004). Demographic factors (age, education, and sex) are typically obtained

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

through self-reports, but native intellectual ability is difficult to measure in a convincing manner. Three approaches have been used to characterize it: the Armed Forces Qualifying Test (AFQT), which is administered on enlistment in the US military (e.g., Storzbach et al. 2001); the information subscale of the Wechsler Adult Intelligence Scale (WAIS), which measures knowledge attained in life (e.g., Proctor et al. 2003); and the National Adult Reading Test (NART) (e.g., David et al. 2002). The WAIS and NART were administered after the veterans returned from the Persian Gulf War, so they yielded estimates based on current performance; the AFQT is the one truly premilitary measure. Each of those tests correlated with measures of overall intelligence, such as the full WAIS, but is not believed to be affected by exposures, such as to neurotoxic chemicals (that idea has been put forward as one possible cause of symptoms attributed to “Gulf War syndrome”). In addition, it is important that the examiners who administer the neurobehavioral or neurocognitive tests be blind to the condition or status of the veterans and the control population. Blinding is of less concern if the tests are administered on a computer.

Studies That Respond to Question 1 (Outcomes in Gulf War-Deployed Veterans vs Veterans Deployed Elsewhere or Not Deployed)

The committee identified two primary and five secondary studies that compared deployed veterans with those deployed elsewhere or not deployed (Table 5.3). David et al. (2002) compared the neurobehavioral test performance of 209 UK soldiers deployed to the Persian Gulf, 54 UK Bosnia peacekeeping soldiers, and 78 UK Gulf War-Era nondeployed soldiers. A broad array of neurobehavioral tests were administered to all participants and the results were analyzed, although evaluation is limited by the lack of standard deviations of the mean test scores. No differences were found among the groups after correction for age, education, intelligence (according to the NART) and the Beck Depression Inventory (BDI) score (Table 5.3).

Proctor et al. (2003) studied 143 Gulf War veterans and 72 nondeployed veterans of the Danish military. A broad array of neurobehavioral tests were administered to participants and the results were analyzed. Proctor et al. (2003), too, did not find any differences in the overall analysis of neurobehavioral test performance (Table 5.3).

Three secondary studies addressed whether deployed veterans differed from nondeployed veterans (Axelrod and Milner 1997; Vasterling et al. 2003; White et al. 2001). Only one of the three (Axelrod and Milner 1997) found reliable differences in neurobehavioral test performance between the groups after correction for age and education (Table 5.3). The study by White et al. (2001) was considered a secondary study for this outcome because the Gulf War group combined two demographically heterogeneous samples—one from Fort Devens and the other from New Orleans—and because the comparison population, Germany-deployed veterans, was small.

In its evaluation of those studies, the committee was concerned that the investigators’ analyses might have masked differences. For example, David et al. (2002) adjusted the results for depression because it is found to coexist with cognitive measures (e.g., Brown et al. 1994). That adjustment could have made it impossible to detect cognitive differences. David et al. (2002), White et al. (2001) and Vasterling et al. (2003) used an overconservative Bonferroni statistical adjustment to correct for multiple comparisons (Sterne and Davey Smith 2001), which also might have masked differences. The committee, therefore, estimated the effect size (d) (Cohen 1992) of the corrected and precorrection significant test differences (or trends) and searched for a pattern or consistency of results among the studies. The percentage of the neurobehavioral tests given that were reported as significant is listed in column 6; it varied from

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

0% to 30% (mean 21.2%). The tests on which performance deficits in deployed veterans were reported as significant by the report authors are listed in boldface type in column 5 of Table 5.3. Tests that were significant before but not after correction (and thus are not considered significant by the article’s authors) are included in column 5, where they are printed in regular type. The effect size (d) for each test difference is listed if it could be calculated. Although the results are subject to interpretation, a consistent pattern of neurobehavioral deficits does not emerge from an inspection of column 5 of Table 5.3. Specifically, deficits in higher cognitive function were seen in pre-Bonferroni correction analyses in the David et al. (David et al. 2002) and Proctor et al. (2003) studies and not in the secondary studies. Other cognitive measures were deficient in one primary study and two secondary studies (Axelrod and Milner 1997; Proctor et al. 2003; White et al. 2001), but the measures were different. Although deficits in the same or similar measures of motor coordination (pegboard) were seen in three studies, including the David et al. (2002) study, it is difficult to infer that wartime events might have changed coordination. The percentage of precorrection significance varied from 0 to 31% across the five studies (see last column of Table 5.3), and overall, 23% of the tests had at least one significant measure. However, the effect sizes ranged from 0.2 to 0.4 (column 5 of Table 5.3); all of which are small according to Cohen’s criteria (1992).

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

TABLE 5.3 Comparisons of Gulf War-Deployed Veterans with Those Not Deployed or Deployed Elsewhere

Reference

Population

Adjustments

Significant Findings and Statistical Adjustment

Significant Test Deficits

Number of Neurobehavioral Tests and % Significance

Primary Studies

David et al. 2002

209 GWVs vs 54 Bosnia-deployed vs 78 nondeployed

Age, education, intelligence, depression

No; ANCOVA & Bonferroni

Purdue pegboard; Verbal IQ; Performance IQ; block design; logical memory; (No SDs, so no d)

16

(31% sig)

Proctor et al. 2003

143 GWVs vs 72 nondeployed (Denmark)

Age

No; MANCOVA

CVLT (d = 0.3 and 0.32); Wisconsin card sort (d = 0.3 and 0.3)

7

(29% sig)

Secondary Studies

White et al. 2001

293 GWVs vs 50 Germany-deployed

Age, sex, education

No; Bonferroni

Continuous performance test (d = 0.2); PASAT (d =0.3); trailmaking A errors (d = 0.3); Purdue pegboard (d = 0.4)

13

(31% sig)

Vasterling et al. 2003

72 GWVs vs 33 nondeployed

Age, rank

No; Bonferroni

 

7

(0% sig)

Axelrod and Milner 1997

44 GWVs vs standard or adjusted T scores (see Table 5.7 for specifics)

Age (Stroop), education (both tests)

Yes; no adjustment

Stroop word, color, color-word; grooved pegboard

(no standard comparators, so no d)

13

(15% sig)

NOTE: ANCOVA = Analysis of Covariance; CVLT = Continuous Verbal Learning Test; d = effect size; GWV = Gulf War veteran; MANCOVA = Multivariate Analysis of Covariance; PASAT = Paced Auditory Serial Arithmetic Test. Measures considered significant after final statistical analyses are shown in boldface. Measures significant before final statistical comparison (Bonferroni) and thus not considered significant by the authors are listed in regular print in column 5.

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

Studies That Respond to Question 2 (Symptomatic vs Nonsymptomatic Veterans)

Table 5.4 lists the symptoms used by each primary and secondary study to categorize participants as symptomatic. Two primary studies and six secondary studies (Table 5.5) compared symptomatic veterans with those who did not report being symptomatic.

TABLE 5.4 Symptoms Serving as Basis for Categorizing Participants into Symptomatic and Not Symptomatic

Study

David et al. 2002 (Primary Study)

Storzbach et al. 2001 (Primary Study)

Axelrod and Milner 1997

Goldstein et al. 1996

Bunegin et al. 2001

Lange et al. 2001

Hom et al. 1997

Sillanpaa et al. 1997

Basis for Inclusion in Symptomatic Group:

Any One

Any One

Probably Any One

Any One

Chemical Intolerance and Any One

One

Highest Factor Scores On

Factor in Regression

Cognitive deficiencies (unspecified)

 

 

X

 

 

 

 

 

Memory loss

X

X

 

X

X

 

(X)

X

Confusion

 

X

 

 

X

 

(X)

 

Attention or concentration problems

X

X

 

 

X

 

(X)

X

Mood swings

X

X

 

 

 

 

 

 

Somnolence

 

X

 

X

 

 

 

 

Gastrointestinal distress

 

X

 

X

 

 

 

 

Fatigue

X

X

 

X

 

X

(X)

 

Muscle and joint pain

X

X

 

X

 

 

(X)

 

Skin or mucous membrane lesions

 

X

 

X

 

 

 

 

Chemical intolerance

 

 

 

 

X

 

 

 

Headache, sensory dysfunction, dizziness, depression, tumor, weight gain, sinus problems, or newborn problems

 

 

 

X

 

 

 

 

Main factors: impaired cognition (attention, memory, reasoning); confusion-ataxia (disturbances in thinking, orientation, balance, impotence), arthromyoneuropathy (joint and muscle pain, muscle fatigue, lifting difficulty, extremity paresthesia)

 

 

 

 

 

 

X

 

NOTE: (X) = symptom listed in publication as example of group of symptoms that was basis for categorization, but report did not list full range of specific symptoms.

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

Table 5.5 below reveals that each case definition is slightly different. However, the core symptoms are cognitive and are the same or would be assumed to be the same (Axelrod and Milner 1997; Hom et al. 1997) in all but the Lange et al. study (2001) in which fatigue was the exclusive symptom. The results are summarized in Table 5.5.

Of the two primary studies, David et al. (2002) categorized the veterans as either ill (n = 151) or well (n = 188) on the basis of their SF-36 physical functioning scale scores. They then recategorized them more stringently on the basis of the Center for Disease Control and Prevention working definition of Fukuda et al.(1998) for reporting (n = 65) or not reporting (n = 33) Gulf War-related symptoms (Table 5.5, column 1), analyzing each separately. (These are additional analyses from the same David et al. (2002) study identified above as comparing veterans on the basis of their deployment status.) David et al. (2002) did not find overall differences after adjusting for age, education, estimated intelligence (NART), and depression (BDI score); but they did report some cognitive test differences before making the final Bonferroni correction for multiple comparisons (column 5 of Table 5.5).

Storzbach et al. (2001) compared 239 deployed veterans who reported having at least one symptom associated with the “Gulf War syndrome” (Table 5.5) with 112 nonsymptomatic deployed veterans in a case-control study. They reported poorer performance on three neurobehavioral tests (Oregon Dual Task Procedure [ODTP] errors and latency, Digit Span Backward, and Simple Reaction Time) after adjustment for age, education, AFQT scores, and a Bonferroni correction for multiple comparisons (Table 5.5). In the analysis, Storzbach et al. identified a group of “slow” responders (12% of the symptomatic group) who had very slow latencies in choosing answers on the memory component of the ODTP, although they made few errors. Anger et al. (1999) reported on the first 100 participants in whom this slow group was identified; see Table 5.5. The slow ODTP subgroup proved to have been responsible for the statistically significant differences between cases and controls in the larger group: removing them from the analysis virtually eliminated the differences. The effect size between the slow responders and the controls on the ODTP was very high (d = 2.9). The effect sizes for the performance differences of the slow ODTP group (n = 30) vs the non-symptomatic (n = 112) ranged from 0.7 to 2.9 (mean d = 1.5) (Table 5.5, column 5).

All other studies that compared symptomatic veterans with nonsymptomatic veterans are listed as secondary studies. They are outlined in the lower six rows of Table 5.5 and are described in more detail in Table 5.7. As can be seen in Table 5.5, some categorized veterans as symptomatic on the basis of subsets of the overall symptom constellation, and some defined symptomatic in a narrower way. Some of the studies were not well described, had design flaws, asked limited questions, and studied small samples (n = 8-48). Nonetheless, as noted above, they used similar or the same neurobehavioral tests and thereby allow an evaluation of the consistency or pattern of results.

Five of the six secondary studies reported performance differences between symptomatic cases and controls or between cases and standard scores. As can be seen in Table 5.5, only two secondary studies (Axelrod and Milner 1997; Sillanpaa et al. 1997) adjusted for age and education, and only one (Lange et al. 2001) adjusted for multiple comparisons. The study that did not report a difference (Sillanpaa et al. 1997) did adjust for age, education, and an estimate of exposure (not clearly described). Each study had additional weaknesses. Axelrod and Milner (1997) used standard scores instead of controls and did not mention whether examiners were blinded. The basis of subject selection was not clear in Goldstein et al. (1996), and the issue of blinding of examiners was not addressed in the publication. The basis of subject selection was

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

not clear in Bunegin et al. (2001), nor was examiner blinding addressed, although this is not a serious problem for the computer-based tests used in this study. In Hom et al. (1997), the basis of selection of nonsymptomatic controls was not described, that is, whether it was random or targeted. Lange et al. (2001) studied veterans with fatiguing illness but did not describe the case definition, and blinding of the examiners was not addressed although individually administered tests were used. Sillanpaa et al. (1997) used standard scores instead of a control group. Notwithstanding those limitations in the secondary studies, they found effects consistent with those found in the primary studies.

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

TABLE 5.5 Comparisons of Symptomatic and Nonsymptomatic Veterans

Reference

Population

Adjustment

Significant and Statistical Adjustment

Significant Test Deficits (Effect Size if Calculable)

Number of Neurologic Tests

Primary Studies

David et al. 2002

151 symptomatic vs 188 nonsymptomatic

Age, education, intelligence, depression

No; ANCOVA, Bonferroni

Performance IQ, digit symbol; trail- making B, trail-making A, sustained attention reaction time (No d; no SDs in Table 3 of study)

16

(25% sig)

David et al. 2002

65 CDC cases (per Fukuda 1998 definition) vs 33 well

Age, education, intelligence

No; ANCOVA, Bonferroni

WAIS vocabulary (d = 0.6), digit symbol (d = 0.8), PASAT [near 0.05] (d = 0.7), Stroop [near 0.05] (d = 0.8)

16

(25% sig)

Storzbach et al. 2001

239 GWVs symptomatic vs 112 nonsymptomatic; analyses of 30 “slow ODTP” vs 112 nonsymptomatic

Age, education, intelligence

Yes; ANOVA, regression, Bonferroni

Symbol Digit (d = 1.0), ODTP errors (d = 1.1) and latency (d = 2.9); Selective Attention (d = 0.7); digit span forward (d = 0.9); backward (d = 2. 5); simple reaction time (d = 1. 3)

6

(83% sig)

Secondary Studies

Axelrod and Milner 1997

17 GWVs symptomatic vs 27 non symptomatic

Age, education

Yes; t-test; no adjustment

Semantic fluency (d = 0.7), Stroop word (d = 0.8), color (d = 0.8), color- word (d = 0.7)

13

(15% sig)

Goldstein et al. 1996

21 GWVs with ½ symptomatic vs 38 locals

 

Yes; MANOVA; no adjustment

Impairment index (d = 0.7) (individual < 0.05) COWAT (d = 0.6); continuous performance/trial 3 (d = 0.6)

11

(18% sig)

Bunegin et al. 2001

8 GWVs with symptoms vs 8 without symptoms

 

Yes; ANOVA; no adjustment

Pattern memory (d = 0.9), digit span forward (d = 0.8) and backward (d = 1.0); switching attention (d = 0.7)

6

(50% sig)

Hom et al. 1997

26 GWVs symptomatic vs 20 non-deployed or nonsymptomatic

 

Yes; l-tail t-test and Mann-Whitney U.; no adjustment

Halstead category test (d = 0.7); WAIS comprehension (d = 0.6); WAIS vocabulary (d = 0.6); bilateral simultaneous tactile-left (d = 0.6); finger recognition-right (d = 0.5) and left-right confusion (d = 0.8); trail-making B (d = 0.7); grip strength dominant (d = 1.0) and nondominant (d = 0.7)

22

(36% sig)

Lange et al. 2001

48 GWVs with fatigue vs 39 without

 

Yes; MANOVA by domain, Bonferroni

Category test (d = 0.5); trail-making B-A (d = 0.8); complex (d = 0.6) and simple (d = 0. 8) reaction time; digit span backward (d = 0.5); PASAT (d = 0.6)

10

(60% sig)

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

Reference

Population

Adjustment

Significant and Statistical Adjustment

Significant Test Deficits (Effect Size if Calculable)

Number of Neurologic Tests

Sillanpaa et al. 1997

16 with memory and attitude symptoms vs 13 with memory or attention symptoms vs 19 nonsymptomatic

Age, education, exposure estimates

No; MANOVA; no adjustment

 

7

(WAIS = 1)

(0% sig)

NOTE: ANCOVA = Analysis of Covariance; CDC = Centers for Disease Control and Prevention; COWAT = Controlled Oral Word Association Test; GWV = Gulf War veteran; MANOVA = Multivariate Analysis of Variance; ODTP= Oregon Dual TaskProcedure; PASAT = Paced Auditory Serial Addition Test; sig = significant; WAIS =Wechsler Adult Intelligence Scale. Measures the authors reported assignificant after all statistical comparisons are in boldface. Measures that are not considered significant (column 5) after final statistical adjustments are in regular print.

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

Taking the primary and secondary studies together (see Table 5.5), differences or trends were seen in numerous measures in at least six and probably seven studies, but not in Sillanpaa et al. (1997). The percentage of neurobehavioral tests that were reported as significant out of the total number of tests administered is listed in column 6 of Table 5.5; the percentage is 0% to 83% (mean = 35.9%). The mean effect size of all the measures in the table is 0.86, which is “large” according to Cohen (1992). Thus, evidence of a consistent and large neurocognitive deficit in returning symptomatic veterans emerges from the disparate studies. That one study (Storzbach et al. 2001) found that a subgroup defined the extreme of the distribution of test scores suggests that a unique experience or exposure may have been responsible for the differences. Hom et al. (1997) may have been studying a similar subgroup in their factor-analytic study of “Gulf War syndrome” in a larger group from which the subgroup was selected.

Related Findings: Malingering and Association of Symptoms with Objective Test Results

Several studies provide ancillary information related to the two main questions. Results of a standard test of malingering (TOMM—Test of Memory Malingering) suggested that veterans were not attempting to fake poor memory performance in the study of White et al. (2001). In fact, all veterans earned a perfect or nearly perfect score on the TOMM (low scores are associated with malingering) (Lindem et al. 2003c). The computer-based ODTP used by Storzbach et al. (2001) was adapted from a standard neuropsychologic test of malingering. The component of the ODTP that was designed to detect poor motivation or malingering (memory errors) led to the exclusion of only two of 351 participants. That suggests that the veterans were not faking poor performance on the tests and supports the validity of the neurobehavioral test results.

Neurobehavioral performance scores were correlated only weakly with self-reported symptoms of poor memory drawn from the SCL-90 test (Binder et al. 1999) or self-reported symptoms of poor overall neuropsychologic performance (Lindem et al. 2003b). Thus, subjective symptoms were not a good predictor of objective test performance. Lindem and colleagues (Lindem et al. 2003a; Lindem et al. 2003b) reported associations between neurobehavioral performance and mood (Lindem et al. 2003b), PTSD symptom severity (Lindem et al. 2003a), and anger and depression measures (Hull et al. 2003).

Those findings suggest that neurobehavioral monitoring or screening tests should be administered to all veterans, not only those who report symptoms of concentration and memory loss, because subjective reports are not accurate predictors of objective lower performance (Binder et al. 1999; Lindem et al. 2003b). Tests that should be considered are those listed multiple times in the tables, particularly tests with the highest effect size (such as PASAT, Stroop, Digit Span, and Trail-making). Having predeployment data on the tests would make them far more useful for individual assessments, and tests that did not reveal differences in veterans should also be included to provide a control for broad-based low performance or intentionally poor individual performance (Storzbach et al. 2001).

Summary and Conclusion

Primary studies found nonsignificant trends of poorer neurobehavioral performance when Gulf War veterans were compared with nondeployed veterans or those deployed to Germany. However, when PTSD (White et al. 2001) or depressed mood (David et al. 2002) was treated as a confounder in the statistical analyses those trends disappeared. The results were adjusted for

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

depression because it is often found to coexist with PTSD. That adjustment could have made it impossible to detect cognitive differences.

One study concluded that Gulf War veterans who report symptoms associated with the Gulf conflict performed more poorly on neurobehavioral tests than veterans who did not report symptoms (Storzbach et al. 2001); another study found substantial neurobehavioral deficits in deployed veterans but had intentionally recruited veterans who experienced a high prevalence of post-Gulf War illness (Hom et al. 1997). That study, however, failed to adjust for key confounders and for the large number of statistical comparisons in their study, raising doubt about the validity of their findings.

With regard to malingering, the component of the ODTP that was designed to detect poor motivation or malingering (memory errors) led to the exclusion of only two of 351 participants (Storzbach et al. 2001). That suggests that the veterans were not faking poor performance on the tests and supports the validity of the neurobehavioral test results.

In conclusion, primary studies of deployed Gulf War veterans vs veterans not deployed to the Gulf do not demonstrate differences in cognitive and motor measures as determined through neurobehavioral testing. However, returning Gulf War veterans who had at least one symptom commonly reported by Gulf War veterans (fatigue, memory loss, confusion, inability to concentrate, mood swings, somnolence, gastrointestinal distress, muscle or joint pain, or skin or mucous membrane complaints) demonstrated poorer performance on cognitive tests than returning veterans who did not report such symptoms. Results of primary and secondary studies are summarized in Tables 5.6 and 5.7.

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

TABLE 5.6 Neurobehavioral or Neurocognitive Outcomes: Primary Studies

Reference

Population

Outcome Measures

Results

Statistics and Adjustment

Comments or Limitations

David et al. 2002 (derivative of Unwin et al. 1999)

Design: case-control

341 randomly selected male deployed; Bosnia peacekeeping; Gulf War-Era non-deployed UK soldiers who were either ill or well (from the first decile of the SF-36 physical functioning subscale)

WAIS-R scaled scores:

Vocabulary

Digit span

Arithmetic

Similarities

Picture arrangement

Block design

Object assembly

Digit symbol

PASAT

Sustained attention to response task

Stroop

Trailmaking A and B

WMS:

Logical memory (Immediate and delayed recall)

Verbal paired associates (Immediate and delayed recall)

Camden recognition memory test

Purdue pegboard

Individually administered tests,

blinded examiners

Gulf War-deployed had significantly lower scores on 5 cognitive tests after demographic confounder and LSD corrections: Digit symbol Trail-making Stroop PASAT Verbal associates Final Bonferroni adjustments for multiple comparisons and BDI revealed the only difference was in the Purdue pegboard

ANCOVA, adjusted for education, age, NART, BDI Multiple comparison adjustment: LSD procedure and Bonferroni

Careful treatment of potential confounders such us depression mood, IQ and education. However, it is difficult to determine the total number (n) in the original Unwin et al. (1999) study or this one. Also, the definition of the controls in this study is confusing

Proctor et al. 2003

(derivative of Ishoy et al. 1999b)

Design: cross-sectional

143 male Gulf War-deployed Danish military veterans vs 72 male non-deployed military forces randomly selected from 84% and 58% of total number of Danish armed forces deployed and nondeployed, respectively, at the time of the Gulf War

WAIS-R scaled scores:

Block design

Information

CVLT

WMS

Visual reproductions

NES continuous-performance test

Trail-making

Wisconsin card sorting test

Purdue pegboard

TOMM

Individually administered tests except test in computer-based

Authors’ conclusion: no overall differences in neuropsychologic domains, Significant tests within domains: CVLT Wisconsin Card Sort

MANCOVA by neuropsychologic domain, adjusted for age

Rationale of initial mailing is unclear (see Ishoy et al. 1999, in Chapter 4)

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

Reference

Population

Outcome Measures

Results

Statistics and Adjustment

Comments or Limitations

 

 

neurobehavioral system (NES); blinded examiners

 

 

 

Storzbach et al. 2001 (derivative of McCauley et al. 1999)

Design: case-control

239 randomly selected Gulf War male and female deployed veterans with symptoms vs 112 deployed with no symptoms; case = one of memory loss, confusion, inability to concentrate, mood swings, somnolence, gastrointestinal distress, fatigue, muscle and joint pain, skin or mucous membrane lesions lasting 1 month or longer, starting during or after service in gulf, and present during 3 months before questionnaire received

Symbol digit

Serial digit learning

ODTP

Selective attention test

Digit span

Simple reaction time

BARS computer-based testing system

Blinded examiners

Cases significantly worse than controls on:

Digit span backward

Simple reaction time

ODTP

Errors

Latency (including a slow group of 13% of sample with scores > 2 SD slower than control mean latency)

PCA showed the slow ODTP (Slow Case in 1999) were responsible for group differences in neurobehavioral performance; 2 of 354 excluded for possible poor motivation because of excess errors in ODTP

ANCOVA, adjusted for age, sex and AFQT, but effect was small so t-tests were used; Bonferroni correction for multiple comparisons

 

NOTE: AFQT = Armed Forces Qualifying Test; ANCOVA = analysis of covariance; BARS = Behavioral Assessment and Research System; BDI = Beck Depression Inventory; CVLT = California Verbal Learning Test; LSD = least significant difference; MANCOVA = multivariate analysis of covariance; NART = National Adult Reading Test; NES = Neurobehavioral Evaluation System; ODTP = Oregon Dual Task Procedure; PASAT = Paced Auditory Serial Addition Test; PCA = principal components analysis; TOMM = Test of Memory Malingering; WAIS = Wechsler Adult Intelligence Scale; WMS = Wechsler Memory Scale.

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

TABLE 5.7 Neurobehavioral or Neurocognitive Outcomes: Secondary Studies

Reference

Population

Outcome Measures

Results

Statistics and Adjustment

Comments and Limitations

White et al. 2001 (derivative of Proctor et al. 1998)

Design: cross-sectional

343 veterans (293 randomly selected deployed who returned to Massachusetts and Louisiana, and 50 from entire National Guard reserves in Maine sent to Germany during the Gulf War)

WAIS-R scores:

Information

Block design

WMS-R:

Digit span

Paired associate learning

Visual reproduction

CVLT

NES CPT

Trail-making A and B

Wisconsin card-sorting test

PASAT

Finger tapping

Purdue pegboard

TOMM

Individually administered tests except test in computer-based NES

Blinding not addressed

No differences between the domains; trends of deficits in Gulf War-deployed groups seen on continuous performance, PASAT, Wisconsin Card Sorting Test, trail-making A errors, Purdue pegboard; no evidence of malingering on TOMM

Multivariate regression controlling for age, sex, education, sampling design; Bonferroni-adjusted differences within neuropsychologic domains

Small sample of 50 controls from Maine reserve unit; combined demographically heterogeneous groups to form deployed group

Vasterling et al. 2003 (derivative of Brailey et al. 1998)

Design: cross-sectional

72 randomly selected male and female south Louisiana National Guard and military reserve unit deployed vs 33 male and female activated but not deployed veterans (estimated 33% response rate) from larger Brailey et al. study

UPSIT

Digit span

Rey Auditory Verbal Learning Test

Continuous Visual Memory Test

Wisconsin card-sorting test

Purdue pegboard

WAIS-R information

Individually administered tests

Examiner blinding not addressed

No differences on neurobehavioral tests

ANCOVA with age and rank as co-covariates; Bonferroni correction for multiple comparisons within neuropsychologic domains

Small sample, large difference in sample size between control and deployed group, lack of adjustment for multiple comparisons

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

Reference

Population

Outcome Measures

Results

Statistics and Adjustment

Comments and Limitations

Axelrod and Milner 1997

Design: case vs standard scores

44 male deployed veterans of Michigan Army National Guard reserve unit vs standard or T scores; subjects were volunteers from 78 veterans in unit; data grouped by (1) self-report of cognitive symptoms or (2) none since Gulf War

Rey Auditory Verbal Learning TestR

Reitan-Indiana Aphasia ScreenAE

WAIS-R (score)S

Controlled Oral Word Association TestR

Semantic (Category) FluencyR

Peabody Individual Achievement TestR, S

Wisconsin card-sort testAE

Stroop Color and Word TestA

Trail-making testAE

WMSR AE

Finger tapping testAE

Grooved pegboardAE

Grip strengthAE

AAge-adjusted T scores

AE Age- and education- adjusted T scores

R Raw scores

S Standard scores

Deployed veterans had deficits on StroopA, Grooved PegboardAE ; symptomatic veterans had deficits on Stroop word, color & color-word namingA, semantic fluencyR

t-test; no adjustment for demographic confounders, but analysis indicated groups did not differ on these issues; no adjustment for multiple comparisons; individual tests were compared with scores according to legend in column 3

Small sample; lack of group differences on demographic confounders does not eliminate need for adjustments; no controls; standard scores do not adjust for occupational group characteristics or local differences, such as in education

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

Reference

Population

Outcome Measures

Results

Statistics and Adjustment

Comments and Limitations

Goldstein et al. 1996

Design: cross-sectional

21 male and female deployed with health-related complaint vs 38 male and female demographically matched community controls (part of study of chemical exposure); basis for selection (such as, random) not addressed

WAIS-R:

Information

Similarities

Digit span

Digit symbol

Picture completion

Block design

Verbal associative learning task

Incidental memory task

Symbol-digit learning task

Trail-making

Grooved pegboard

CPT

Brown-Peterson distractor-type memory task

Individually administered tests, examiner blinding not addressed

Impairment index comparison revealed significant deficits in deployed with health-related complaints vs controls (no complaints), but SCL-90-R covariates attenuated other differences

MANOVA; no adjustments for confounders or multiple comparisons

Small sample; selection basis for participants not clear

Bunegin et al. 2001

Design: case-control

Eight deployed male veterans with one or more neurobehavioral symptoms and eight asymptomatic male controls who had no negative responses to odors; participants recruited from newspaper ads and from San Antonio veterans hospital

While exposed to acetone or not: Horizontal addition Pattern memory Switching attention Digit span NES computer-based tests Examiner blinding not addressed

Symptomatic veterans had less correct responses on digit span, pattern memory, switching attention (side direction side) than controls; (acetone exposure had no effect on performance)

RMANOVA; no differences in groups on age, but no adjustment for age or other confounders or multiple comparisons

Very small sample; selection basis for participants not clear; community participants were respondents to newspaper ads

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

Reference

Population

Outcome Measures

Results

Statistics and Adjustment

Comments and Limitations

Hom et al. 1997 (derivative of Haley et al. 1997b)

Design: case-control

26 male deployed veterans with highest attention, memory, reasoning “syndrome” scores (syndrome based on (Haley et al. 1997b) factor analysis of 249 Naval Mobile Construction Batallion veterans) vs 10 deployed with no health problems plus 10 nondeployed from same battalion (the 20 controls were selected to be demographically similar to “syndrome” group in age and education and were treated as single group for analysis)

WAIS-R:

Information

Comprehension

Digit span

Arithmetic

Similarities

Vocabulary

Picture arrangement

Picture completion

Block design

Object assembly

Halstead category test

Tactual performance tests A & B

Speech-sounds perception test

Finger Oscillation test

Trail-making A & B

Reitan-Indiana aphasia screening examination

Reitan-Klove sensory-perception and lateral dominance examinations

Reitan word-finding test

WMS-Russell Revision

Verbal and Figural immediate and delayed recall

Wide range achievement test 3

Individually administered tests

Blinded examiners

One-tail t tests showed deficits in deployed in Halstead category test, trailmaking B, grip measures, WAIS comprehension, WAIS vocabulary bilateral, simultaneous tactile-left, finger recognition-right, left-right confusion

One-tail t tests for neurobehavioral tests; two-tail tests for psychologic tests; no adjustment for demographic confounders or multiple comparisons

Small sample; many neurobehavioral measures without statistical adjustment for multiple comparisons; used one-tail tests for those comparisons

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

Reference

Population

Outcome Measures

Results

Statistics and Adjustment

Comments and Limitations

Lange et al. 2001

Design: case-control, secondary study

48 deployed veterans who reported fatiguing illness vs 39 veterans who did not

NES Simple Reaction Time

NES Complex Reaction Time

PASAT

WAIS-R:

Digit span

Block design

CVLT

Rey-Osterrieth complex figure test

Trail-making A & B

Judgment of line orientation test

Grooved pegboard test

Individually administered tests except tests in computer-based NES; blinding not addressed

Significant deficits in deployed fatigued on simple and complex reaction time, digit span backward, PASAT, trail-making B, category test; followed by multiple regression fitted with psychopathology variables MDD, anxiety disorders, PTSD as independent variables and cognitive measures as dependent variables, revealing that MDD predicted simple and complex reaction-time performance, and MDD and anxiety disorders predicted PASAT performance

MANOVA followed by t tests adjusted for multiple comparisons by Bonferroni

Source of subjects unclear (may not be randomly selected); usual limitation of a small sample, but statistical analysis is sound, research questions are clearly spelled out

Sillanpaa et al. 1997

Design: case vs standard scores (rather than a control group), secondary study

82 deployed Army Reserve Military Police referred for neuropsychologic analyses vs normative data; somatic complaints on MMPI, SCL-90, and neurologic symptom checklist were basis for defining participants as symptomatic or not

WAIS-R-full scale IQ

Rey auditory verbal learning test

Wisconsin card-sorting test

NES CPT

Grip strength

Grooved pegboard

Vibratory sense-neurologic screen

Fingertip number writing perception

Individually administered tests, except tests in computer-based

NES, blinded examiners

No differences

MANOVA; no confounder adjustments, although all explored and education was significant; no adjustment for multiple comparisons

No control group; standard scores do not adjust for occupational group characteristics or local differences, such as in education

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

Reference

Population

Outcome Measures

Results

Statistics and Adjustment

Comments and Limitations

Lindem et al. 2003c (derivative of Proctor et al. 1998)

Design: cross-sectional, secondary study

77 veterans (58 randomly selected deployed from Louisiana, and 19 from National Guard reserves in Maine deployed to Germany during Gulf War) divided into those with high and low scores on TOMM; their neurobehavioral test performance was compared

WAIS-R full scale IQ

Rey Auditory Verbal Learning TestWisconsin card-sorting test

NES CPT

Grip strength

Grooved pegboard

Vibratory sense (“neurologic screen”)

Fingertip number-writing perception

Individually administered tests except tests in computer-based

NES; blinded examiners

21 veterans scoring high on TOMM had significantly lower scores on some neuropsychologic tests compared to 56 scoring lower; overall, most veterans’ scores indicated they were well motivated

MANCOVA by domain; also adjusted for multiple comparisons by requiring significant univariate and grouping variables; adjusted for age and education

Small sample limits conclusions

Binder et al. 1999 (derivative of McCauley et al. 1999)

Design: case-control, secondary study

100 randomly selected northwest US male and female deployed veterans grouped as symptomatic (cases) vs asymptomatic (controls) (first 100 of sample from Storzbach et al. 2000)

Correlated self-ratings of cognition (8 SCL-90 R items indicative of cognitive complaints) and affective stress (Beck Depression and Anxiety Inventories) with neurobehavioral performance: Symbol digit

ODTP

Selective attention task

Digit span

Simple reaction time

BARS computer-based testing system; blinded examiners

Correlations between neurobehavioral performance and subjective cognitive complaints on SCL-90 R were 0.28 or less; affective complaints correlated 0.58 with subjective cognitive complaints; thus, self-ratings of cognitive deficits are poorly correlated with objective measures of cognitive performance

R, plus q statistic that showed cognitive symptoms (SCL-90 R) had a stronger relationship with affective distress (Beck scales) than with cognitive test performance

This suggests that epidemiologic studies that use self-reported symptoms are hard to interpret and at least must adjust for affective distress (such as depression)

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

Reference

Population

Outcome Measures

Results

Statistics and Adjustment

Comments and Limitations

Lindem et al. 2003b (derivative of Proctor et al. 1998)

Design: case-control analysis of cross-sectional study, secondary study

240 veterans (193 randomly selected deployed who returned to Massachusetts and Louisiana and 47 from entire National Guard reserves in Maine sent to Germany during Gulf War; veterans were grouped by degree of symptom reporting into high, medium, and low on HSC

WAIS-R:

Information

Block design

WMS-R:

Digit span

Paired Associate Learning

Visual reproduction

CVLT

NES CPT

Trail-making A & B

Wisconsin card-sorting test

PASAT

Finger tapping

Purdue pegboard

TOMM

Individually administered tests except test in computer-based NES; blinding not addressed

Subjective complaint (on HSC) not associated with objective neurobehavioral performance deficits; veterans deployed to gulf reported more severe neuropsychologic symptoms than those deployed to Germany

MANCOVA by domain; also adjusted for multiple comparisons by requiring significant univariate and grouping variables; adjusted for age, education, WAIS-R information, deployment status, medical conditions, PTSD severity, depression

Sound statistical analysis

Lindem et al. 2003a (derivative of Proctor et al. 1998)

Design: cross-sectional, secondary study

225 veterans (178 randomly selected deployed who returned to Massachusetts and Louisiana and 47 from entire National Guard reserves in Maine sent to Germany during Gulf War); PTSD symptom severity (based on CAPS) was related to objective measures of neuropsychologic performance

WAIS-R:

Information

Block design

WMS-R:

Digit span

Paired Associate Learning

Visual reproduction

CVLT

NES CPT

Trail-making A & B

Wisconsin card-sorting test

PASAT

Finger tapping

Purdue pegboard

TOMM

Individually administered tests except test in computer-based NES; blinding not addressed

PTSD symptom severity (based on CAPS) was correlated with neuropsychologic performance on: WAIS-R information, CPT, tapping, Purdue pegboard, CVLT

Partial correlational analyses controlled for age, education, WAIS-R information, deployment status, depression as measured by the Structured Clinical Interview for DSM-III-R (SCID); R2 measures obviate multiple comparison adjustment

 

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

Reference

Population

Outcome Measures

Results

Statistics and Adjustment

Comments and Limitations

Hull et al. 2003 (derivative of Unwin et al. 1999)

Design: case-control, secondary study

136 males from three groups: deployed UK soldiers, nondeployed UK soldiers, UK soldiers deployed to Bosnia–divided into symptomatic cases (from lowest decile of SF-36 physical functioning subscale) and nonsymptomatic cases

WAIS-R – scaled scores:

Vocabulary

Digit span

Arithmetic

Similarities

Picture arrangement

Block design

Object sssembly

Digit symbol

WMS-R:

Logical memory--Immediate

and delayed recall

Verbal Paired Associates--

Immediate and delayed recall

Letter-number sequencing task

PASAT

Sustained Attention to Response

Task

Stroop

Trail-making A & B

Camden Recognition Memory Tests

Purdue pegboard

Individually administered tests; blinded examiners

Moderate (0.41) relationships between anger and depression with neuropsychologic measures

Structural equation modeling

Small sample for structural equation modeling

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

Reference

Population

Outcome Measures

Results

Statistics and Adjustment

Comments and Limitations

Anger et al. 1999 (derivative of McCauley et al. 1999) (also see Storzbach et al. 2001 in primary studies, Table 5.6 above)

Design: case-control, secondary study

101 randomly selected northwest deployed male and female veterans grouped (first 101 scheduled for study) as symptomatic (cases, n = 66) vs asymptomatic (controls, n = 35); cases: symptoms lasting 1 month or longer, began during or after service in gulf, present during 3 months before questionnaire received

Symbol digit

Serial digit learning

ODTP

Selective Attention Test

Digit span

Simple reaction time

BARS computer-based testing system; blinded examiners

Cases significantly worse on ODTP after Bonferroni correction, including “slow group” of 13% of the sample with very low scores; cases inferior on all other neurobehavioral tests, and three were significant before Bonferroni correction; performance test results and ODTP performance suggested good motivation

t tests; Bonferroni adjustment for multiple comparisons

Report of “slow case” group after first 100 participants; also described in full sample by Storzbach et al. (2000); a limitation is that slow group has only 13 veterans, so it is hard to identify important determinants of slow cases

NOTE: ANCOVA = analysis of covariance; BARS = Behavioral Assessment and Research System; CAPS = Clinician Administered PTSD Scale; CPT = Continuous Performance Test; CVLT = California Verbal Learning Test; HSC = Health Symptom Checklist; MANOVA = multivariate analysis of variance; MDD = Major Depressive Disorder; MMPI = Minnnesota Multiphasic Personality Inventory; NES = Neurobehavioral Evaluation System; ODTP = Oregon Dual Task Procedure; PASAT = Paced Auditory Serial Addition Test; R = Revised; RMANOVA = Repeated Measures Analysis of Variance; SCL = Symptom Check List; TOMM = Test of Memory Malingering; UPSIT = University of Pennsylvania Smell Identification Test; WAIS = Wechsler Adult Intelligence Scale; WMS = Wechsler Memory Scale.

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

DISEASES OF THE NERVOUS SYSTEM (ICD-10 G00-G99)

The neurologic outcomes reviewed in this section are amyotrophic lateral sclerosis (ALS), peripheral neuropathy, and other neurologic outcomes.

Amyotrophic Lateral Sclerosis

ALS is a neuromuscular disorder; it is often referred to as Lou Gehrig’s disease and might also be called motor neuron disease or Charcot’s disease. It affects approximately 20,000 to 30,000 people in the United States (NINDS 2006; The ALS Association 2006). ALS affects all races and ethnic backgrounds and the risk is higher in men than women of the same age (Annegers et al. 1991). The disease is often relentlessly progressive and almost always fatal. The rate of progression is quite variable from patient to patient.

ALS causes degeneration of the motor neurons in the cerebral motor cortex (called upper motor neurons) and the brain stem and spinal cord (called lower motor neurons) (Rowland 2000). The motor neurons are nerve cells that provide communication between the highest levels of the nervous system and the voluntary muscles of the body (NINDS 2006). When the upper motor neurons degenerate, their connections to the lower motor neurons and spinal interneurons are disrupted. That disruption leads to weakness of muscles in a characteristic distribution and the development of spasticity. Lower motor neuron degeneration disrupts nerve contact to the muscles resulting in muscle atrophy. Spontaneous muscle activity, called fasciculation, occurs. Eventually, affected people are unable to move their arms and legs and cannot speak or swallow. When the connection is disrupted between the neurons and the muscles responsible for breathing, patients either die from respiratory failure or require mechanical ventilation to continue to breathe. The majority of persons with ALS die from respiratory failure within 5 years from the onset of symptoms. To be diagnosed with ALS, patients must have signs and symptoms of both upper and lower motor neuron damage that cannot be attributed to other causes (such as progressive muscular atrophies and varieties of peroneal muscular atrophy, Kennedy’s syndrome, or multifocal motor mononeuropathy).

Five to ten percent of ALS cases are familial (inherited) and the remainder are sporadic (Rowland 2000; Siddique et al. 1999). Only one parent needs to carry the mutant gene for ALS to occur in about half of the children in cases of familial ALS (NINDS 2006). The specific gene mutations causing the majority of familial ALS cases are unknown. However, about 20% of familial cases are believed to be caused by a mutation in a gene that encodes the enzyme superoxide dismutase 1 (NINDS 2006).

The cause of sporadic ALS is unknown. Despite a number of epidemiologic studies examining occupation (for example, Italian professional soccer players, farmers, and electricians), physical trauma, strenuous physical activity, lifestyle factors (for example, diet, body mass index, cigarette use, and alcohol consumption), ethnic group, and socioeconomic status, there are no consistent findings (Armon 2003; Armon 2004a; Chio et al. 2005; Rowland 2000; Valenti et al. 2005). It has been suggested that the risk of ALS may decrease with decreasing latitude (north-south gradient of risk) (McGuire et al. 1996a).

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Primary Studies

Horner and colleagues (2003) conducted a nationwide, epidemiologic case-ascertainment study to determine if Gulf War veterans have elevated rates of ALS. They used active and passive methods of case ascertainment. Active methods included screening of inpatient, outpatient, and pharmacy medical databases of VA or DOD. Passive methods included establishment of a toll-free telephone number, solicitations through relevant Internet sites, and mass mailings of study brochures to practicing neurologists in the VA and to members of the American Academy of Neurology. The ALS diagnosis was verified by medical-record review.

Among nearly 2.5 million eligible military personnel, nearly 700,000 had been deployed to the Gulf War. In the deployed population, there were 107 cases of ALS identified, for an overall occurrence of 0.43 per 100,000 persons per year. Most of the cases were found with active ascertainment methods. An increased risk of ALS was found among all deployed personnel (RR 1.92, 95% CI 1.29-2.84). The attributable risk (that is, the risk difference or excess risk) associated with Gulf War deployment was 18% (95% CI 4.9-29.4%). The foremost limitation of the study was potential underascertainment of cases, particularly among nondeployed veterans, because nondeployed veterans had less incentive to participate. Because of the rarity of ALS, underascertainment of a few cases, particularly if it is greater among the nondeployed, can substantially exaggerate results. Underascertainment of nondeployed cases would make the risk among the deployed appear higher by comparison. Contributing to this concern is the finding that the incidence of ALS in deployed veterans was actually lower than that of an age-matched sample from Washington state (McGuire et al. 1996b). Another study limitation was failure to consider smoking as a possible confounding factor in the study design. Smoking may be a risk factor for the development of ALS (Nelson et al. 2000). Those study limitations were raised in a letter to the editor by an ALS researcher (Armon 2004b). In response, the study authors pointed out the low likelihood that smoking rates would be different among deployed and nondeployed veterans. They also undertook a secondary analysis to address concerns about differential case ascertainment among deployed vs nondeployed veterans.

The secondary analysis (Coffman et al. 2005) estimated the occurrence of ALS with capture-recapture statistical methods. The study used three methods of capture-recapture: log-linear models, sample coverage, and ecologic models. The investigators found modest underascertainment of cases in nondeployed military personnel and little underascertainment in the deployed. Correcting the rates for underascertainment, the investigators still found a higher age-adjusted risk of ALS among the deployed than among the nondeployed (RR 1.77, lower bound 1.21, with log-linear model). The results confirmed the original findings of Horner et al. (2003) of an increase in ALS among deployed veterans. Their estimates also address the criticism that the incidence among the deployed is lower than that in the study of Washington State. See Table 5.8 for a summary of the primary ALS studies discussed above.

Secondary Studies

Haley (2003) found an excess incidence of ALS among deployed veterans in comparison with the expected incidence based on US vital statistics. His analysis spanned 1991-1998. In the first half of that period the increased incidence was not apparent; from 1995 to 1998, the incidence more than doubled (standardized mortality ratio [SMR] 2.27, 95% CI 1.27-3.88). Although the study used passive and active means of case ascertainment similar to those of Horner et al., it differed in several key respects: it restricted cases to those below the age of 45 years (instead of all ages); it used 8 years of followup (instead of 10 years), and it used as a

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

comparison population the age-adjusted rates from US mortality statistics (instead of age-adjusted rates in nondeployed veterans). The major criticism of the study is its use of mortality statistics for the general population to estimate the “expected” incidence (Armon 2004b). That may have underestimated the expected rates for the comparison population, thereby making the SMRs appear higher than they actually were.

Several US and UK mortality studies have not found an excess risk of ALS, but they did not have sufficiently long followup or sufficiently detailed methods (Defence Analytical Services Agency 2005; Kang and Bullman 1996; Macfarlane et al. 2000). A hospitalization study (Smith et al. 2000) also found no difference in relative risk of ALS (RR 1.66, 95% CI 0.62-4.44), but the authors acknowledge that they had too few cases to make valid comparisons between deployed and nondeployed veterans. The study was also limited by inclusion of only active-duty military personnel and only 6 years of followup. Nicolson and colleagues, studying eight Gulf War veterans with ALS and two other comparison populations, found that ill Gulf War veterans had the highest frequency of systemic mycoplasm infections (Nicolson et al. 2002). Although the authors suggest that mycoplasma might be involved in the pathogenesis or progression of ALS, insufficient information was given regarding the selection of cases and controls to determine whether they are representative.

Summary and Conclusion

Two primary studies and one secondary study found that deployed veterans appear to be at increased risk for ALS. The primary study, by Horner et al., which had the possibility of underascertainment of cases in the nondeployed population, was confirmed by a secondary analysis by Coffman et al. That analysis, using capture-recapture analysis, confirmed the nearly doubled risk. A secondary study by Haley, using general population estimates as the comparison group, found a slightly higher relative risk but within the same general range. Other US and UK mortality studies and a hospitalization study have not found excess risk of ALS. The committee concludes that the further followup is warranted.

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

TABLE 5.8 Amyotrophic Lateral Sclerosis (ALS)

Reference

Design

Population

Outcome

Results

Adjustment

Comments or Limitations

Horner et al. 2003

Retrospective cohort

All active, Gulf War-deployed military personnel (1990-1991), compared withnon-Gulf War-deployed veterans

ALS

All deployed forces, significant increased risk of ALS

RR = 1.92% 95% CL -1.29-2.84

Age-adjusted average, annual

10-year incidence; attributable risk

Case ascertainment method through screening of VA and DOD medical databases and benefit files (and TriCare) by ICD-9 code for ALSor riluzole use; toll-free telephone enrollment; Internetnotices; mass mailings to neurologists, VA centers, and veteran service organizations

Coffman et al. 2005

Capture-recapture reanalysis of Horner et al. cohort

See Horner et al. 2003

ALS

Found no under as certainment of ALS cases among deployed

Log-linear models; sample coverage; ecologic models

Possible slight undercounts not likely to substantively affect results

NOTE: DOD= Department of Defense; VA =Department of Veterans Affairs.

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

Peripheral Neuropathy and Other Neurologic Outcomes

Peripheral Neuropathy

This section reviews studies of peripheral neuropathy, polyneuropathy, or neuromuscular symptoms, as identified by the investigators’ conducting the studies. Peripheral neuropathy, broadly defined, is a disease of the peripheral nerve tissues (that is, nerve fibers ensheathed by Schwann cells, including nerve roots), which transmit information from the brain and spinal cord to other parts of the body.

Numerous types of peripheral neuropathy have been characterized, each with its own set of symptoms, patterns of development, and prognosis. Peripheral neuropathy can be classified by a variety of factors, such as the population of nerves fibers affected (for example, motor, sensory, or autonomic). Additionally, neuropathy can be classified by the time course (acute, subacute and chronic, remitting, or relapsing); and by pathology (axonal, demyelinating, or other). Peripheral neuropathy might be inherited (for example, resulting from inborn errors in the genetic code or mutations) or acquired (for example, from physical injury, tumors, toxins, autoimmune responses, nutritional deficiencies, alcoholism, vascular and metabolic disorders, or infections from conditions such as leprosy, HIV, herpes simplex and zoster, or hepatitis associated) (National Institute of Neurological Disorders and Stroke 2006). Polyneuropathy is a neurologic disorder characterized by progressive weakness and impaired sensory function in arms and legs. The committee notes that an objective inquiry of peripheral neuropathy depends on clinical recognition of absent ankle reflexes; distal symmetric leg and foot weakness and atrophy; sensation loss in toes and feet; and abnormalities in nerve conduction. The committee also regards studies with objective and quantitative measures, such as those with nerve-conduction tests, to be optimal.

The best population-based questionnaire study for assessment of peripheral neuropathy is that of Cherry and colleagues (Cherry et al. 2001a), who studied UK troops deployed to the Gulf War. Almost 35% of Gulf War veterans who reported handling pesticides for more than a month indicated numbness or tingling on mannequin diagrams compared with 13.6% of veterans who did not report handling pesticides. However, although the study was well-designed and suggested a dose-response relationship, it was limited by recall bias, lack of clinical evaluations, and the absence of nerve-conduction studies. Self-reporting of peripheral neuropathy symptoms has poor diagnostic accuracy (Franse et al. 2000). Because of those limitations, the committee defined primary studies as requiring, at a minimum, medical evaluations and objective nerve testing.

Primary Studies

In the medical evaluation component of the large, population-based cohort assembled by VA, Davis and colleagues (2004) reported on the presence of distal symmetric polyneuropathy in deployed and nondeployed veterans. That condition was evaluated through history, physical examination, and standardized electrophysiologic assessment of motor and sensory nerves in 1,061 deployed veterans and 1,128 nondeployed veterans. Spouses of deployed and nondeployed veterans were also studied, as was a population of 240 Khamisiyah-exposed deployed veterans. Exposure to potential nerve agents was assessed with one of the first DOD models of atmospheric dispersion (Winkenwerder 2002). Blood studies were performed to rule out metabolic causes of neuropathy. Although the study provided results on distal symmetric polyneuropathy as distal sensory or motor neuropathy identified on the basis of the neurologic examination, nerve conduction study, or both, the committee favored distal symmetric

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

polyneuropathy identified with a nerve conduction study as the best, most reliable measure of peripheral neuropathy. No significant differences between adjusted population prevalence of distal symmetric polyneuropathy in deployed and nondeployed veterans were found (OR 0.65, 95% CI 0.33-1.28). There also were no differences on physical examination or self-reported peripheral neuropathy, although at the time of examination, deployed veterans reported more numbness and tingling than did nondeployed veterans. The veterans exposed to the Khamisiyah ammunition depot explosion did not differ significantly from nonexposed deployed veterans (OR 1.04, 95% CI 0.25-4.37). The prevalence of distal symmetric polyneuropathy in the spouses of deployed and non-deployed veterans also did not differ; however, the measure of distal symmetric polyneuropathy was obtained through self-reports as opposed to medical evaluation or nerve conduction study. One limitation of the study is potential participation bias: only 53% of deployed veterans and 39% of nondeployed veterans invited to participate were actually examined.

Neuromuscular symptoms of UK veterans were evaluated with objective testing of peripheral nerves, skeletal muscles, or neuromuscular junctions in a case-control study (Rose et al. 2004; Sharief et al. 2002). Ill veterans (with more than four neuromuscular symptoms and lower functioning according to the SF-36) were compared with healthy deployed veterans, 13 symptomatic Bosnian veterans, and 22 symptomatic Gulf-War-era controls. All groups had been randomly selected from 8,195 male military personnel. In the first publication, veterans underwent nerve-conduction studies, quantitative sensory and autonomic testing, and concentric needle and single-fiber electromyography. In the second, they underwent quantitative myometry through the ischemic forearm exercise test, the subanaerobic bicycle exercise test, and a muscle biopsy. The studies revealed no statistically significant differences between deployed and nondeployed veterans who had symptoms of Gulf War illness. The sole exception was the greater effort required for the bicycle exercise test with increased lactate production; this finding could reflect mitochondrial damage or inactivity resulting from ill health. See Table 5.9 for a summary of the peripheral neuropathy findings.

Secondary Studies

Other studies supporting the absence of findings include those of Eisen et al. (2005), Joseph et al. (2004), Rivera-Zayas et al. (2001), and Amato et al. (1997). Eisen et al. (2005) appears to have reported in less detail on the peripheral neuropathy findings in the same cohort as previously reported by Davis and colleagues (2004) but the precise relationship between the two publications is not clear. The Joseph et al. study, a retrospective review of electrophysiologic testing of 56 Gulf War veterans and 120 nondeployed veterans, showed no objective evidence of a higher incidence of neuromuscular disease in deployed veterans than in nondeployed veterans. In the Rivera-Zayas et al. study, 12 of 162 Gulf War veterans tested electrophysiologically with positive questionnaires for neuropathy had normal results except for two subjects who had carpal tunnel syndrome. Amato et al. showed that in 20 Gulf War veterans who had severe muscle fatigue, weakness, and myalgias, nerve conduction studies, repetitive nerve stimulation, quantitative and single-fiber electromyography, and muscle biopsies were inconclusive.

Other Neurological Outcomes

Haley and colleagues performed detailed neurologic assessments in several case-control studies of the original cohort of Seabee reservists. The cases were veterans who had met criteria for factor-derived syndromes. Under the hypothesis that those veterans were ill from neurotoxic exposures, especially to organophosphates, the assessments covered broad neurologic function

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

(Haley et al. 1997a), autonomic function (Haley et al. 2004), vestibular function (Roland et al. 2000), basal ganglia injury (Haley et al. 2000a; 2000b), and paraoxonase (PON) genotype and serum concentrations (Haley et al. 1999). Separate groups of investigators also studied PON genotype or activity (Hotopf et al. 2003; Mackness et al. 1997). A case-control study of neuropsychologic functioning (Hom et al. 1997) is discussed elsewhere in this chapter.

The committee regarded those case-control studies as secondary studies primarily because of their lack of generalizability and strong potential for selection bias. Although their study design was characterized as nested case-control, the studies of Haley et al. were not true nested case-control studies. Cases were, appropriately, selected from the original cohort, but controls were not. Ten of the 20 controls were from 150 newly discovered members of the battalion who had not been deployed. Those 10 were not from the original cohort and there is no indication that they were tested to determine whether they should be treated as cases. The selection of those controls raises the possibility of selection bias. With regard to the other concern, lack of generalizability, the authors selected as cases the most severely affected veterans—that is, those who scored highest on factor analysis-derived syndromes—rather than a random sample of those who met a particular case definition.

Summary and Conclusion

One large, well-designed study conducted by VA did not find evidence of excess peripheral neuropathy in Gulf War veterans. The study used the most thorough and objective case definition. Several other secondary studies supported a conclusion of no excess risk. Some studies (such as that of Cherry et al. (2001a)) did report higher rates of peripheral neuropathy, but they used self-reports, which the committee did not accept as a reliable measure of peripheral neuropathy. Furthermore, because researchers use different case definitions of peripheral neuropathies, there are problems of ascertainment, which makes comparisons among groups difficult.

The committee finds no increase in the prevalence of peripheral neuropathy in deployed vs nondeployed veterans, as defined by history, physical examination, and electrophysiologic studies. Regarding other neurologic outcomes, Haley and colleagues found evidence of basal ganglia injury and other abnormalities with detailed neurologic assessments, in several case-control studies. The committee regarded the studies as secondary because of their lack of generalizability and their strong potential for selection bias.

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

TABLE 5.9 Peripheral Neuropathy

Reference

Design

Population

Outcomes

Results

Adjustments

Comment or Limitations

Davis et al. 2004

Cross-sectional, prevalence, medical evaluation

1,061 deployed veterans vs 1,128 nondeployed veterans; 240 Khamisiyah-exposed deployed veterans vs 807 non-Khamisiyah-exposed deployed veterans; 482 spouses of deployed vs 527 spouses of nondeployed veterans as controls

Distal symmetric polyneuropathy identified by nerve-conduction studya

Deployed vs. nondeployed veterans: OR 0.65 (95% CI 0.33-1.28); Khamisiyah-exposed deployed veterans vs non-Khamisiyah-exposed deployed veterans: OR 1.04 (95% CI 0.25-4.37); no difference between spouses of deployed vs spouses of nondeployed veterans

Excludes coexisting conditionsb

Low participation rate: 53% in deployed veterans, 39% in nondeployed veterans

Rose et al. 2004, Sharief et al. 2002

Case-control

49 symptomatic deployed veterans vs 26 healthy deployed veterans, 13 symptomatic Bosnia deployed veterans, 22 symptomatic Gulf War-era veterans

Nerve-conduction studies, quantitative sensory and autonomic testing, concentric needle and single-fiber, electromyography, ischemic forearm exercise test, subanaerobic bicycle exercise test, muscle biopsy

No significant differences except bicycle exercise test more effortful and produced higher plasma lactate

 

One positive finding could reflect mitochondrial damage or inactivity resulting from ill health

  

aAlthough study defined distal symmetric polyneuropathy as distal sensory or motor neuropathy identified on basis of neurologic examination, nerve conduction study, or both, committee defined it by nerve-conduction study alone.

  

bAlcoholdependence, diabetes mellitus, renal insufficiency, hypothyroidism, AIDS/HIV, collagen vascular disease, and neurotoxic medications.

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

CHRONIC FATIGUE SYNDROME

Chronic fatigue syndrome (CFS) is marked by severe and persistent fatigue with a cluster of other symptoms. Fatiguing syndromes were chronicled 100 years ago and have long been the focus of considerable controversy in the medical establishment (Straus 1991; Wessely 1998). The study of unexplained fatiguing illnesses was greatly facilitated and legitimized in the last decade with the development of a case definition sponsored by the Centers for Disease Control and Prevention (CDC). That case definition has helped clinicians and scientists to recognize and classify CFS (Table 5.10). CDC’s case definition, first published in 1988 and revised in 1994 (see below), requires fatigue and related impairment in function, and the occurrence of four of eight other defining symptoms over at least 6 months (Fukuda et al. 1994; Holmes et al. 1988). Of the eight symptoms, the most commonly reported are headaches, post-exertional malaise, impaired cognition, and muscle pain (Buchwald and Garrity 1994).

The etiology of CFS is unknown, and no widely accepted laboratory tests or pathologic physical signs are widely accepted (Epstein 1995). Several biologic correlates of the syndrome have emerged, including dysregulation of the hypothalamic-pituitary-adrenal axis, immune activation, and other measures (Goshorn 1998), but they might be present in only a minority of patients; and those findings are not specific to CFS. Although infectious agents may trigger some cases of CFS, a complex, multifactorial etiology that incorporates biologic, psychologic, and social factors is likely (Wessely 1998). The degree of disability associated with CFS is striking, with high rates of unemployment (Bombardier and Buchwald 1996; Buchwald et al. 1996) and poor quality of life related to health (Hardt et al. 2001; Komaroff et al. 1996).

Before defining a primary and secondary study, it should be noted that CFS is a diagnosis of exclusion. The CDC criteria require that three elements be completed as part of a comprehensive evaluation. The first element, determining whether the symptom criteria for CFS are present, requires that a person be queried specifically about length and severity of fatigue and about eight ancillary symptoms. The second element, determining whether other medical conditions are present, mandates a complete physical examination, a battery of specified laboratory tests, and a medical history. The third element, assessing exclusionary psychiatric conditions, requires an interview by a trained professional to obtain diagnostic information.

Thus, in this report, a primary study for CFS is one in which CFS has been diagnosed. A secondary study is one in which a CFS-like condition has been documented. Both primary and secondary studies needed to include a suitable control group so that findings could be interpreted. Other studies that estimated the prevalence of symptoms of “chronic fatigue” (Gray et al. 1999a; Unwin et al. 1999), or multisymptom illness (Fukuda et al. 1998), are not considered further in this section. Likewise, studies that used scalar measures of disability and poor quality of life related to health (Reid et al. 2001) as surrogates for the CDC criteria are not included. Finally, self-reports of CFS (Unwin et al. 1999) and self-reports of a physician diagnosis of CFS (Gray et al. 2002) were not included among the secondary studies, because diagnostic data obtained that way are highly inaccurate. For example, in the Eisen et al. (2005) study, which the committee considered to be the only primary study, only two or three of 38 deployed and eight nondeployed veterans who self-reported CFS received a formal diagnosis after a comprehensive examination. Others, using a method of classifying a case of CFS based on cutoff scores on a fatigue scale and a functional status instrument, found that only 11% of veterans reporting a diagnosis of CFS met operational CFS study criteria.

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

TABLE 5.10 Case Definition of Chronic Fatigue Syndrome

Reference

Case Definition

CDC Criteria

Presence of both the following:

Clinically evaluated, unexplained, persistent or relapsing chronic fatigue that is of new or definite onset (that is, has not been lifelong), is not the result of ongoing exertion, is not substantially alleviated by rest, and results in substantial reduction in levels of occupational, educational, social, or personal activities. Clinical evaluation includes medical history, physical examination, laboratory studies, and psychiatric assessment.

Concurrent occurrence of four or more of the following, which must have persisted or recurred during 6 or more consecutive months of illness and must not have predated the fatigue: self-reported impairment of short-term memory or concentration severe enough to cause substantial reduction in levels of occupational, educational, social, or personal activities; sore throat; tender cervical or axillary lymph nodes; muscle pain; multijoint pain without joint swelling or redness; headaches of a new type, pattern, or severity; unrefreshing sleep; and postexertional malaise lasting more than 24 hours.

SOURCE: Fukuda et al. 1994.

Primary Studies

Only one primary study was identified (see Table 5.12). CFS was one of 12 primary health outcome measures studied by Eisen and colleagues (2005), who conducted medical evaluations in phase III of the nationally representative, population-based VA study. In the period 1999-2001, 1,061 of 11,441 deployed and 1,128 of 9,476 nondeployed veterans selected were evaluated. Those veterans had participated in the phase I survey study conducted in 1995 (Kang et al. 2000). The veterans were randomly selected, and the researchers were blinded to their deployment status. The diagnosis of CFS was based on in-person interviews, examinations, and the strict application of the CDC criteria (Fukuda et al. 1994). One study limitation was that deployed veterans were significantly younger, less educated, less likely to be married, and of lower income than nondeployed veterans, although the analysis adjusted for those factors. Another limitation was that, despite three recruitment waves, participation in the 2005 study was low: only 53% of Gulf War veterans and 39% of nondeployed veterans participated. However, to examine biases associated with nonparticipation, the authors obtained sociodemographic and self-reported health data previously collected in 1991 on participants and nonparticipants from the DOD Manpower Data Center and the 1995 VA study (Kang et al. 2000). Because both deployed and nondeployed participants were more likely than nonparticipants to report CFS, this disparity was adjusted for in the analysis of prevalence. The authors concluded that the population prevalence of CFS was higher in deployed than in nondeployed veterans: 1.6% vs 0.1% (OR 40.6, 95% CI 10.2-161.15). Study strengths are its large, population-based design, stratified sampling method, analysis of participation bias, comprehensive examination, and use of computer-based algorithms by researchers who were blinded to deployment status.

Secondary Studies

An earlier questionnaire study, conducted during phase I of the VA study, surveyed 11,441 deployed and 9,476 nondeployed veterans (Kang et al. 2003). Several items in the 48-item symptom questionnaire served as the basis for meeting the case definition for CFS. After exclusion of veterans who self-reported exclusionary medical conditions that could explain their

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

fatigue, 4.9% of deployed and 1.2% of nondeployed veterans (OR 4.2, 95% CI 3.3-5.5) met the case definition. The investigators found that CFS was not related to the severity of combat stressors. The latter was assessed according to responses to questions on wearing chemical protective gear or hearing chemical alarms, being involved in direct combat duty, or witnessing any deaths. The study was limited by its reliance on solely self-reported symptoms without a physical or laboratory examination and on self-reported physician-diagnosed conditions. Those shortcomings resulted in a higher rate of CFS-like illness than was observed when the same cohorts were sampled and underwent more rigorous medical evaluations as in Eisen et al. (2005).

Proctor and colleagues (2001) conducted in-person interviews of 180 Army veterans selected from the larger Fort Devens cohort to determine the prevalence of CFS. The deployed veterans were compared with 46 members of an air ambulance company deployed to Germany during the Gulf War. The prevalence was determined only according to the symptom criteria specified by the CDC case definition (Fukuda et al. 1994). With that approach, the rate was higher in the Gulf-deployed than the Germany-deployed group (7.5% vs 0%, p = 0.02). When additional information from self-reported medical or psychiatric conditions (such as substance abuse and bipolar disorder) and clinical psychiatric interviews was considered, the prevalence in Gulf veterans decreased to 2%, which was no longer statistically significant. The study demonstrated the importance of performing psychiatric assessments, but it was limited by the relatively small sample and the lack of medical or laboratory evaluations.

Canada deployed more than 3,000 sea, land, and air troops to the gulf region; they participated in a naval blockade and were responsible for one-fourth of enemy interceptions in the gulf. A survey of the entire Canadian Gulf War forces found that deployed veterans were at least 5 times as likely as nondeployed veterans to report symptoms of CFS (OR 5.27, 95% CI 3.95-7.03) (Goss Gilroy Inc. 1998). Veterans were not interviewed or examined, and all data were obtained from self-reports. The CFS-like illness was based on responses to questions derived from the CDC criteria and a score above zero on the Chalder fatigue scale.4 With only minor modifications, the items used in this study were the same as those used by the Iowa Persian Gulf Study Group (1997). The study was limited by the lack of in-person interviews and examinations and by the nontraditional assessment of CFS.

The Iowa study (1997) surveyed 1,896 deployed and 1,799 nondeployed veterans who listed Iowa as their home state at the time of enlistment. The presence of a CFS-like condition was based on a combination of symptoms used in the CDC criteria (Fukuda et al. 1994) and scores on a fatigue scale (Chalder et al. 1993). The investigators found that the prevalence differed by 1.4% (95% CI 0.9-2.0) after adjusting for age, sex, race, branch of military, and rank. Study limitations were the use of self-reports of symptoms on a questionnaire and the lack of medical evaluations. Although rigorously conducted and analyzed, the study suffers shortcomings similar to those of the Canadian study.

Summary and Conclusion

Because the diagnosis of CFS depends entirely on symptoms, not on physical or laboratory findings, the prevalence is highly variable from study to study. In addition, some of the secondary studies reviewed were not limited to CFS but included fatigue or CFS-like illnesses. One primary study demonstrated a higher prevalence of CFS in deployed than in nondeployed veterans although the absolute difference in risk was very small (1.6% vs 0.1%).

4

The Chalder fatigue scale is widely used to measure physical and mental fatigue in CFS patients.

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

Secondary studies showed higher prevalences of CFS and CFS-like illnesses among veterans deployed to the Persian Gulf than among their counterparts who were not deployed to the gulf or who were deployed elsewhere.

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

TABLE 5.11 ChronicFatigue Syndrome (CFS)

Study

Design

Population

Outcomes

Results

Adjustments

Comments or Limitations

Eisen et al. 2005

Population-based; cross-sectional; prevalence; in-person medical and psychiatric evaluations

1,061 deployed and 1,128 nondeployed

CFS based on in-person interviews using CFS criteria of CDC (Fukuda et al. 1994) and exclusionary diagnoses from history, interviews, examinations, laboratory testing

Prevalence: 1.6% vs 0.1% (OR 40.6, 95% CI 10.2-161.15)

Age, sex, race, education, cigarette-smoking, duty type, service branch, rank

Low participation rates, especially among nondeployed; but analysis of nonparticipants and participants reveals that participants, both deployed and nondeployed, are more likely to report symptoms of CFS

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

DISEASES OF THE CIRCULATORY SYSTEM (ICD-10 I00-I99)

Cardiovascular disease is a broad, all-encompassing term for any disorder in the heart or the blood vessels. Heart disease is the leading cause of death of both women and men in the United States. In 2002, 696,947 people (51% of them women) died of heart disease, accounting for 29% of all US deaths. The age-adjusted death rate was 241 per 100,000 of population. Several of the studies that examined cardiovascular disease also included reports of diabetes. Type 2 diabetes is the most common form of diabetes and usually appears in adults, often in middle age (over the age of 45 years). Type 2 diabetes is a progressive disease that contributes to nearly 200,000 US deaths per year. It occurs frequently in people who are overweight; are of black, Latino or Hispanic, American Indian, Asian or Pacific Islander descent; or have a family history of the disease.

Primary Studies

There are two primary studies of cardiovascular effects or diabetes; one study examined the effect of possible nerve agent exposure (Smith et al. 2003) and the other examined deployment as the exposure (Eisen et al. 2005). The latter of the two studies is one of few studies that conducted medical evaluations.

The medical evaluation study is the culmination (phase III) of the nationally representative, population-based study sponsored by VA. Eisen and colleagues (2005) examined 12 primary health-outcome measures, two of which are relevant here: hypertension and diabetes. The study evaluated 1,061 Gulf War and 1,128 non-Gulf War veterans who had been randomly selected from 11,441 Gulf War deployed and 9,476 Gulf War-nondeployed veterans who previously participated in the phase I questionnaire conducted in 1995 (Kang et al. 2000). Neither VA study tested for relationships between exposures and health outcomes. Researchers were blind to deployment status. One limitation was that deployed veterans were significantly younger, less educated, less likely to be married, and of lower income, although the analysis attempted to adjust for those factors. Another limitation was that, despite three recruitment waves, the participation rate in the 2005 study was low, with only 53% of Gulf War veterans and 39% of nondeployed Gulf War veterans participating. To determine nonparticipation bias, the study authors obtained previously collected findings on participants and nonparticipants from the DOD Manpower Data Center and gathered sociodemographic and self-reported health findings from the 1995 VA study (Kang et al. 2000). The lack of difference in hypertension and diabetes between deployed participants and nonparticipants and between nondeployed participants and nonparticipants suggested the lack of participation bias. The medical evaluation revealed no statistically significant differences between deployed and nondeployed in the prevalence of hypertension (OR 0.90, 95% CI 0.60-1.33) or diabetes (OR 1.52, 95% CI 0.81-2.85). The nonsignificant statistical increase in diabetes may warrant further investigation.

Two additional DOD hospitalization studies examined hospitalizations in relation to possible exposure to sarin and cyclosarin as a result of demolishing weapons at Khamisiyah, Iraq, in March 1991. The first, by Gray and colleagues (1999b), was superseded by the second, by Smith and colleagues (2003), which used a more advanced exposure model. Those researchers analyzed hospitalizations in 1991-2000 among 431,762 active-duty military deployed

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

to the Gulf War Theater during the time of the Khamisiyah demolition. Exposure status was determined by whether active-duty military were within the plume area defined by meteorologic-dispersion modeling, according to DOD's revised modeling in 2000 (Rostker 2000), and whether they were within a military unit determined, according geographic information systems data, to have been exposed during a 3-day period at concentrations set by CDC for the “general population limit”(GPL), below which no symptoms are expected. The GPL was adjusted because of the briefer duration of troops’ potential exposure. Troops were considered exposed if concentrations were above the adjusted GPL of 0.000003 mg/m3 for sarin and 0.00001 mg/m3 for cyclosarin.

Investigators studied hospitalization diagnoses from 15 ICD-10 categories, including “circulatory system diseases” and “endocrine, nutritional, and metabolic diseases”. On the basis of Cox's proportional hazard modeling, only one category of disease circulatory—system disease—was related to exposure (RR 1.10, 95% CI 1.05 -1.16), after adjustment for other variables in the model. The increased RR was due to one of 10 cardiac diagnoses: cardiac dysrhythmias (RR 1.23, 95% CI 1.04-1.44). While acknowledging that the finding could have been due to chance, the investigators concluded that the excess in dysrhythmia was "small in comparison with potential observational variability, but the findings are provocative and warrant further evaluation." The authors identified study limitations as use of hospitalizations because outpatient data were unavailable, restriction to DOD hospitals (because of the availability of computerized records), and restriction to hospitalizations of those who remained on active duty after the war. It also was not possible to adjust for confounding exposures. A further problem is the uncertainty in the exposure model (see Chapter 2). See Table 5.12 for a summary of the primary papers reviewed for cardiovascular disease and diabetes.

Secondary Studies

Several large-scale epidemiologic studies included self-reported cardiovascular outcomes or diabetes with new onset after the Gulf War. The committee regarded studies that relied on self-reports of cardiac outcomes as secondary studies. None of the studies, except the VA study, verified the diagnosis with physician records.

No postwar differences between deployed and nondeployed were found in two population-based cohort studies. Australian veterans deployed to the Gulf War were as likely as nondeployed veterans to report physician-diagnosed high blood pressure (OR 1.2, 95% CI 0.9-1.6) (Kelsall et al. 2004a). Similarly, Kansas veterans were as likely as their nondeployed counterparts to report physician-diagnosed high blood pressure (OR 1.24, 95% CI 0.82-1.89), heart disease (OR 1.56, CI 0.69-3.56), or diabetes (OR 1.22, 95% CI 0.45-3.30) (Steele 2000). Other studies had mixed findings. The large VA study, with its phase I questionnaire (Kang et al. 2000), found no statistical increase in physician-diagnosed stroke or diabetes in deployed veterans, but did find increases in physician-diagnosed hypertension; the difference in incidence between deployed and nondeployed was 3.84% (95% CI 3.75-3.93). That finding, however, occurred in the same cohort in which later medical evaluations found no statistical increase in hypertension (Eisen et al. 2005). Studying all Seabee commands, Gray and colleagues (2002) found no statistical increase in physician-diagnosed diabetes but did find increased hypertension (OR 1.63, 95% CI 1.36-1.95). McCauley and colleagues (2002) examined the effects of potential nerve-agent exposure on a cohort of veterans from Portland and Washington. The investigators found no statistical differences in self-reporting of physician-diagnosed diabetes, high blood pressure, or heart disease when they divided their cohort into potentially exposed and unexposed,

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

with exposure defined as being within 50 miles of the Khamisiyah demolition. But when dividing their cohort into deployed and nondeployed to the Gulf War, they found more self-reported physician-diagnosed high blood pressure (OR 1.7, 95% CI 1.3-2.4) and heart disease (OR 2.5, 95% CI 1.1-6.6), but no increase in diabetes (OR 1.0, 95% CI 0.5-2.4). Finally, one study of UK veterans (Unwin et al. 1999) found higher prevalence of self-reported high blood pressure: 8.8% of deployed veterans reported high blood pressure compared with 4.3% of Bosnia veterans and 6.6% of Gulf War-era veterans.

Two postwar hospitalization studies did not detect any differences between deployed and nondeployed active-duty personnel regarding circulatory system or endocrine diseases (Gray et al. 1996; Gray et al. 2000). The methods of those studies are described in the section on hospitalizations and are listed in Table 5.13. Those hospitalization studies had several limitations, including lack of generalizability because they covered only active-duty personnel; inability to capture illnesses with longer latency, such as cancer; and inability to measure outpatient treatment, which is more likely to detect diabetes and cardiovascular effects if they occur. Medical evaluation studies, like that by Eisen and colleagues, have greater capacity to detect cardiac outcomes or diabetes.

Fukuda et al. (1998) studied an Air National Guard unit from Pennsylvania in relation to three comparison groups. Their major goal was to develop a case definition of multisymptom illness and to correlate it with clinical and laboratory examinations. Their case definition required one or more chronic symptoms in two of three categories: fatigue, mood and cognition, and musculoskeletal. Some 45% of deployed and 15% of nondeployed veterans met the case definition. Results of clinical examinations were essentially negative in both groups.

Summary and Conclusion

Primary studies found no statistically significant differences in cardiovascular disease between deployed and nondeployed veterans. However, the specific cardiac outcomes under study were limited to hypertension and abnormal blood sugar. The one study that found an increase in cardiovascular disease was limited to hospitalizations and compared deployed veterans who were possibly exposed to the Khamisiyah plume and those who were not exposed. The increase was due entirely to an increase in cardiac dysrhythmia (Smith et al. 2003). The small but significant increase seen in this single study bears watching, though the study did not address the risk in deployed vs nondeployed veterans. In the secondary studies, deployed veterans were generally more likely to self-report hypertension and palpitations, but those reports were not confirmed in medical evaluations. Thus, it does not appear that there is a difference in the prevalence of cardiovascular disease or diabetes between deployed Gulf War veterans and nondeployed.

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

TABLE 5.12 Cardiovascular Disease or Diabetes

Study

Design

Population

Outcomes

Results

Adjustments

Comments or Limitations

Eisen et al. 2005

Population-based; cross-sectional; prevalence; medical evaluation

1,061 deployed, 1,128 nondeployed

Hypertension = blood pressure > 140/90 mm Hg or history of hypertension and use of antihypertensive medications; diabetes = fasting glucose of 6.99 mmol/L or greater (≥126 mg/dL) or taking hypoglycemic medication

Hypertension: OR 0.90 (95% CI 0.60-1.33)

Diabetes: OR 1.52 (95% CI 0.81-2.85)

Age, sex, race, years of education, cigarette-smoking, duty type, service branch, rank

Low participation rates, especially among nondeployed, but analysis of nonparticipants and participants reveals no differencesin hypertension or diabetes

Smith et al. 2003

DOD hospitalization study (1991-2000) of those potentially exposed to nerve agent

99,614 active-duty military considered exposed vs 318,458 nonexposed, according to revised DOD exposure model

ICD-10 diagnostic groups plus some individual codes

Circulatory system diseases: RR 1.10 (95% CI 1.05-1.16);

Cardiac dysrhythmia: RR 1.23 (95% CI 1.04-1.44);

Endocrine, nutritional, and metabolic diseases: RR 1.00 (95% CI 0.94-1.06)

One or more hospitalizations in a specific diagnostic category

Restricted to DOD hospitals; restricted to hospitalizations foronly Gulf War veterans who remained on active duty after the war; not possible to adjustfor confounding exposures

Gray et al. 1999b

DOD hospitalization study

Three exposure categories, according to revised DOD exposure model: not exposed (n=224,804), uncertain low dose exposure (n =75,717), estimated subclinical exposure (n =48,770)

ICD-10 diagnostic groups plus some individual codes

Circulatory system diseases: RR 1.12 (95% CI 0.93-1.33);

Endocrine, nutritional, metabolic diseases: RR 0.75 (95% CI 0.52-1.07)

Sex, age group, prewar hospitalization, race, service type, marital status, pay grade, occupation

Same as for Smith et al. 2003

NOTE: DOD = Department of Defense; VA = Department of Veterans Affairs.

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

DISEASES OF THE RESPIRATORY SYSTEM (ICD-10 J00-J99)

This section covers respiratory outcomes according to the major types of studies published in the Gulf War literature: respiratory outcomes in deployed vs nondeployed personnel without regard to specific exposures, and respiratory outcomes in relation to two types of exposures: to oil-well fires and to nerve agents. Table 5.14 contains a summary of all the studies reviewed in this section.

Associations of Respiratory Outcomes with Deployment in the Gulf War Theater

Primary Studies

Two publications reported on respiratory outcomes among participants in the medical evaluation component (phase III) of the large, population-based VA study (Eisen et al. 2005; Karlinsky et al. 2004). Eisen and colleagues (2005) evaluated 1,061 Gulf War and 1,128 non-Gulf War veterans who had been randomly selected from 11,441 Gulf War-deployed and 9,476 Gulf War-nondeployed veterans who previously had participated in a 1995 questionnaire survey (Kang et al. 2000). No statistically significant increase in the prevalence of self-reported asthma, bronchitis, or emphysema was observed among deployed veterans in models that adjusted for smoking and demographic variables (adjusted OR 1.07, 95% CI 0.65-1.77, for the three diseases combined). Obstructive lung disease was defined by the investigators as a history of lung disease (asthma, bronchitis, or emphysema) or pulmonary symptoms (wheezing, dyspnea on exertion, or persistent coughing with phlegm), and either the use of bronchodilators or at least 15% improvement in FEV15 after use of a short-acting bronchodilator. No increase in obstructive lung disease was observed among deployed personnel (adjusted OR 0.91, 95% CI 0.52-1.59). Limitations of the study include potential selection bias owing to low participation rates – 53% and 39% of deployed and nondeployed veterans, respectively.

Karlinsky and colleagues (2004) reported on results of pulmonary function tests (PFTs) on the same VA population that Eisen and colleagues studied. PFT results were classified into five categories: normal pulmonary function, nonreversible airway obstruction, reversible airway obstruction, restrictive lung physiology, and small airway obstruction. The patterns of PFT results were similar in deployed and nondeployed veterans; there were no statistically significant differences. The pattern of PFT results was also reported to be similar in those exposed and those not exposed on the basis of DOD exposure estimates developed in 2002 (see Chapter 2), of exposure to nerve agents through destruction of munitions at the storage site at Khamisiyah in 1991. Prevalences of self-reported pulmonary symptoms were higher in deployed veterans; however, self-reported diagnoses, use of asthma medications, and self-reported physician visits and hospitalizations for pulmonary conditions were similar in deployed and nondeployed. Although no adjustments were made for covariates, demographic variables were similar in the two groups and a history of tobacco-smoking was more common in deployed veterans than in nondeployed veterans (51% vs 44%, p = 0.03). Limitations include the inadequacy of the sampling strategy description to evaluate bias and no explanation of “matching” in the analysis.

5

Forced expiratory volume (FEV1) measures how much air a person can exhale during the first second of a forced breath.

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

Kelsall et al. (2004b) reported respiratory outcomes from a cross-sectional study of Australian Gulf War veterans. The gulf-deployed cohort comprised 1,456 participants and the nongulf military cohort included 1,588 participants. Response rates differed greatly between deployed (80.5%) and nondeployed (56.8%). Deployed veterans reported higher prevalences of all respiratory symptoms and some self-reported symptom-based respiratory diagnoses. Lung function measures adjusted for smoking and other covariates were somewhat higher in the deployed group (for example, FEV1/FVC% <70%; OR 0.8, 95% CI 0.5-1.1). FVC6, but not FEV1, was associated with self-report of exposure to oil-fire smoke. Although generally well conducted, the study was limited by the potential for selection bias, the lack of doctor-diagnosed respiratory conditions, and, with respect to effects of exposure to oil-fire smoke, the lack of availability of modeled exposure estimates.

Gray and colleagues (1999a) enrolled 1,497 study subjects from 14 Seabee commands in the US Navy, 527 of whom were Gulf War veterans and 970 were nondeployed veterans. Although respiratory symptoms were reported more frequently by the Gulf War veterans, pulmonary function measures adjusted for age, height, race, and smoking status were not associated with Gulf War status (mean FVC: Gulf War 4.96 L, non-Gulf War 4.99 L, p = 0.77; mean FEV1: Gulf War 4.05 L, non-Gulf War 4.04 L, p = 0.81).

A cross-sectional study of military personnel from Denmark, involved primarily in peacekeeping or humanitarian roles after the end of the Gulf War, also found increased respiratory symptoms among gulf-deployed personnel (n = 686) compared with nondeployed (n = 231) but no statistically significant differences in pulmonary function (FVC percent of expected: Gulf War 100.7, non-Gulf War 100.7; FEV1 percent of expected: Gulf War 95.6, non-Gulf War 96.4). Smoking patterns were very similar in the two groups (Ishoy et al. 1999b).

Secondary Studies

The overwhelming majority of secondary studies conducted among Gulf War veterans have found that several years after deployment veterans report higher levels of respiratory symptoms and of respiratory illnesses than nondeployed troops whether from the United States (Doebbeling et al. 2000; Gray et al. 2002; Iowa Persian Gulf Study Group 1997; Kang et al. 2000; Karlinsky et al. 2004; Kroenke et al. 1998; Petruccelli et al. 1999; Steele 2000), the UK (Cherry et al. 2001b; Nisenbaum et al. 2004; Simmons et al. 2004; Unwin et al. 1999), Canada (Goss Gilroy Inc. 1998), Australia (Kelsall et al. 2004b), or Denmark (Ishoy et al. 1999b). For example, the findings of the 1997 survey (Goss Gilroy Inc. 1998) mailed to the entire cohort of Canadian Gulf War veterans found an increase in self-reported respiratory disease (OR 1.35, 95% CI 1.16-1.57), bronchitis (OR 2.81, 95% CI 2.22-3.55), and asthma (OR 2.64, 95% CI 1.97-3.55) when adjusted for tobacco-smoking. The study by Eisen et al. (2005), described previously, is an exception in finding few differences among US veterans in respiratory symptoms and self-reported respiratory diagnoses between deployed and nondeployed troops 10 years after the Gulf War. And some of the many studies that have conducted factor analyses on reported symptoms have found respiratory factors (e.g., Cherry et al. 2001a; Cherry et al. 2001b; Hotopf et al. 2004). Of particular interest, some of the UK reports that found differences in respiratory symptoms and self-reported respiratory diagnoses have included comparisons of those deployed to the Gulf War with those deployed to Bosnia (Nisenbaum et al. 2004; Unwin et al. 1999). The Unwin et al. (1999) study found that the risk of self-reported asthma and bronchitis was higher in the Gulf

6

Forced vital capacity (FVC) is the total amount of air exhaled during an FEV test.

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

War cohort than in the Bosnia cohort (asthma: RR 1.2, 95% CI 0.8-1.6; bronchitis: RR 1.5, 95% CI 1.0-2.3). The Nisenbaum et al. (2004) study found most respiratory symptoms reported at least twice as often by the Gulf War cohort as by the Bosnia cohort.

Associations of Respiratory Outcomes with Specific Exposures Experienced by Gulf War Veterans During Their Deployment

Exposure to Oil-Well Fire Smoke

In February 1991, retreating Iraqi forces set fire to more than 600 oil wells. Fires burned over a 10-month period until November 1991, exposing thousands of US troops to combustion products. Several studies of US Gulf War veterans exposed to oil-well fires stand out from most other Gulf War studies by virtue of their focus on a narrow set of respiratory health outcomes and on a single type of exposure (smoke from oil-well fires) and their exposure assessment on the basis of models of troop-unit movements in relation to air-quality models that incorporated ground-based monitoring data, as well as satellite imagery. The vast majority of Gulf War health studies focused on multiple health outcomes, multiple exposures, and self-reporting of exposures without validation (see Chapter 2). The studies reviewed below examined long-term respiratory effects after deployment. Most of them did not distinguish between new cases and exacerbation of pre-existing respiratory illnesses. Asthma, for example, would not have been grounds for exclusion from entry into the military, so cases that predated deployment would be expected to be present in the deployed population. These studies are important because their exposure estimates are probably relatively robust inasmuch as data were actively collected at at least eight sites for many months and were integrated with remote-sensing imagery, and the locations of units on particular days are expected to have been reasonably valid.

Primary Studies

Cowan et al. (2002) conducted a case-control study to identify cases of physician-diagnosed asthma in the DOD registry (n = 873) and controls without asthma (n = 2,464). The DOD registry was established for active-duty Gulf War military who wished to receive a comprehensive physical examination. Cases were defined by using a military physician-assigned diagnosis of asthma (ICD-9-CM [Clinical Modification] codes 493 and 493.91) based on patient complaints, signs and symptoms, and a physical examination. Specified diagnostic criteria were not used. Pulmonary function testing data were not available. Exposure to smoke from oil-well fires was estimated by linking troop locations with modeled oil-fire smoke exposure. National Oceanic and Atmospheric Administration (NOAA) researchers modeled exposure on the basis of meteorologic and ground-station air-monitoring data (Draxler et al. 1994; McQueen and Draxler 1994) with a spatial resolution of 15 km and a temporal resolution of 24 hours (see Chapter 2). DOD personnel records were used to ascertain each study subject’s unit and dates of service. Only Army personnel were included in the study, because their location data were more precise. Two exposure measures were used: cumulative smoke exposure (based on the estimated concentrations on all days when each subject was in the Gulf War Theater) and number of days when the subject was exposed at 65 μg/m3 or greater.

Self-reported oil-well-fire smoke exposure was associated with a higher risk of asthma (OR 1.56, 95% CI 1.23-1.97). In addition, modeled cumulative oil smoke exposure was related to a greater risk of asthma (OR 1.21, 95% CI 0.97-1.51 for the intermediate-exposure group of up to 1.0mg-day/m3; and OR 1.40, 95% CI 1.12-1.76 for the high-exposure group of over 1.0 mg-day/m3) after controlling for sex, age, race or ethnicity, rank, smoking history, and self-

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

reported exposure. When exposure was classified as number of days with exposure at 65 μg/m3 or greater, the risk of asthma also increased. Both exposure metrics showed statistically significant associations with asthma when treated as continuous variables. Smoking appeared to modify the effect: the effect of oil-well-fire smoke exposure was observed among never-smokers and former smokers but not among current smokers. Study strengths include the objective exposure assessment and the use of physician-diagnosed asthma on the basis of clinical evaluations. Limitations include the lack of pulmonary function data and of specified criteria for the diagnosis of asthma and self-selection into the DOD registry, which could have introduced selection bias (for example, if the cohort was enriched with persons who both experienced exposure and had respiratory conditions, the risk estimate could be biased upward). Moreover, the study examined current asthma cases, so a higher incidence of asthma cannot be distinguished from exacerbation or recrudescence of pre-existing disease. The study did not ask about chronic bronchitis or other respiratory effects.

In contrast, the population-based Iowa cohort of 1,560 Gulf War veterans found no statistical association between modeled oil-well-fire exposure and the risk of asthma (Lange et al. 2002). Five years after the war, veterans were asked about their exposures and current symptoms. Exposure was modeled with an approach similar to that of Cowan et al. (2002). Each veteran’s exposure was modeled on the basis of the identified unit and its location during the period of oil-well fires (February-October 1991). Cases of asthma were defined by questions aimed at assessing wheezing and chest tightness. Cases of bronchitis were assessed on the basis of self-reported cough and phlegm production. Both questions pertained to symptoms in the preceding month, so it is not possible to determine whether symptoms were chronic. Self-reported exposure to oil-well fires was associated with a greater risk of asthma and bronchitis. However, there was no statistical association between modeled exposure and the risk of asthma or bronchitis in models that controlled for sex, age, race, military rank, smoking history, military service, and level of preparedness. The correlation between self-reported exposure and modeled exposure was modest (0.40-0.48, p < 0.05). The authors ascribed the different results for self-reported and objective exposure measurement to recall bias. A study strength is the population-based sampling, which implies that findings can be generalized to all military personnel in the Persian Gulf. A limitation is that the study addressed the outcome of asthma symptoms rather than an asthma diagnosis. And, chronic bronchitis was not defined with the standard epidemiologic definition, so it was impossible to distinguish between acute and chronic symptoms.

In the postwar hospitalization study of 405,142 active-duty Gulf War veterans, Smith et al. (2002) examined the effect of oil-well-fire exposure. Exposure was estimated by using troop location data and estimated oil-smoke concentrations based on the same NOAA modeling used in the Cowan and Lange studies. There was no association between exposure to oil-well fires and the risk of hospitalization for asthma (RR 0.90, 95% CI 0.74-1.10), acute bronchitis (RR 1.09, 95% CI 0.62-1.90), or chronic bronchitis (RR 0.78, 95% CI 0.38-1.57). There was modest increase in the relative risk of emphysema from oil-well fire smoke (RR 1.36, 95% CI 0.62-2.98). Because most adults who have asthma or chronic bronchitis are never hospitalized for the condition, the study would not be expected to have captured most cases. No information was available on tobacco-smoking or other exposures that may be related to respiratory symptoms.

Secondary Studies

Several other studies on smoke from oil-well fires in the Persian Gulf are methodologically less robust. A cohort study of Gulf War veterans evaluated self-reported

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

combustion exposure but examined pulmonary symptoms only as a broad class; asthma and bronchitis were not specifically evaluated (Proctor et al. 1998). A prospective study of 125 British royal engineer bomb disposal units stationed in Kuwait City found no change in forced expiratory flow (FEF 25-75%) after the oil-well fires were set, but asthma and bronchitis were not specifically evaluated (Coombe and Drysdale 1993). Finally, two ecologic studies of Kuwaiti residents found no increase in the rate of asthma hospitalization after the Gulf War (Abul et al. 2001; Al-Khalaf 1998).

Exposure to Nerve Agents

A study by Gray and colleagues (1999b) reported on the occurrence of postwar hospitalization from 1991-1995 for respiratory system disease from three levels of estimated exposure to nerve agents due to destruction of munitions at the storage site at Khamisiyah in 1991. Their exposure estimates were developed by DOD in 1997. A small increase in risk was seen only in the comparison of the highest exposed group (0.09657-0.51436 mg-min/m3) with the unexposed group (RR 1.26, 95% CI 1.05-1.51). Limitations of the study include the likely exposure misclassification based on late revised DOD estimates, lack of control for tobacco-smoking, lack of a clear dose-response pattern, and low biologic plausibility of effects on the respiratory system.

As described earlier, Karlinsky et al. (2004) found no associations between pulmonary function measures and exposure to nerve agents at Khamisiyah based on the most recent DOD exposure estimates developed in 2000.

Other Exposures

A study of the UK veteran cohort 7 years after the Gulf War (Cherry et al. 2001a) reported a weak but statistically significant (p < 0.01) effect of exposure to insect repellent on a respiratory-health index based on self-reports.

A cohort study of US Gulf War veterans evaluated associations of self-reported exposures with a pulmonary symptom score; asthma and bronchitis were not specifically evaluated (Proctor et al. 1998). Exposure to smoke from tent heaters and to smoke from burning human waste were significantly associated with pulmonary symptom score (p < 0.001, and p = 0.015, respectively).

Summary and Conclusion

Associations of Respiratory Outcomes with Deployment in the Gulf War Theater

Five studies (Eisen et al. 2005; Gray et al. 1999b; Ishoy et al. 1999b; Karlinsky et al. 2004; Kelsall et al. 2004b) representing four distinct cohorts from three countries (United States, Australia, and Denmark) examined associations of Gulf War deployment with pulmonary function measures or respiratory disease diagnoses based in part on such measures. In none of the studies were such associations found. Each study suffers from various methodologic weaknesses, but the uniformity of findings is striking, especially given that the same studies found Gulf War deployment status to be significantly associated with self-reports of respiratory symptoms among three of the four cohorts. Indeed, the overwhelming majority of studies conducted among Gulf War veterans—whether from the United States (Doebbeling et al. 2000; Gray et al. 1999b; Gray et al. 2002; Iowa Persian Gulf Study Group 1997; Kang et al. 2000; Karlinsky et al. 2004; Kroenke et al. 1998; Petruccelli et al. 1999; Steele 2000), the UK (Cherry

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

et al. 2001b; Nisenbaum et al. 2004; Simmons et al. 2004; Unwin et al. 1999), Canada (Goss Gilroy Inc. 1998), Australia (Kelsall et al. 2004b), or Denmark (Ishoy et al. 1999b)—have found that, several years after deployment, deployed veterans report higher levels of respiratory symptoms and of self-reported respiratory illnesses than nondeployed troops. Of particular interest is the UK cohort study reported in Nisenbaum et al. (2004) and Unwin et al. (1999) which found substantially higher prevalences of respiratory symptoms and self-reports of respiratory disease among those deployed in the Gulf War than among those deployed in another war theater, Bosnia.

Associations of Respiratory Symptoms, Signs, and Illnesses with Specific Exposures Experienced by Gulf War Veterans During Their Deployment

The study of Gulf War veterans of Cowan et al. (2002), which used objective exposure measure and methods, found associations between oil-well-fire smoke and doctor-assigned diagnosis of asthma in veterans. Limitations of the study include the lack of pulmonary function data and of specified criteria for the diagnosis of asthma and the self-selection into the DOD registry. Exposures were well estimated and high but brief, and the exposed population was healthier than in most studies of combustion products. At least aggravation of asthma appears biologically plausible as effect of this exposure, and causation of asthma less certain (the Cowan study did not include questions that would have enabled differentiation of aggravation from causation). The other key Gulf War study of oil-well-fire smoke, based on the Iowa cohort (Lange et al. 2002), which found no relationship between the same objective exposure and respiratory health outcomes, had the advantage of avoiding the potential selection biases of the Cowan et al. study. However, its definitions of respiratory diseases were based entirely on self-reports of symptoms and cannot be viewed as adequate. The study of Smith et al. (2002) found no significant associations between the same objective measures of exposure to smoke from oil-well fires and hospitalization for asthma, acute bronchitis, chronic bronchitis, or emphysema. Limitations of the study include the lack of information on tobacco-smoking, and that most adults in the study age range are seldom hospitalized for those diagnoses which imply that most cases would not be expected to be captured.

The study by Gray and collaborators (1999b) found a small increase in postwar hospitalization for respiratory system disease associated with modeled exposure to nerve agents at Khamisiyah. Limitations of the study include the likely exposure misclassification based on later revised DOD exposure estimates, lack of control for tobacco-smoking, lack of a clear dose-response pattern, and low biologic plausibility of effects on the respiratory system. Karlinsky et al. (2004) found no associations between pulmonary function measures and exposure to nerve agents at Khamisiyah based on the improved DOD exposure estimates developed in 2002; the lack of finding casts further doubt on the validity of the findings of the Gray et al. study.

In conclusion, as is the case for a number of other organ systems, respiratory symptoms and self-reported diseases are strongly associated with Gulf War deployment in most studies addressing this question and used comparison groups of nondeployed veterans. However, the findings of no statistical association of objective pulmonary function measures with Gulf War deployment, in the four cohorts in which this has been investigated, leaves the clinical interpretation of the increased symptoms and self-reported diseases uncertain.

With respect to associations of specific exposures in the Gulf War Theater with pulmonary outcomes, the positive study by Cowan et al. of objective measures of oil-well fire smoke and doctor-assigned respiratory diagnoses is methodologically the strongest to have

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

addressed the question. A well-conceived study that examined associations of pulmonary function measures, and specific-criteria-based physician-diagnosed respiratory diseases and used the same objective measures of exposure to smoke from oil-well fire would be useful. With respect to nerve agents at Khamisiyah, no study that used valid objective estimates of exposure has found statistically significant associations with pulmonary function measures or physician-diagnosed respiratory disease (see Table 5.13 for a summary of respiratory outcomes).

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

TABLE 5.13 Respiratory Outcomes

Study

Design

Population

Outcomes

Results

Adjustments

Comments or Limitations

Eisen et al. 2005

Population-based, cross-sectional, prevalence, medical evaluation

1,061 US deployed vs 1,128 non-deployed

Self-reported asthma, bronchitis, or emphysema; obstructive lung disease (history of disease or symptoms plus use of bronchodilators or 15% improvement in FEV1 after bronchodilator use)

Asthma, bronchitis, or emphysema: OR 1.07 (95% CI 0.65-1.77) Obstructive lungdisease: OR 0.91 (95% CI 0.52-1.59)

Age, sex, race, years of education, cigarette-smoking, duty type, service branch, rank

Low participation rates, especially among nondeployed

Karlinsky et al. 2004

Cross-sectional, medical evaluation

1,036 US deployed vs 1,103 non-deployed

PFT results classified into five categories: normal, nonreversible obstruction, reversible obstruction, restrictive, small-airways obstruction

No association of PFT- based classifications with deployment status, nor with exposure to nerve agents at Khamisiyah based on 2002 DOD exposure models

 

No adjustment for smoking or other confounders; description of sampling strategy inadequate to evaluate bias; no explanation of "matching" or control of matching in analysis

Kelsall et al. 2004b

Cross-sectional, medical evaluation

1,456 Australian deployed vs 1,588 nondeployed

Asthma; bronchitis; FEV1/FVC% <70%

Asthma: OR 1.2 (95% CI 0.8-1.8);

Bronchitis: OR 1.9 (95% CI 1.2-3.1);

FEV1/FVC% <70%: OR 0.8 (95% CI 0.5-1.1); FVC, but not FEV1, associated with self-report of oil-well-fire exposure

Service type, rank, age, education, marital status

Generally well done; substantial potential for selection bias (response rates: deployed 81%, comparison 57%); no use of modeled oil-fire exposures

Gray et al. 1999a

Cross-sectional, medical evaluation

527 Gulf War veterans vs 970 nondeployed from 14 US Navy Seabees commands

Cough; shortness of breath; FVC (L); FEV1(L)

Cough : OR 1.8 (95% CI 1.2-2.8)

Shortness of breath: OR 4.0 (95% CI 2.2-7.3)

Age, height, race, smoking status

No use of modeled oil-fire exposures

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

Study

Design

Population

Outcomes

Results

Adjustments

Comments or Limitations

 

 

 

 

FVC (L): 4.96 vs 4.99, p = 0.77 FEV1(L): 4.05 vs 4.04, p = 0.81

 

 

Ishoy et al. 1999b

Cross-sectional, population-based, medical evaluation

686 peace-keeping Danish deployed to Gulf War Theater vs 231 nondeployed controls

Shortness of breath; FVC; FEV1; peak flow

14% vs 3.5% Percent of predicted: 100.7 vs 100.7, NS 95.6 vs 96.4, NS 94.0 vs 92.8, NS

None

Appropriate population-based controls but differential participation: 84% deployed vs 58% nondeployed; smoking histories similar in deployedand nondeployed

Studies of respiratory outcomes specifically associated with modeled oil-well-fire exposure

Cowan et al. 2002

Case-control study of exposure to smoke from oil-well fires; DOD registry, Army only

873 with asthma vs 2,464 controls

Physician-assigned diagnosis of asthma 3-6 years after war

Self-reported exposure: OR 1.56 (95% CI 1.23-1.97)

Cumulative modeled exposure: OR 1.24 (95% CI 1.00-1.55) for intermediate

cumulative modeled exposure: OR 1. 40 (95% CI 1.11-1.75) for high exposure;

Number of days at > 65μg/m3: OR 1.22(95% CI 0.99-1.51) for 1-5 days; 1.41 (95% CI 1.12-1.77) for 630 days

Sex, age, race, military rank, smoking history, self-reported exposure

Effect seen in former smokers and never-smokers, but not current smokers; key strength: modeled exposure rather than only self-reported exposure; limitations: self-selected population; no specified criteria for asthma diagnosis and no pulmonary function data; pre-exposure asthma status unknown

Lange et al. 2002

Cross-sectional study of exposure to smoke from oil-well fires; derived from cohort study

1,560 Iowa veterans

Asthma symptoms; bronchitis symptom

Fors modeled exposure, adjusted ORs for quartiles of exposure, 0.77-1.26 with no dose-response relationship; for self-reported exposure, asthma ORs 1.77-2.83, bronchitis ORs 2.14-4.78

Sex, age, race, military rank, smoking history, military service, level of preparedness for war

Structured interviews conducted 5 years after war; key strengths: modeled exposure rather than only self-reported exposure, population-based sample; key limitation: symptom-based case definitionof bronchitis and asthma

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

Study

Design

Population

Outcomes

Results

Adjustments

Comments or Limitations

Smith et al. 2002

DOD hospitalizations 1991-1999; exposure modeling for oil- well fire smoke

405,142 active-duty Gulf War veterans

ICD-9 CM codes

Exposed vs non-exposed:

“Influential predictors” of p< 0.15 included in analyses

Objective measure of disease not subject to recall bias; no issues with self- selection; however, only DOD hospitals, only active duty, no information on smoking or other exposures that may be related to respiratory symptoms, most adults with asthma or chronic bronchitis have never been hospitalized for that condition

Asthma

OR 0.90 (95% CI 0.74-1.10)

Acute bronchitis

OR 1.09 (95% CI 0.62-1.90)

Chronic bronchitis

OR 0.78 (95% CI 0.38-1.57)

Emphysema

OR 1.36 (95% CI 0.62-2.98)

Respiratory conditions due to chemical fumes and vapors

OR 0.71 (95% CI 0.23-2.17)

Other respiratory diseases

OR 1.45 (95% CI 0.86-2.46)

Study of respiratory outcomes specifically associated with exposure to nerve agent

Gray et al. 1999b

DOD hospitalizations 1991-1995, exposure to nerve agents at Khamisiyah based on 1997 DOD exposure models

Not exposed (n = 224,804), uncertain low- dose exposure (n = 75,717), exposed (n = 48,770)

Respiratory system disease (vs not exposed):

 

Sex, age group, prewar hospitalization, race, service type, marital status, pay grade, occupation

See Smith et al. 2002 also, probable substantial exposure misclassification as models were revised, lack of a clear dose- response pattern, little biologic plausibility given that no effect was seen for nervous system diseases

Uncertain low dose

OR 0.92 (95% CI (0.85- 0.99)

<0.013 mg-min/ m3

OR 0.90 (95% CI 0.77-1.04)

0. 013- 0.097 mg-min/ m3

OR 0.89 (95% CI 0.79-1.02)

0. 097- 0.514 mg-min/ m3

OR 1.26 (95% CI 1.05-1.51)

Note: DOD = Department of Defense; NS = not significant; PFT = pulmonary function test.

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

DISEASES OF THE DIGESTIVE SYSTEM (ICD-10 K00-K93)

Primary Studies

There were no excess hospitalizations for digestive system disorders, as broadly defined by a range of ICD codes, according to the first postwar hospitalization study (1991-1993) of Gray and colleagues (1996). That study compared hospitalizations of almost 550,000 Gulf War veterans and almost 620,000 nondeployed veterans. The major limitations of the study are its focus on DOD hospitalizations (see earlier discussion) and its inability to capture any but the most severe digestive system disorders (most would be treated on an outpatient basis).

Another hospitalization study conducted by Gray (2000) covered the years 1991-1994 and examined DOD, VA, and California hospitals. The study examined hospitalizations at nonfederal hospitals in California to eliminate potential bias related to veterans seeking care outside DOD and VA facilities. Because of the unreliability of residence data in DOD and VA datasets, the authors could not directly compare rates of hospitalization in the three categories. Therefore, they compared proportional morbidity ratios (PMRs) of hospital-discharge diagnoses (14 diagnostic categories from ICD-9) in Gulf War-deployed and nondeployed veterans. PMRs of most disease categories were not increased in deployed veterans. However, digestive system diseases were increased in VA hospitals (PMR 1.12, 95% CI 1.05-1.18) and in California hospitals (PMR 1.11, 95% CI 0.97-1.24), but not in DOD hospitals. Dyspepsia was one of the conditions studied by Eisen and colleagues (2005), who conducted medical evaluations in phase III of the VA’s nationally representative, population-based study. From 1999-2001, 1,061 deployed and 1,128 nondeployed veterans were evaluated. They had been randomly selected from 11,441 deployed and 9,476 nondeployed veterans, who had participated in the phase I questionnaire, which was used in 1995 (Kang et al. 2000). Researchers were blind to deployment status. The diagnosis of dyspepsia was made by in-person interviews according to a history or symptoms of dyspepsia (frequent heartburn and recurrent abdominal pain) and use of antacids, H2 blockers, or other medications to treat dyspepsia. One study limitation was that deployed veterans were significantly younger, less educated, less likely to be married, and of lower income, although the analysis adjusted for those factors. Another limitation was that, despite three recruitment waves, the participation rate in the 2005 study was low: only 53% of Gulf War veterans and 39% of nondeployed veterans participated. To determine nonparticipation bias, the authors obtained previously collected findings from participants and nonparticipants from the DOD Manpower Data Center and gathered sociodemographic and self-reported health findings from the 1995 VA study (Kang et al. 2000). Both deployed and nondeployed participants were more likely than nonparticipants to report heartburn or indigestion. That could limit the generalizability of findings, but the authors adjusted for the disparity in their analysis of population prevalence. The prevalence of dyspepsia was higher in deployed than in nondeployed veterans: 9.1% vs 6.0% (OR 1.87, 95% CI 1.16-2.99). See Table 5.14 for a summary of the primary studies reviewed above.

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

Secondary Studies

In a study by Sostek and colleagues (1996), veterans from the same unit, deployed and not deployed to the Gulf, were compared for gastrointestinal complaints. Fifty-seven deployed unit members were compared with 44 nondeployed, and a higher prevalence of chronic gastrointestinal symptoms was noted in the deployed group. Especially common were abdominal pain, loose or frequent stools, and excessive gas. Every nongastrointestinal symptom asked about was more common in the deployed group (Sostek et al. 1996).

Ishoy and colleagues (1999a) analyzed self-reported gastrointestinal symptoms in relation to Gulf War exposures among 686 deployed veterans and 257 controls who were not deployed to the gulf. The groups were matched for age, sex, and profession within the military. The study was a further analysis of Danish Gulf War veterans and matched controls (Ishoy et al. 1999b). The earlier study found that on a questionnaire, eight of 14 gastrointestinal symptoms were reported significantly more frequently by veterans than by controls. After adjustment for the interrelationship of variables, only two of the eight gastrointestinal symptoms remained significant: prevalence of recurrent diarrhea for one year and rumbling in the stomach more than two times per week. The later study used both symptoms as the combined main outcome measure and investigated its relationship to 24 environmental exposures. Fifteen of the 24 exposures were significantly associated (p < 0.05) with that combined measure. After multivariate adjustment, only two were significantly associated with the outcome, exposure to insecticides against cockroaches (OR 2.3, 95% CI 1.2-4.4) and the burning of waste or manure (OR 2.5, 95% CI 1.3-5.0). Two others were nearly significant (p < 0.10): teeth-brushing with water contaminated with chemicals or pesticides and bathing in or drinking contaminated water. When those four significant or nearly significant exposures were examined in relation to the main outcome, a dose-response trend was found: the prevalence of the main outcome was greater after three or four exposures (18.9%) than after one or two exposures (7.4 % and 12.8%, respectively). In separate analyses of population attributable risk, 85% of the main outcome could be attributed to either environmental exposures or neuropsychologic symptoms. The limitations of the study are the use of self-reported symptoms and exposures.

Summary and Conclusion

There were many reports of gastrointestinal disturbances in Gulf War-deployed veterans. The disturbances seem to be linked to contaminated water and burning of animal waste. The committee notes that several studies reported an increase in the rate of self-reported dyspepsia. There appears to be a higher prevalence of dyspepsia in deployed Gulf War veterans than in nondeployed veterans. See Table 5.14 for a summary of the primary studies reviewed above.

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

TABLE 5.14 Gastrointestinal Outcomes

Reference

Design

Population

Outcomes

Results

Adjustments

Comments or Limitations

Eisen et al. 2005

Population-based, cross-sectional, prevalence, medical evaluation

1,061 US deployed and 1,128 nondeployed

Dyspepsia

Deployed: 9.1%; nondeployed: 6.0%; OR 1.87 (95% CI 1.16- 2.99)

Age, sex, race, years of education, smoking, duty type, service branch, rank

Low participation rates, especially among nondeployed

Gray et al. 2000

Retrospective cohort study (hospitalization records)

Gulf War-deployed (August 1990-July 1992, n = 652, 979) and Gulf War nondeployed (n = 652,922) stratified by California residence, service, and service branch of allnon deployed veterans (n = 2,912,737)

Digestive system diseases

VA hospitals: PMR 1.12 (95% CI 1.05-1.18);

DOD hospitals: PMR 0.98 (95% CI 0.96- 0.99);

California hospitals: PMR 1.11 (95% CI 0.97-1.24)

Hospitalization records were matched on sex, age

Findings might be influenced by chance or by potential confounders, including health-registry participation

Gray et al. 1996

Retrospective cohort study (hospitalization records)

DOD hospitals: 547,076 Gulf War veterans; 618,335 era veterans who did not serve in Gulf War

Digestive system diseases

All ORs < 1.0

Hospitalization rates and rate ratios adjusted for age, sex; multiple logistic-regression models adjusted for all observed demographic differences between groups

Study data reflect only hospitalization experience of persons who remained on active duty through September 1993

NOTE: DOD = Department of Defense; PMR = proportional mortality ratio; VA = Department of Veterans Affairs.

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

DISEASES OF THE SKIN AND SUBCUTANEOUS TISSUE (ICD-10 L00-L99)

Primary Studies

For dermatologic outcomes, the committee defined a primary study as one that used a dermatologic examination, whereas secondary studies were based on self-reports. In the most nationally representative US study, Eisen and colleagues (2005) divided dermatologic conditions into two categories. Group 1 consisted of many common conditions (for example, freckles, seborrheic keratoses, moles, cherry hemangiomas, skin tags, and surgical scars), and Group 2 consisted of more unusual diagnoses not included in Group 1. Diagnoses were determined by a board-certified dermatologist, who evaluated the Group 2 conditions through teledermatology, at least two digital photographs, and the results of a standardized history and physical examination. In adjusted analyses, two Group 2 skin conditions were diagnosed more frequently among deployed than among nondeployed veterans: atopic dermatitis (1.2% vs 0.3%, OR 8.1, 95% CI 2.4-27.7) and verruca vulgaris (warts) (1.6% vs 0.6%, OR 4.02, 95% CI 1.28-12.6).

UK researchers performed dermatologic evaluations on UK Gulf War veterans (111 disabled and 98 nondisabled) and 133 disabled non-Gulf War veterans; disability was defined as reduced physical functioning. They found no differences among groups in any dermatologic conditions other than seborrheic dermatitis (8.1% in deployed vs 2.3% in nondeployed), which was more common in Gulf War veterans, irrespective of disability status (Higgins et al. 2002). See Table 5.15 for a summary of the primary papers that the committee considered for dermatologic outcomes.

Secondary Studies

Two other studies found higher prevalences of dermatologic conditions. Kang et al. (2000), using a stratified random-sampling method, compared 693,826 Gulf War-deployed veterans with 800,680 non-Gulf War veterans from the DOD Defense Manpower Data Center. They found increases in self-reported eczema or psoriasis (7.7% vs 4.4%; 95% of rate difference CI 3.26-3.42) and other dermatitis (25.1% vs 12.0%; 95% CI of rate difference 13.04-13.28). Proctor et al. (1998) looked at 252 Gulf War-deployed veterans—186 from the Fort Devens cohort and 66 from the New Orleans cohort—and compared them with 48 veterans deployed to Germany during the Gulf War. They reported on increased prevalence (15.5% vs 11.7% vs 1.9%) of dermatologic conditions such as rashes, eczema, and skin allergies. However, both studies relied primarily on self-reports or questionnaires and are thus vulnerable to recall or reporting bias.

Summary and Conclusion

On the basis of the few studies of dermatologic conditions, unrelated skin conditions occur more frequently among Gulf War-deployed veterans, but the findings are not consistent among the studies. The committee notes that there is some evidence in Gulf War-deployed veterans of a higher prevalence of two distinct dermatologic conditions—atopic dermatitis and warts. See Table 5.15 for a summary of the primary papers that the committee considered for dermatologic outcomes.

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

TABLE 5.15 Dermatologic Outcomes

Reference

Design

Population

Outcomes

Results

Adjustments

Limitations/Comments

Eisen et al. 2005

Population-based, cross-sectional, prevalence, medical evaluation

1,061 US deployed and 1,128 nondeployed

Atopic dermatitis and verruca vulgaris (warts)

Atopic dermatitis: 1.2% vs 0.3%, OR 8.1 (95% CI 2.4-27.7); verruca vulgaris (warts): 1.6% vs 0.6%, OR 4.02 (95% CI 1.28-12.6)

Age, sex, race, years of education, smoking, duty type, service branch, rank

Low participation rates, especially among nondeployed

Higgins et al. 2002

Prospective case-comparison study

Disabled Gulf War veterans (n = 111), nondisabled Gulf War veterans (n = 98), and 133 disabled non-Gulf War veterans

Seborrheic dematitis

8.1% in deployed vs 2.3% in nondeployed

Socioeconomic status, military rank, current service history, smoking, alcohol use

 

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

DISEASES OF THE MUSCULOSKELETAL SYSTEM AND CONNECTIVE TISSUE (ICD-10 M00-M99)

Arthritis and Arthralgia

Arthritis is the most common form of joint disease. Several powerful risk factors are major trauma, repetitive joint use, and age. Arthritis is diagnosed according to a combination of clinical features and radiographic findings. Arthralgia, which is a self-reported symptom of arthritis, refers to painful joints. In the absence of other clinical features and radiographic findings, arthralgias are not necessarily diagnostic of arthritis.

Primary Studies

Arthralgias were one of 12 primary health outcome measures studied by Eisen and colleagues (2005). They conducted medical evaluations in phase III of VA’s nationally representative, population-based study. From 1999-2001, 1,061 deployed and 1,128 nondeployed veterans were evaluated. They had been randomly selected from 11,441 deployed and 9,476 nondeployed veterans who had participated in the phase I questionnaire in 1995 (Kang et al. 2000). Researchers were blinded to deployment status. Arthralgias were defined as persistent and clinically significant bone or joint symptoms with or without joint effusion, and treatment with anti-inflammatory agents, narcotic pain medications, or nonnarcotic pain medications. There was no statistically significant difference in arthralgias between deployed and nondeployed veterans (OR 1.15, 95% CI 0.70-1.89).

One study limitation was that despite three recruitment waves, the participation rate in the 2005 study was low: only 53% of Gulf War veterans and 39% of nondeployed veterans participated. To determine nonparticipation bias, the authors obtained previously collected findings on participants and nonparticipants from the DOD Manpower Data Center and gathered sociodemographic and self-reported health findings from the 1995 VA study (Kang et al. 2000). Both deployed and nondeployed participants were more likely than nonparticipants to report arthritis of any kind. See Table 5.16 for a summary of the studies reviewed in this section.

Secondary Studies

Two other studies examined differences in prevalence of arthritis, but they relied on self-reporting. Kang et al. (2000), using a stratified random-sampling method, compared data from the DOD Defense Manpower Data Center on 693,826 Gulf War deployed veterans and 800,680 non-Gulf War veterans, and asked about arthritis as a self-reported condition. They found a significant difference in such reporting between Gulf War-deployed and non-Gulf War deployed veterans (22.5% vs 16.7%, rate difference of 5.87, 95% CI 5.74-6.00). Gray et al. (2002) looked at 3,831 Gulf War deployed veterans, 4,933 veterans deployed elsewhere, and 3,104 nondeployed Seabees. The authors found an increase in reporting of arthritis among Gulf War than among Seabees deployed elsewhere (5.87% vs 4.42%). The latter, in turn, were similar to other nondeployed Seabees (4.42% vs 4.38%). The OR for Gulf War veterans vs veterans deployed elsewhere was 1.44 (95% CI 1.17-1.76), and that for Gulf War-deployed vs non-deployed veterans was 1.63 (95% CI 1.29-2.08).

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

Summary and Conclusion

Among those examined, there was no statistically significant difference in arthralgias, a surrogate for arthritis, but data on self-reports indicate that arthritis was more common among those deployed to the gulf. The data suffer, however, from the problem of self reporting of a common condition that can be easily confused with other symptoms. There appears to be no statistically significant increase in the prevalence of arthralgias among veterans who underwent a medical examination.

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

TABLE 5.16 Arthralgia

Study

Design

Population

Outcomes

Results

Adjustments

Comments or Limitations

Eisen et al. 2005

Population-based, cross-sectional, prevalence, medical evaluation

1,061 deployed vs 1,128 nondeployed

Persistent and clinically significant bone or joint symptoms with or without joint effusion, and treatment with anti-inflammatory agents, narcotic pain medications, ornonnarcotic pain medications

Prevalence: 6.4% vs. 6.8%, OR 1.15 (95% CI 0.70-1.89)

Age, sex, race, yearsof education, cigarette-smoking, duty type, service branch, rank

Low participation rates, especially among nondeployed

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

FIBROMYALGIA

The hallmarks of fibromyalgia are widespread muscle and skeletal pain and tenderness at numerous soft tissue sites on the body upon palpation, according to classification criteria promulgated by the American College of Rheumatology (ACR) (Wolfe et al. 1990). The case definition requires both widespread pain (pain on both sides of the body, above and below the waist, and including axial skeletal pain) lasting for at least 3 months and pain (not just tenderness) in at least 11 of 18 tender point sites on palpation with an approximate force of 4 kg. The presence of a second clinical disorder does not exclude a diagnosis of fibromyalgia. Other symptoms of fibromyalgia include fatigue, sleep disturbance, morning stiffness, and cognitive impairment, but those are not sensitive and specific enough to use for classification (Wolfe et al. 1990). Early characterization of the condition as an inflammation of muscle (hence the label fibrositis) have not been borne out through research (Goldenberg 1999). There is no pathologic or laboratory test with which to confirm the diagnosis. And there are no widely accepted causative factors. Fibromyalgia’s prevalence in the general population is about 3.4% in women, and 0.5% in men, so it is one of the more common rheumatologic disorders (Wolfe et al. 1995). Its prevalence increases with age (Wolfe et al. 1995). On the basis of longitudinal studies, the course is chronic but variable in intensity (Wolfe et al. 1997). It should be noted that the existence of fibromyalgia as a distinct disease entity is considered controversial by some expert commentators (Nimnuan et al. 2001; Pearce 2004).

Primary Studies

For consideration as a primary study, the basis of diagnosis of fibromyalgia has to include symptom reporting and physical examination, rather than only symptom-based criteria. Fibromyalgia was one of 12 primary health-outcome measures studied by Eisen and colleagues (2005), who conducted medical evaluations in phase III of VA’s nationally representative, population-based study. From 1999- 2001, 1,061 Gulf War veterans and 1,128 non-Gulf War veterans were evaluated. They had been randomly selected from 11,441 deployed and 9,476 nondeployed veterans, who had participated in the phase I questionnaire in 1995 (Kang et al. 2000). Researchers were blinded to deployment status. The diagnosis of fibromyalgia was based on diffuse body pain and pain on physical examination, following the ACR criteria (Wolfe et al. 1990). Self-reported diagnoses of fibrositis or fibromyalgia did not vary between deployed and nondeployed veterans (0.6% and 0.8% respectively; OR 1.21, 95% CI 0.36-4.10, adjusted for age, sex, race, cigarette-smoking, duty type, service branch, and rank). However, fibromyalgia diagnosed on the basis of physical examination was present in 2.0% of deployed and 1.2% of nondeployed veterans (adjusted OR 2.32, 95% CI 1.02- 5.27). Strengths of the study include the population-based sampling strategy, blinding of evaluating physicians, and use of validated diagnostic criteria based on physical examination. Limitations include the potential for substantial selection bias due to modest participation rates of 53% of Gulf War veterans and 39% of non-Gulf War veterans and the deployed veterans being significantly younger, less educated, less likely to be married, and of lower income, although the analysis adjusted for most of those factors.

Smith and colleagues (Smith et al. 2000) performed a study of postwar hospitalizations (1991-1997) among 551,841 deployed and 1,478,704 nondeployed active duty personnel. The

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

association of hospitalization for fibromyalgia (ICD-9 code 729) with deployment status was investigated with Cox proportional-hazards models. The study found higher risk of fibromyalgia hospitalization among the deployed (RR 1.23, 95% CI 1.05-1.43). However, survival curves showed that the higher rate of hospitalization for fibromyalgia among the deployed occurred only between the inception of the Comprehensive Clinical Evaluation Program (CCEP) on June 30, 1994, through the middle of 1995. During that roughly 1-year period, many CCEP participants were admitted to the hospital only for purposes of evaluation. In fact, CCEP participants had more than 26 times the risk of being hospitalized for fibromyalgia than did nonparticipants. For the 3-year period before the inception of the CCEP, hospitalization for fibromyalgia was unrelated to Gulf War status (RR 0.92, 95% CI 0.74-1.13). The Smith et al. study has the advantage of a large, population-based sample and good statistical power for the detection of an effect. Its major limitations are the inclusion of only active-duty personnel, changes in hospitalization rates for fibromyalgia in association with the practices of the CCEP, and the fact that few cases of fibromyalgia are severe enough to warrant hospitalization. The findings on fibromyalgia are summarized in Table 5.17.

Secondary Studies

The Iowa study (Iowa Persian Gulf Study Group 1997) surveyed 1,896 deployed and 1,799 non-deployed veterans who listed Iowa as their home state at the time of enlistment. Fibromyalgia was assessed according to the symptom criteria of Wolfe and colleagues (Wolfe et al. 1995). Those criteria include the presence of widespread pain for at least 3 months. No physical examinations were conducted. Symptoms of fibromyalgia were present in about 21% of deployed veterans and 11% of nondeployed veterans. The authors found a prevalence difference of 9.3% (95% CI 7.3-11.2) after adjustment for age, sex, race, branch of military, and rank. The main strength of the study was the population-based sample. The main study limitation was “diagnosis” of fibromyalgia based on reported symptoms without physical examination to assess pain on digital palpation.

Canada deployed more than 3,000 troops to the Gulf region. A survey of the entire cohort of Canadian Gulf War-deployed veterans found that they were more likely than nondeployed veterans—group-matched to cases on sex, age, and regular vs reserve status—to report symptoms of fibromyalgia (Goss Gilroy Inc. 1998). The criteria for fibromyalgia, adapted from several published studies (e.g., Wolfe et al. 1995), were having “overall” body pain lasting 3 months or longer and having pain at a level of 1 or higher (on a scale of 1-10) over the preceding 24 hours or self-report of fibromyalgia or fibrositis. Symptom-defined fibromyalgia was present in about 16% of deployed veterans and 10% of nondeployed veterans (adjusted OR 1.81, 95% CI 1.55-2.13). The main strength of the study was the population-based sample. The main study limitation was diagnosis of fibromyalgia based on symptoms or self-reported diagnosis without physical examination to assess pain on digital palpation.

Bourdette et al. (2001) studied 244 Oregon and Washington Gulf War veterans who had unexplained illness after clinical evaluation to exclude “explainable” illness. Fifty (20.8%) fulfilled the ACR criteria for fibromyalgia (although it is not stated, presumably these included clinical examination to see whether pain was present at at least 11 of 18 body points on digital palpation). Those 50 represent 2.5% of 2022 Gulf War veterans solicited for participation (of the 2,022 solicited, 1,760 were located, 1,119 responded, and 799 were deemed eligible for the clinical study). The study's main limitations are its lack of a nondeployed comparison group and lack of clarity about the nature of the clinical examination for fibromyalgia.

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

Summary and Conclusion

The diagnosis of fibromyalgia is based entirely on symptoms and physical examination; there are no pathologic or laboratory tests with which to confirm it. Among the available cross-sectional studies that include both Gulf War-deployed and non-deployed veterans, only Eisen and colleagues (2005) used the full ACR case definition of fibromyalgia, including criteria based on physical examination. Fibromyalgia was diagnosed in 2.0% of deployed and 1.2% of nondeployed veterans, for an adjusted OR of 2.32 (95% CI 1.02-5.27). A strength of this study is the population-based sampling strategy. An important limitation is the modest participation rates (53% of Gulf War veterans and 39% of non-Gulf War veterans) with the potential to introduce selection bias. After accounting for a 1-year period during which many CCEP participants were admitted to the hospital only for purposes of evaluation, the study by Smith and colleagues (2000) found no association between Gulf War deployment and hospitalization for fibromyalgia. That finding does not appear inconsistent with positive findings in the Eisen et al. study, in that few cases of fibromyalgia are severe enough to warrant hospitalization. Notably, the prevalence of a diagnosis of fibromyalgia in the Eisen et al. study is about 300 times the prevalence of hospitalization for fibromyalgia in the Smith et al. study. The Iowa study and the Canadian study both found significantly increased fibromyalgia symptoms among deployed veterans compared with nondeployed veterans. The findings of those two studies, although generally supportive of the findings of the Eisen et al. study, are of limited value owing to the lack of a physical examination to enable the use of the full criteria for diagnosis. The Bourdette et al. study, which had no nondeployed comparison group, estimated that at least 2.5% met the full ACR case definition of fibromyalgia. In conclusion, largely on the basis of the Eisen et al. study, which used the criteria of the ACR for diagnosis of fibromyalgia but could have been subject to unrecognized selection bias, there is a higher prevalence of fibromyalgia among deployed Gulf War veterans than among nondeployed veterans.

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

TABLE 5.17 Fibromyalgia

Study

Design

Population

Outcomes

Results

Adjustments

Comments or Limitations

Eisen et al. 2005

Population-based, cross-sectional, prevalence, medical evaluation

1,061 deployed, 1,128 nondeployed

Symptoms and physical examination using criteria of American College of Rheumatology (Wolfe et al. 1990)

Prevalence: 2.0% vs 1.2%, OR 2.32 (95% CI 1.02-5.27)

Age, sex, race, years of education, cigarette smoking, duty type, service branch, rank

Uses gold standard for diagnosis of fibromyalgia; low participation rates, especially among nondeployed

Smith et al. 2000

Postwar hospitalization study

551,841 deployed, 1,478,704 nondeployed

Hospitalization (1991-1997); Cox proportional-hazards models ICD-9codes for fibromyalgia (729.1)

RR 1.23 (95% CI 1.05-1.43); however, survival curves indicate excess due to hospitalization only for purposes of evaluation during the CCEP; before CCEP: RR 0.92 (95% CI 0.74-1.13)

Sex, age, branch of service

No increase after accounting for CCEP effect; limited to active duty; most cases offibromyalgia are not severe enough to warrant hospitalization

NOTE: CCEP = Comprehensive Clinical Evaluation Program.

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

BIRTH DEFECTS AND ADVERSE PREGNANCY OUTCOMES (ICD-10 O00-Q99)

This section evaluates the findings on birth defects in the offspring of veterans, on adverse reproductive outcomes, on the risk of male infertility, and on sexual problems. As appropriate, the major results from each study are organized by whether the father or the mother served in the gulf and by outcome. Table 5.18 summarizes all the primary studies on birth defects and adverse reproductive outcomes reviewed by the committee.

Birth Defects

Birth defects occur in about 3% of live births. The numerous types of serious or disabling birth defects include structural defects, chromosomal abnormalities, and birth defect syndromes (California Birth Defects Monitoring Program 2006). Because of that diversity, epidemiologists attempting to calculate whether birth defects are increased in a particular group such as deployed veterans, encounter the problem of making multiple comparisons; that is, the greater the number or the more types of comparisons, the greater the likelihood that one or more of them will appear statistically significant when no true differences exists. Several statistical techniques are used to adjust for, or minimize, the problem of multiple comparisons but they are not foolproof.

Primary Studies

In the most comprehensive population-based study, Araneta and colleagues (2003) identified birth defects among infants of military personnel born from January 1, 1989 to December 31, 1993, from population-based birth defect registries in six states: Arizona, Hawaii, Iowa, and selected counties of Arkansas, California, and Georgia (metropolitan Atlanta). They compared the prevalence of 48 selected congenital anomalies diagnosed from birth to the age of 1 year between Gulf War veterans’ and nondeployed veterans’ infants conceived before the war; between Gulf War veterans’ and nondeployed veterans’ infants conceived during or after the war; and between Gulf War veterans’ infants conceived before and after the war. The authors performed separate analyses on the basis of whether the mother or the father was engaged in military service. If both parents were in the military then the birth was categorized as an infant of a military mother. The study found higher prevalence of three cardiac defects (tricuspid valve insufficiency, aortic valve stenosis, and coarctation of the aorta), and one kidney defect (renal agenesis and hypoplasia) among infants conceived after the war to Gulf War veteran fathers. There also was a higher prevalence of hypospadias (malformation of the urethra and urethral groove), a genitourinary defect among sons conceived postwar to Gulf War veteran mothers compared to their nondeployed counterparts. Aortic valve stenosis, coarctation of aorta, and renal angenesis and hypoplasia were also elevated among infants conceived among the Gulf War veteran fathers postwar compared to those conceived prewar. There was only 1 birth defect recorded among 142 births conceived prewar to Gulf War veteran mothers, and this precludes comparisons with this group.

This study is particularly informative because it relies on medically confirmed outcomes diagnosed through the first year, rather than at birth, and uses information from population registries, as opposed to information from voluntary participation by study subjects. Because both nonmilitary and military hospitals participated in the registries in all states except California

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

(nonmilitary only), births among reservists, National Guard, and former military personnel were eligible, as well as among those on active duty. The study also included comparisons of births to Gulf War veterans before and after deployment. One limitation is that the study relied on availability of unique personal identifiers in military and birth certificate data, which leads to the possibility that some military offspring might be missed among the cases, and that would make the observed prevalence more conservative than the actual. Another is the study’s low power to assess individual defects that are rare. The authors also published the results of the pilot study of their method, which was performed in only Hawaii (Araneta et al. 2000); because the data are incorporated in the larger six-state study, we did not review them separately here.

Secondary Studies

Additional studies of birth defects are considered secondary either because they rely on self-reports (and thus introduce potential recall bias) or because they consider only groups of birth defects. Studying groups of birth defects, although useful in identifying patterns, makes it difficult to determine which specific defects may be increased. Doyle and colleagues (2004) evaluated the prevalence of self-reported birth defects among the offspring of all UK veterans (male and female) deployed to the gulf and among the offspring of nondeployed veterans who responded to a postal questionnaire. Response rates were higher among the Gulf War veterans (53% of men, 72% of women) than the comparison group (42% of men, 60% of women). They considered pregnancies conceived after deployment (after January 1, 1991, for nondeployed veterans) through November 8, 1997. Medical confirmation was requested for all fetal deaths at 16 weeks or more or of unknown gestation and for liveborn children in whom a congenital abnormality, serious childhood medical condition, or death was reported. Among infants conceived by fathers deployed to the gulf compared with infants of fathers not deployed, the OR for any malformation was 1.5 (95% CI 1.3-1.7). Elevated risks were observed specifically for malformations of the genital system, urinary system, musculoskeletal system, and cranial neural crest; for “other” malformations of the digestive system; for “other” non-chromosomal malformations; and for metabolic and single-gene defects. The risks of urinary system and musculoskeletal system defects remained increased when the cases were restricted to the 55% that had clinical confirmation. No statistically significant increased risk of birth defects in infants of mothers deployed to the gulf was found.

Three additional secondary studies assessed groups of birth defects. In a population-based survey in the United States, Kang et al. (2001) observed excess risks of self-reported “likely birth defects” and specifically “moderate to severe defects” among infants of Gulf War-deployed fathers and mothers compared with nondeployed fathers and mothers. Most defects were isolated anomalies, and no clear patterns were found. First pregnancies ending after June 30, 1991, were considered in this analysis. Another US study that included live births at 135 military hospitals from 1991-1993 did not find evidence of a statistically significant increased risk of “any birth defects” or “severe birth defects” in infants born to fathers or mothers deployed to the gulf (Cowan et al. 1997). However, because this study included only births at military hospitals, only parents on active duty at time of the birth were included, and it is likely that higher risk pregnancies were referred to civilian hospitals. The observed number of birth defects among children (liveborn and stillborn) born after deployment to National Guard personnel in two units in southeast Mississippi was not greater than expected on the basis of population-based registries (Penman et al. 1996).

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

A population-based study of male Canadian veterans (Goss Gilroy Inc. 1998) surveyed deployed and nondeployed veterans for self-reported birth defects. Overall, deployed veterans reported higher rates of birth defects (a combined category that includes births before, during, and after the Gulf War). Birth defects that occurred at similar rates include urogenital and kidney defects.

In a study of Kuwaiti nationals, a higher prevalence of congenital heart disease (CHD) was observed in babies born after invasion than before invasion (Abushaban et al. 2004). However, no data were available on the residence status of the parents during the war (inside or outside Kuwait), or of their proximity to potential environmental pollutants, for example, oil-well fires. It is also possible that there was underreporting of CHD before to the war. This study is considered secondary because the study population is not Gulf War veterans.

Goldenhar Syndrome

Anecdotal reports raised the possibility of increased prevalence of Goldenhar syndrome, a rare craniofacial abnormality, among children of Gulf War veterans. External features of the syndrome are ear abnormalities, such as microtia, anotia, and preauricular tags. Among infants conceived after the Gulf War (or December 31, 1990 for nondeployed veterans) through September 30, 2003 and born to active-duty personnel in military hospitals, five cases with Gulf War veteran fathers and two cases with NDV fathers were identified (Araneta et al. 1997). Given those small numbers, it is difficult to determine whether an excess risk is associated with service in the gulf. In a case-control study of hemifacial microsomia (of which Goldenhar syndrome is one type) in craniofacial clinics in 24 US cities, the OR of 2.4 for parental army service in general (95% CI 1.4-4.2) was statistically significant. The association with Gulf War army service in particular (OR 2.8, 95% CI 0.8-9.6) was increased but not statistically significant (Werler et al. 2005).

Summary and Conclusion

Primarily on the basis of the Araneta et al. and Doyle et al. studies, because of the availability of medical confirmation in those studies, there is some evidence of increased risk of birth defects among offspring of Gulf War veterans. However, with the possible exception of urinary tract abnormalities, the specific defects with increased prevalence in the two studies were not consistent. The reported association of Gulf War service with Goldenhar syndrome was inconclusive. Overall, the studies are difficult to interpret because specific birth defects are relatively rare, multiple comparisons were performed, and sample sizes were small when divided by timing of exposure (before or after conception) and whether the mother or the father was exposed.

Thus the committee concludes that there is no consistent pattern of higher prevalence of birth defects among offspring of male or female Gulf War veterans and that no single defect, except urinary tract abnormalities, has been found in more than one well-designed study.

Adverse Pregnancy Outcomes

Primary Study

The prevalence of spontaneous abortions, stillbirths, and ectopic pregnancies has been studied in deployed and nondeployed women. Araneta and colleagues (2004) recruited women

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

admitted to military hospitals for pregnancy-related diagnoses (including livebirths, spontaneous and induced abortions, ectopic pregnancies, and pregnancy-related complications) from August 2, 1990-May 31, 1992. Among women who conceived after the war, the risk of spontaneous abortions and ectopic pregnancies was higher among deployed than among nondeployed women. The risk of those outcomes among so called “Gulf War exposed” conceptions was increased, but not statistically significantly. Self-reported outcomes were confirmed by hospital-discharge data. Because only military hospitals were included, only information on active-duty personnel was included.

Secondary Studies

Doyle and colleagues (2004) also studied the risk of self-reported miscarriages and stillbirths among Gulf War-deployed fathers and mothers. They observed no effect of Gulf War service on the risk of stillbirths or of miscarriages in pregnancies reported by female veterans. There was a 40% increase in the risk of miscarriages among pregnancies reported by male Gulf War veterans compared with their nondeployed counterparts (95% CI 30%-50%), and the effect was stronger for early miscarriages (OR 1.5, 95% CI 1.3-1.6). However, in a study of nuclear industry workers (the Nuclear Industry Family Study) there was evidence of underreporting of miscarriages among the nonexposed workers. The potential selection bias could explain the increased risks observed among the Gulf War veterans.

In the Kang et al. (2001) study described above, there was an excess prevalence of self-reported spontaneous abortions and stillbirths among pregnancies conceived by Gulf War veteran fathers and of spontaneous abortions among pregnancies conceived by Gulf War veteran mothers. As that study relied on self-reports and inconsisten participation rates, the results are difficult to interpret.

Gray and colleagues (1996) conducted a hospitalization study (1991-1993) in which they compared almost 550,000 Gulf War veterans with almost 620,000 nondeployed veterans. The study found increased hospitalizations, in 1991 only, for the broad category “genitourinary system diseases”. More specifically, the increase was due to female veterans’ being hospitalized for inflammatory diseases of the ovary, fallopian tube, pelvic cellular tissue, and peritoneum. Male Gulf War veterans were hospitalized for redundant prepuce and phimosis (ICD code 605). This code often accompanies hospitalizations for circumcision. The major limitation of this study is its focus on DOD hospitalizations.

Summary and Conclusion

Although the results from the Araneta et al. study, which had hospital-discharge data available, are suggestive of an increased risk of spontaneous abortions and ectopic pregnancies, the results may not be generalizable to deployed women who left the service or to pregnancy-related admissions to nonmilitary hospitals. Thus, it is difficult to conclude whether there is a higher prevalence of adverse pregnancy outcomes in Gulf War-deployed than in nondeployed veterans.

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

Male Fertility Problems and Infertility

Primary Studies

Two studies have addressed fertility problems among men who served in the Gulf War. A group of UK Gulf War veterans (drawn from the same population as the Doyle et al. study described above) who fathered or tried to father children after the war and before August 1997 reported excess risk of infertility (defined as consulting a doctor after trying unsuccessfully for more than 1 year) compared to their nondeployed counterparts. The risks of type 1 infertility (never achieving pregnancy) and type II infertility (never achieving a live birth) were also significantly higher. Furthermore, more Gulf War veterans than non-Gulf War veterans experienced time to conception for planned pregnancies of more than 1 year (Maconochie et al. 2004). Those results are difficult to interpret because of low response rates, possible recall bias, and lack of information on partners’ fertility status.

Ishoy and colleagues evaluated serum concentrations of reproductive hormones in a study of 661 Danish Gulf War veterans and 215 nondeployed veterans. There were no statistically significant differences in concentrations of luteinizing hormone, follicle stimulating hormone, testosterone, or inhibin B between the two groups (Ishoy et al. 2001a). There was a higher prevalence of reporting of sexual problems, specifically decreased libido, in the deployed veterans (Ishoy et al. 2001b). Sexual problems were associated with self-reported combat-related exposures, such as being threatened with arms or witnessing killing or wounding of victims and colleagues.

Summary and Conclusion

For the most part, the findings on fertility and sexual problems relied on self-reports which entail a substantial opportunity for recall bias. There was no evidence of statistically significant differences in concentrations of male reproductive hormones between Gulf War veterans and nondeployed veterans.

Although it appears that there is no difference in the prevalence of male fertility problems or infertility between the deployed Gulf War veterans and their nondeployed counterparts, it is difficult to draw conclusions from the small number of available studies.

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

TABLE 5.18 Birth Defects and Adverse Reproductive Outcomes

Study

Design

Population

Outcomes

Results

Adjustments

Comments or Limitations

Birth defects

Araneta et al. 2003

Prevalence, population-based, birth-defect registry (active surveillance all cases identified from birth to 1 year)

Infants of military personnel born 1/1/1989-12/31/1993 in Arizona, Iowa, Hawaii, and participating counties of Arkansas, California, Georgia to GWV mothers (n = 450), NDV mothers (n = 3,966), GWV fathers (n = 11, 511), NDV fathers (n = 29, 086)

48 birth defects identified by CDC as occurring frequently or of public health importance, excluding pulmonary artery anomalies and adding dextrocardia, chromosomal anomalies (other than trisomies 13, 18, and 21), and Goldenhar syndrome

Unadjusted RRs: Postwar conceptions, GWVs vs NDVs:

father: tricuspid valve insufficiency, 10/4648 vs 9/11,164 (RR 2.7, 95% CI 1.1-6.6); aortic valve stenosis, 5/4,648 vs 2/11,164 (RR 6.0, 95% CI 1.2- 31.0); coarctation of aorta, 5/ 4,648 vs 3/11,164 (RR 4.0, 95% CI 0.96-16.8); renal agenesis or hypoplasia, 5/4,648 vs 5/11,164 (RR 2.4, 95% CI 0.7-8.3)

mother: hypospadias 4/154 vs 4/967 (RR 6.3, 95% CI 1.5-26.3)

GWVs postwar vs prewar conceptions

father: aortic valve stenosis 5/4,648 vs 0/6,863 (RR 16.3, 95% CI 0.9-294); coarctation of aorta, 5/4,648 vs 1/6,863 (RR 7.4, 95% CI 0.9- 63.3); renal agenesis and hypoplasia, 5/4,648 vs 0/ 6,863 (RR 16.3, (95% CI 0.9-294); adjustment did not change results

State, maternal and paternal age, race, marital status, education, plurality, parity, prenatal visits, gestational weight gain, branch of service, military rank, prenatal alcohol exposure, intrauterine growth retardation, low birth weight, small for gestational age, pre-eclampsia

Limitations: California limited to diagnoses in nonmilitary hospitals; relies on availability of unique personal identifiers in military and birth certificate data limited power to assess individual defects, multiple comparisons, limited to live births

Strengths: population-based, including reservists, National Guard, former military personnel; includes defects diagnosed through first year, medically confirmed as opposed to self-reports, comparisons with prewar experience

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

Study

Design

Population

Outcomes

Results

Adjustments

Comments or Limitations

Doyle et al. 2004

Prevalence

All UK GWVs and randomly selected cohort of NDVs responding to postal questionnaire; conceptions from postdeployment (for NDVs – conceived after 1/1/1991) through 11/8/1997 GWV fathers (n = 16,442) NDV fathers (n = 11,517) GWV mothers (n = 484) NDV mothers (n = 377)

External comparison populations: (1) NIFS; (2) annual registered stillbirths in England and Wales, 1991-1998

Fetal death: early and late miscarriage, stillbirth; congenital malformations excluding minor abnormalities among live births; self-report with clinical confirmation attempted for fetal deaths and livebirths with reported abnormalities

Adjusted ORs: GWVs vs NDVs

father: all miscarriages 2,829/15,539 vs 1,525/10,988 (OR 1.4,95% CI 1.3-1.5); any congenital malformation, 686/13,191 vs 342/9,758 (OR 1.5, 95% CI 1.3-1.7); other malformations of digestive system, 69/13,191 vs 31/9,758 (OR 1.6, 95% CI 1.0-2.5);

genital system, 45/13,191 vs 19/9,758 (OR 1.8, 95% CI 1.0-3.0);

urinary systema, 103/13, 191 vs 48/9,758 (OR 1.6,95% CI 1.1-2.3);

musculoskeletal systema, 194/13, 191 vs 78/ 9,758 (OR 1.8, 95% CI 1.4-2.4);

other non-chromosomal malformations, 45/13, 191 vs 19/ 9,758 (OR 1.7, 95% CI 1.0-3.0);

cranial neural crest, 184/13, 191 vs 101/9,758 (OR 1.3, 95% CI 1.0-1.7);

metabolic and single gene defects, 22/13,191 vs 8/9,758 (OR 2.0, 95% CI 0.9-4.8);

mothers: no significant associations

Stratum matched on branch of service, sex, age, serving status, rank; ORs adjusted by year of pregnancy end, paternal/maternal pregnancy order, maternal age, service, rank, previous fetal death, multiplicity

Response rates: GWVs: men 53%, women 72%; NDVs: men 42%, women 60%

Limitations: poor response rates among men and response rates lower in NDVs, low numbers of miscarriages in NDVs compared with NIFS population could mean participation and reporting bias; multiple comparisons

Strengths: medical confirmation for some cases; fetal deaths as well as live births; external comparison groups to evaluate possible biases

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

Study

Design

Population

Outcomes

Results

Adjustments

Comments or Limitations

Goldenhar Syndrome

Araneta et al. 1997

Prevalence

Infants conceived after GW (or 12/31/1990 for NDVs) through 9/30/2003, born to active-duty military personnel in military hospitals; GWVs (n = 34,069), NDVs (n = 41,345)

Goldenhar syndrome criteria: presence of microtia, anotia, or preauricular tag; presence of either hypoplasia of mandible or a physical feature peculiar to syndrome, such as equibulbar dermoid or coloboma of upper lid

Unadjusted RRs: GWV fathers (5 cases) vs NDV fathers (2 cases) (RR 3.03, 95% CI 0.63-20.57)

Sex, race, history of fetal loss, maternal and paternal age, military occupation

Limitations: only military hospitals, may miss high-risk pregnancies, included cases only among liveborn infants diagnosed at birth

Werler et al. 2005

Case-control

HFM cases ∞3 years old (born 1996-2002) from craniofacial clinics in 24 US cities (n = 232); controls matched by age and pediatrician (n = 832)

HFM, facial asymmetry, or Goldenhar syndrome and no evidence of Mendelian inherited or chromosomal anomaly

Adjusted ORs: cases vs controls; parental army service, 22/232 vs 45/832 (OR 2.4, 95% CI 1.4-4.2); parental GW army service, 4/232 vs 9/832 (OR 2.8, 95% CI 0.8-9.6)

Family income, race, BMI in early pregnancy, multiple gestation

Limitations: unmeasured lifestyle factors

Strengths: included cases diagnosed up to of 3 years age

Adverse pregnancy outcomes

Araneta et al. 2004

Prevalence

Deployed women admitted to military hospitals for pregnancy-related diagnoses (including live births, abortions, ectopic pregnancies, pregnancy-related complications) from 8/2/1990 to 5/31/1992 and who responded to mailed survey:

Self-reported stillbirths, spontaneous abortions, ectopic pregnancies, pregnancy-related complications (ICD-9-CM codes 640-676); confirmed by discharge

Adjusted RRs: mothers: GWV vs NDV postwar conceptions: spontaneous abortions, 68 vs 39 (RR 2.92, 95% CI 1.87-4.56); ectopic pregnancies, 32 vs 6 (RR 7.70, 95% CI 3.00-19.8); GWV vs NDV exposed conceptions:

Age, race, education, marital status, branch of service, military rank, parity, history of adverse outcome

Overall response rate: 50%

Limitations: low response rate; no information on smoking, alcohol, caffeine, other known risk factors for fetal loss; possible limited generalizability due to

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

Study

Design

Population

Outcomes

Results

Adjustments

Comments or Limitations

 

 

GW-exposed conceptions (n = 415), GW postwar conceptions (n = 298), NDVs (n = 427)

diagnostic data

spontaneous abortions, 48 vs 39 (RR 1.44, 95% CI 0.91-2.29); ectopic pregnancies, 10 vs 6 (RR 1.91, 95% CI 0.67-5.46)

 

restriction to military hospital admissions; recall bias

Strengths: confirmation with discharge data, assessed GW-exposed and postwar conceptions

Male fertility problems and infertility

Maconochie et al. 2004 (same cohort as Doyle et al. 2004)

Retrospective reproductive cohort

Male UK veterans fathering or trying to father pregnancies after GW and before 8/97 GWV (n = 10,465) NDV (n = 7,376)

Self-reported fertility problems: tried unsuccessfully for >1 year and consulted doctor; type I infertility: never achieving pregnancy; type II infertility: never achieving live birth; semen quality; time to conception; attempted clinical confirmation from both partners’ physicians

Adjusted ORs: fertility problems, 732/10,465 vs 370/7,376: (OR 1.38, 95% CI 1.20-1.60); type I 259/10,465 vs 122/7,376 (OR 1.41, 95% CI 1.05-1.89); type II 356/10,465 vs 166/7, 376 (OR 1.50, 95% CI 1.18-1.89); time to conception >1 year for planned pregnancies, 845/9,968 vs 528/7,408 (OR 1.18, 95% CI 1.04-1.34) (increase in risk stable with time since GW)

Maternal and paternal age at first infertility consult or post-GW conception, year of first consult or conception, pre-GW pregnancy history, military service and rank, smoking, alcohol, pregnancy order

Response rates: GWVs, 53%; NDVs, 42%

Limitations: low response rates, possible recall bias, clinically evaluated only 40%

Strengths: attempted clinical evaluation, information on nonresponders available

Ishoy et al. 2001a

Cross-sectional

Danish Gulf War Study, GWVs (n = 661) NDVs (n = 215)

Self-reports of sexual problems (e.g., including reduced libido); measured male reproductive hormones; serum concentrations of LH, FSH

Male GWVs vs NDVs: self-reported sexual problems, 12.0% vs 3.7% (p<0.001); reproductive hormones, no significant difference; suspected oligospermia, FSH∞10 IU/L, inhibin B∞80 pg/ml,

Age; BMI available; stratified on deployment organization, duration of deployment

Participation rates: GWVs, 83.6%; NDVs, 57.8%

Limitations: limited control for confounding, small numbers for study of fertility rates, congenital

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

Study

Design

Population

Outcomes

Results

Adjustments

Comments or Limitations

 

 

 

testosterone, inhibin B

1.6% vs 1.6%; fertility rates, spontaneous abortion, congenital malformations: no differences

 

malformations

Strengths: measurement of hormones objective and unbiased

Ishoy et al. 2001b (elaboration of findings in Ishoy et al. 2001a)

Cross-sectional

Danish Gulf War Study: GWVs (n = 661), NDVs (n = 215)

Self-reported sexual problems

Male GWVs vs NDVs: sexual problems (80% decreased libido), 79/661 vs 8/215 (OR 2.9. (95% CI 1.4-6.0) (among GWVs associated with “having seen killed or wounded victims”; “having been threatened with arms”; “having watched colleagues being seriously threatened or shot at”; water hygienic environment)

Age

Limitations: small study, self-reported soft outcomes and exposures

Fertility problems

Ishoy et al. 1999b

Increased prevalence of self-reported sexual problems up to 6 years after war

NOTE: BMI = body-mass index; CDC = Centers for Disease Control and Prevention; FSH = follicle-stimulating hormone; GW = Gulf War; GWV = Gulf War veteran; HFM = hemofacial microsomia; LH = luteinizing hormone; NDV = nondeployed veteran; NIFS = Nuclear Industry Family Study.

  

aAssociations attenuated but still statistically significant with clinical confirmation

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

SYMPTOMS, SIGNS, AND ABNORMAL CLINICAL AND LABORATORY FINDINGS (ICD-10 R00-R99)

Unexplained Illness

Many of the symptom clusters that Gulf War veterans report are based on factor analysis of survey data. Those symptom clusters have been referred to as Gulf War Syndrome, chronic multisymptom illness, “unexplained” illnesses, etc. The term “unexplained” is not meant to imply that the illnesses are unique in being of unknown etiology, as that is true of many medical conditions, but rather that the illnesses do not fit into established medical diagnostic categories.

Several lines of inquiry, described and evaluated in this section, have tried to overcome reliance on established diagnoses by seeking to determine whether veterans have increased hospitalizations for any unexplained illness and are suffering from a potentially new syndrome. To uncover the existence of a new syndrome, most research has used a statistical technique known as factor analysis. The technique probes a cluster of symptoms to answer the question, Is the symptom cluster best studied and treated as a new and unique syndrome, or is it a variant of a known syndrome? The technique, its nomenclature, and its purposes are explained in Chapter 3.

Hospitalizations for Unexplained Illness

A hospitalization study (1991-1996) examined DOD’s hospital discharge dataset to search for excess admissions for unexplained illnesses in deployed veterans (n = 552,111) and nondeployed veterans (n = 1,479,751) (Knoke and Gray 1998). The authors reasoned that their earlier study of hospitalizations (Gray et al. 1996) might have missed those for a new or poorly recognized syndrome. Hospital discharge coding might have inconsistently classified such hospitalizations by many diagnoses, thereby masking an effect if one were present. The study operationally defined unexplained illnesses as diagnoses in several catchall ICD-9 diagnostic categories that comprised nonspecific infections and other ill-defined conditions. It examined only first hospitalizations to avoid overcounting medical conditions that required repeated hospitalizations among the subset of patients who had at least one unexplained illness coded on the discharge summary. Up to eight discharge diagnoses were examined per hospitalization. The authors found that deployed active-duty military members were less likely to have been hospitalized for unexplained illnesses than nondeployed (RR 0.93, 95% CI 0.91-0.96) (Knoke and Gray 1998). That finding adjusted for a variety of covariates and removed the effect of participation in the CCEP after June 1, 1994. Participants in that voluntary program had been admitted to the hospital only for evaluation. The Knoke and Gray (1998) study has the advantage of large study groups and high statistical power for the detection of an effect. Its major limitations are its inclusion of only active duty personnel and its inability to detect illnesses not severe enough to warrant hospitalization.

In another hospitalization study Gray et al. (2000) extended the examination of hospitalizations (1991-1994) to cover not only active duty but also reserve and former military personnel who had been deployed to the Gulf War. The study investigated hospitalizations at DOD, VA, and nonfederal hospitals in California to eliminate potential bias related to veterans seeking care outside DOD and VA facilities. Because of the unreliability of state-of-residence data in DOD and VA datasets, the authors could not directly compare rates of hospitalization

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

among the three sources. Therefore, they compared PMRs of hospital discharge diagnoses (14 diagnostic categories from ICD-9) in deployed vs nondeployed veterans. For VA hospitals, but not for DOD or California hospitals, the PMR was increased for ill-defined diseases (PMR 1.24, 95% CI 1.16-1.33).

Factor-Analysis Derived Syndromes

We focus now on the extensive literature using factor analysis and cluster analysis to determine whether veterans’ symptoms might constitute a new syndrome or whether they are a variant of a known syndrome. Secondary studies are included in tables and text, a departure from most other sections of the report. Similar to the neurobehavioral section, secondary studies provide valuable supplementary information that helps to increase or decrease confidence in the conclusions drawn from the primary studies, thus secondary studies also are included in this section.

The largest and most nationally representative survey of US veterans, conducted by VA, found that nearly 30% of veterans meet a case definition of “multisymptom illness”, compared with 16% of nondeployed veterans (Blanchard et al. 2006). Those figures indicate that unexplained illnesses are the most prevalent outcome following service in the Gulf War.

The committee identified five primary studies (Cherry et al. 2001b; Doebbeling et al. 2000; Forbes et al. 2004; Ismail et al. 1999; Kang et al. 2002) and four secondary studies. Those secondary studies have similar methodologic limitations that apply elsewhere in this volume, such as non-representative samples, selection bias, recall bias and small samples.

Primary Studies
University of Manchester (UK)

Through factor analysis, Cherry et al. (Cherry et al. 2001b) identified seven distinct factors in a large, population-based study of British Gulf War-era service members who answered 95 interval-scale symptom questions. Deployed veterans—two random samples of Gulf War veterans (main and validation cohorts)—were compared with a stratified sample of service members who had not been deployed. The seven factors, which accounted for 48% of the variance, could be found in all three groups separately and in combination: psychologic (24 symptoms), peripheral (10 symptoms), neurologic (13 symptoms), respiratory (11 symptoms), gastrointestinal (6 symptoms), concentration (10 symptoms) and appetite (5 symptoms). Deployed veterans’ mean factor scores7 were significantly higher for five factors: psychologic, peripheral, respiratory, gastrointestinal, and concentration. No difference was found in the neurologic factor, and appetite was significantly lower than in the nondeployed cohort. None of the factors was exclusive to Gulf War veterans, so the investigators concluded that their findings did not support a new syndrome (Cherry et al. 2001b). It was a large, diverse, population-based cohort study with high participation rates (86% overall) and had the added benefits of using a 21-point interval scale to record severity of symptoms and a mannequin to ease recording of peripheral dysesthesias and pain. The study included both retired and active-duty personnel and went to extraordinary lengths to find and account for all potential participants, thus minimizing the risk of selection bias. The investigators also divided the Gulf War veterans into two groups:

7

Mean factor scores were computed by adding the sum of mean symptom scores (from 0-21) for each symptom that loaded onto the factor and dividing by the number of symptoms.

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

the main and validation cohorts. They found the same factors in both cohorts, suggesting consistency. Finally, they took the additional step of conducting cluster analysis (see the next section, Cherry et al. 2001b) to find out whether symptom complexes were similar if a different statistical technique was used. Nonetheless, the analyses, as in other studies, relied on self-reported symptom data and thus are subject to usual concerns about recall bias.

Department of Veteran Affairs

The nationally representative VA study searched for potentially new syndromes through factor analysis by Kang and colleagues (Kang et al. 2002). Data were from a study drawn from 15,000 deployed and 15,000 nondeployed active-duty, reserve, National Guard, and retired service members from all four branches. The authors inquired, through questionnaires, about 47 symptoms on a three-point ordinal scale. Factor analysis of the deployed and nondeployed cohorts yielded the same six factors.8 The six factors were fatigue and depression9; neurologic10; musculoskeletal and rheumatologic11; gastrointestinal12; pulmonary13; and upper respiratory14. Several symptoms loaded onto more than one factor. In the deployed group, the fatigue or depression factor had an eigenvalue of 12.82 and accounted for 79% of the variance, and the neurologic factor had an eigenvalue of 1.27 and accounted for 8% of the variance. In the nondeployed group the corresponding numbers were 10.39 and 71% and 1.39 and 10%, respectively. Rather than examining differences between factors that loaded or did not load between the deployed and nondeployed groups, the authors examined which symptoms within each factor loaded and did not load. In the neurologic factor, four symptoms—loss of balance or dizziness, speech difficulty, blurred vision, and tremors or shaking—loaded for the deployed but not for the nondeployed group. A group of 277 deployed veterans (2.4%) and a group of 43 nondeployed veterans (0.45%) met a case definition that subsumed the four symptoms. The authors interpreted their findings as suggesting a possible unique neurologic syndrome related to Gulf War deployment that would require objective supporting clinical evidence. The study was a large, diverse, population-based study that had a relatively high response rate of 70%. The use of low eigenvalues to create a six-factor solution was unconventional and may have led to overinterpretation of the data. In the authors’ words, “as with the U.K. study (Ismail et al. 1999), we observed that the Gulf War and non-Gulf War veterans displayed virtually identical factor solution in the 5-factor analysis. Only the 6-factor solution model produced different results for the Gulf War and non-Gulf War veterans.” Notably the six-factor solution necessitated the use of very low eigenvalues. In addition, the authors’ examination of symptoms that loaded onto the neurologic factor for deployed veterans was unconventional in the factor analysis literature but was a unique approach to extracting potentially important symptom differences.

8

It should be noted that in reaching the six-factor solution, Kang et al. went far below the conventional eigenvalue cutoff value of 1.0 (the so-called Kaiser-Guttman rule) and included four factors with eigenvalues below 1.0 for the deployed group and three for the nondeployed group.

9

Awakening tired and worn out, concentration and memory problems, excessive fatigue, fatigue more than 24 hours after exertion; feeling anxious, irritable, or upset; feeling depressed or “blue”; sleep difficulty; sleepiness during daytime.

10

Blurred vision, concentration or memory problems, irregular heartbeat, loss of balance or dizziness, speech difficulty, sudden loss of strength, tremors or shaking, excessive fatigue, fatigue more than 24 hours after exertion.

11

Back pain or spasms, generalized muscle aches, joint aches, numbness in hands or feet, swelling in joints, swelling in extremities.

12

Constipation, diarrhea, nausea; reflux, heartburn, or indigestion; stomach or abdominal pain, vomiting.

13

Coughing, irregular heartbeat, shortness of breath, tightness in chest, wheezing.

14

Coughing, runny nose, sore throat, swollen glands, trouble swallowing.

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

A later study by VA researchers (Blanchard et al. 2006) sought to determine the prevalence of unexplained illnesses in the VA study. This study was not a factor analysis itself but applied CDC’s definition of chronic multisymptom illness, which had been derived in part by factor analysis (Fukuda et al. 1998), to determine its prevalence. The Fukuda et al. study also provided prevalence figures, but because of its nonrepresentative sample, the committee considered it a secondary study. The VA investigators first assessed veterans’ responses to the 47 symptom questions. Participants were asked about symptoms in face-to-face interviews 10 years after the war. Several other publications resulting from VA's study are discussed elsewhere in this volume (for example, Eisen et al. 2005). Participants who reported one or more symptoms from each of three clusters (fatigability, mood and cognition, and musculoskeletal) were considered to meet the case definition.15 Cases were classified as severe if at least one symptom in each cluster was rated as severe. The investigators found that overall 29% of deployed participants met the criteria for chronic multisymptom illness as opposed to 16% of nondeployed participants. Severe chronic multisymptom illness was found in 7% of deployed veterans and 1.6% of nondeployed veterans. Among deployed veterans a higher score on the Combat Exposure Scale was associated with chronic multisymptom illness. Among nondeployed veterans, female sex, less than a college education, and a higher score on the Combat Exposure Scale were associated with chronic multisymptom illness. Deployed veterans who met the case definition had lower mean scores on the SF-36 for physical and mental health, more nonroutine clinic visits and a higher mean number of prescriptions, and they were more likely to be using psychotropic medications. Chronic multisymptom illness in both deployed and nondeployed veterans was more likely to be associated with fibromyalgia syndrome, CFS, symptomatic arthralgias, dyspepsia, the metabolic syndrome, PTSD, anxiety disorders, major depression, nicotine dependence, and more than one psychiatric diagnosis during the year preceding the examination but, with the exception of CFS, was no more likely to be associated with deployment than with nondeployment. The report was based on the same cohort as Fukuda et al. and had the same problems of low participation rates and self-reporting of symptoms. Nonetheless, it provides evidence that the cluster of symptoms previously studied by Fukuda et al. also existed in this group and 10 years after the Gulf War was twice as common in participants who had been deployed as in those who had not been.

The Iowa Cohort

The Iowa study (Iowa Persian Gulf Study Group 1997) was the first major population-based study to group symptoms into categories suggestive of existing syndromes or disorders, such as fibromyalgia or depression. Its finding of a considerably higher prevalence among Gulf War veterans of symptom groups suggestive of fibromyalgia, depression, and cognitive dysfunction motivated the first applications of factor analysis to group and classify veterans’ symptoms. Several years later, the same team of Iowa investigators performed a factor analysis on the Iowa cohort (Doebbeling et al. 2000). They studied the frequency and severity of 137 self-reported symptoms among 1,896 Gulf War veterans and 1,799 veterans who had not been deployed. They randomly divided the Gulf War veterans into two groups that they called the derivative sample and the validation sample. They identified three symptom factors in deployed veterans in the derivative sample that accounted for 35% of the variance: somatic distress (joint

15

Persistent fatigue for at least 24 hours after exertion was the single symptom in Cluster A (fatigability). Feeling depressed; feeling irritable; difficulty in thinking or concentrating; feeling worried; tense, or anxious; problems in finding words; and problems in going to sleep were the symptoms included in Cluster B (mood and cognition). Joint pain and muscle pain were the two symptoms included in Cluster C (musculoskeletal).

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

stiffness, myalgia, polyarthralgia, numbness or tingling, headaches, and nausea), psychologic distress (feeling nervous, worrying, feeling distant or cut off, depression, and anxiety), and panic (anxiety attacks; a racing, skipping or pounding heart; attacks of chest pain or pressure; and attacks of sweating). They confirmed them in a separate factor analysis in the validation sample. They also conducted factor analysis in the nondeployed group and found the same three factors, which accounted for 29% of the variance. Thus, the study did not support the existence of a new syndrome. The authors noted the difficulty of attributing to a single condition the increased reporting of nearly every symptom in every bodily system. The strengths of this study included the size and diversity of its study population and the inclusion of nondeployed veterans. It also had a substantially higher participation rate (90.7%) than earlier studies.

Guy's, King's, St. Thomas’s Schools of Medicine (UK)

Ismail and colleagues (Ismail et al. 1999) applied factor analysis to a large representative sample of UK veterans. They were able to identify three fundamental factors, which they classified as related to mood and cognition (headaches, irritability or outbursts of anger, sleeping difficulties, feeling jumpy or easily startled, unrefreshing sleep, fatigue, feeling distant or cut off from others, forgetfulness, loss of concentration, avoiding doing things or situations, and distressing dreams), the respiratory system (unable to breathe deeply enough, faster breathing than normal, feeling short of breath at rest, and wheezing), and the peripheral nervous system (tingling in fingers and arms, tingling in legs and arms, and numbness or tingling in fingers or toes). The pattern of symptom reporting by Gulf War veterans differed little from the patterns reported by Bosnia and Gulf War-era comparison groups, although the Gulf War cohort reported a higher frequency of symptoms and greater symptom severity. The UK authors interpreted their results as arguing against the existence of a unique Gulf War syndrome. Strengths of the study were its two comparison groups and its ability to compare how well its three-factor solution fit its Bosnian and nondeployed (“era”) cohorts. As with the study by Haley et al. (1997b, see below), however, a lower-than-ideal response rate of 65% may have introduced selection bias.

Australian Cohort

In a population-based study of all Australian Gulf War veterans, Forbes and colleagues (Forbes et al. 2004) applied factor analysis to findings from a 62-item symptom questionnaire. Symptom reporting was ordinal: “none”, “mild”, “moderate”, and “severe”. Three factors were found that accounted for 47.1% of the variance: psychophysiologic distress (23 symptoms), cognitive distress (20 symptoms), and arthroneuromuscular distress (six symptoms). They were broadly similar to factors in previous analyses and were the same as factors found among nondeployed Australian veterans. However, although the prevalence was similar among deployed and nondeployed veterans, factor scores were higher among the deployed than among the nondeployed. That indicates greater severity of symptoms. The authors concluded that there was no evidence of a unique pattern of self-reported symptoms in deployed veterans. One limitation of this study is that most members of the Australian cohort were from the Navy, so its generalizability to services and personnel from other countries, particularly the United States, may be limited. Nonetheless, its inclusion of all Australian Gulf War veterans and a stratified random sample of nondeployed Gulf War-era Defence Force personnel eliminated the potential for selection bias that other studies had more difficulty in controlling. It is also valuable in setting a baseline of unexplained illness as an effect of deployment itself without the overlay of direct combat and environmental exposures more commonly encountered on land.

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Secondary Studies
Air National Guard

Fukuda and colleagues (1998) used factor analysis and other methods to assess the health status of Gulf War Air Force veterans in response to a request from DOD, VA, and the state of Pennsylvania. Their focus was to assess the prevalence and causes of an unexplained illness in members of one Air National Guard unit. By studying that unit and three comparison Air Force populations, the investigators aimed to organize symptoms into a case definition and to carry out clinical evaluations on participants from the index Air National Guard unit. They administered a 35-item symptom inventory that included symptom severity (mild, moderate, or severe) and duration (less than 6 months or 6 months or longer) and divided the 3,255 participants who had answered all symptom questions into two subsamples of 1,631 and 1,624. They conducted an exploratory factor analysis of the first subsample that yielded 10 factors with eigenvalues greater than 1.0; three of the factors accounted for 39.1% of the total variance. When the three were examined in a confirmatory factor analysis in the second subsample, two could be confirmed. The first, called mood-cognition-fatigue, consisted of the symptoms: feeling depressed, feeling anxious, feeling moody, difficulty in remembering or concentrating, trouble in finding words, difficulty in sleeping, and fatigue. The second, called musculoskeletal, consisted of the symptoms: joint stiffness, joint pain, and muscle pain. They used those 10 symptoms from the two confirmed factors to develop a preliminary case definition having a combined factor score in the top 25th percentile. That was compared with an alternative clinical case definition of having one or more symptoms in each of two of three symptom categories: fatigue, mood-cognition, and musculoskeletal. Forty-five percent of deployed veterans met the factor-score-based case definition, whereas only 15% of nondeployed veterans met it. The same percentages met the clinical case definition, which because of its greater clinical simplicity than the factor-score-based case definition, was then used to create a case definition of chronic multisymptom illness. The new definition, which was used in later studies, was having one or more chronic symptoms (present for 6 months or longer) from at least 2 of the 3 categories: fatigue, mood-cognition16, and musculoskeletal17. A case was classified as severe if each symptom reported that was used to meet the case definition was rated as severe.

Of the participants surveyed, those deployed to the Gulf War experienced a higher prevalence of chronic symptoms (33 of 35 symptoms with more than 6-month duration were reported to be more prevalent) than nondeployed veterans. According to the case definition of chronic multisymptom disease, 39% of Gulf War-deployed veterans and 14% of nondeployed veterans had mild-to-moderate cases, and 6% and 0.7%, respectively, had severe cases. On the basis of a total of 158 clinical examinations performed in one unit, there were no abnormal physical or laboratory findings that differentiated those who met the case definition from those who did not meet the case definition. Cases, however, reported significantly lower functioning and well-being.

Because such a large fraction (14%) of nondeployed veterans met the mild-to-moderate case definition, the investigators concluded that the case definition could not specifically characterize Gulf War veterans who had unexplained illnesses (Fukuda et al. 1998). The study, however, had several limitations, the most important of which was its coverage of only current

16

Symptoms of feeling depressed, difficulty in remembering or concentrating, feeling moody, feeling anxious, trouble finding words or difficulty in sleeping.

17

Symptoms of joint pain, joint stiffness, or muscle pain.

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

Air Force personnel several years after the Gulf War (Air National Guard, Air Force Reserve, and-active duty personnel), which limits its generalizability to other branches of service and to those who left the service possibly because of illness. The use of self-reported symptoms introduced the possibility of reporting bias, and the low participation rates in two of the four units (62% and 35%) introduced the possibility of selection bias. Nonetheless, symptom reporting and prevalence were similar among the four units. A particular strength of this study was its use of a symptom inventory rather than asking veterans about specific diagnoses, such as CFS, MCS, depression, and various neurologic abnormalities. Its use of a more intensive examination of Gulf War veterans from the index unit—including an additional clinical questionnaire; interviewer-administered modules on major depression, somatization disorder, and panic disorder; a screening physical examination with blinded examiners; and a variety of laboratory tests—provided important additional data even though participation rates were low (62%).

Seabee Cohort

Knoke et al. (Knoke et al. 2000) applied factor analysis to active-duty Seabees in response to the factor analysis conducted by Haley et al. (1997b). The study population was drawn from US Navy construction battalion personnel (Seabees) who were on active duty in 1990 and remained on active duty in 1994, when the study was conducted. The instrument contained 98 symptom questions. Among the 524 Gulf War veterans and 935 nondeployed Seabees, Knoke and colleagues performed three factor analyses: the first on the deployed Seabees, the second on the nondeployed Seabees, and the third on both. Each factor analysis identified five factors that accounted for 80%, 89%, and 93% of the total variance respectively. The factors were insecurity or minor depression (27 symptoms), somatization (13 symptoms), depression (10 symptoms), obsessive-compulsive (seven symptoms), and malaise (seven symptoms). Scores among the three analyses were similar for insecurity or minor depression; higher in Gulf War veterans for somatization, depression and obsessive-compulsive; and higher in nondeployed Seabees for malaise.18 Somatization, depression, and obsessive-compulsive affected an excess of about 20% of Gulf War veterans. The findings were similar to those of Doebbeling (2000), Fukuda (1998) and Ismail (1999) and consistent with findings in a civilian population with CFS (Nisenbaum et al. 1998). They concluded that, unlike the results of the Haley et al. (1997b) study of Seabee reservists from one unit, there was no evidence of a unique spectrum of neurologic injury. While the Knoke et al. study used a larger population than Haley et al., this study still used personnel from a single service, so its generalizability is limited. Because participants were active-duty personnel by design, the results cannot be generalized to retired or reserve personnel who might have been more symptomatic. Nonetheless, the authors’ careful examination of the methods and findings of Haley et al. in a very similar population makes this study quite useful.

Haley et al. Seabee Cohort and Validation Study

Haley and collaborators (Haley et al. 1997b) studied a battalion of naval reservists called to active duty for the Gulf War (n = 249). More than half the battalion had left the military by the time of the study. Of those participating, 70% reported having had a serious health problem since returning from the Gulf War and about 30% reported having no serious health problems. The

18

Factor scores used to compare the groups were computed from the regression coefficients of the Gulf War veteran factor analysis, standardized for both groups by subtracting the median and dividing by the semi-interquartile range of the score for the Gulf War veteran group.

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

study was the first to examine groupings of symptoms in Gulf War veterans using factor analysis. Through standardized symptom questionnaires and a two-stage exploratory factor analysis, the investigators defined what they considered to be either six syndromes or six variants of a single syndrome, which they labeled impaired cognition, confusion-ataxia, arthromyoneuropathy, phobia-apraxia, fever-adenopathy, and weakness-incontinence. One-fourth of the veterans in this uncontrolled study (n = 63) were classified as having one of the six syndromes. The first three syndromes had the strongest factor clustering of symptoms. The study was limited by its lack of a comparison group; the authors were unable to comment on the uniqueness of the factors in relation to other groups of veterans. The findings were based entirely on symptoms self-reported in a mail survey; no in-person interview, physical examination, or laboratory data assisted in the characterization of the factors. In addition, the low participation rate (58%) could have introduced selection bias in that people more or less symptomatic may have participated preferentially. The authors comment that the only evidence they found of differential participation was that nonparticipants were less likely to report a serious illness and to be unemployed. If that can be generalized to all nonparticipants, it would mean that the prevalence of symptoms among participants may have been systematically overestimated. Finally, the study population came from a single unit, so findings cannot necessarily be generalized to all Gulf War veterans.

Haley and colleagues (2001) attempted to replicate their factor analysis findings in a validation cohort, which was separate from their original cohort of Seabees. The validation cohort (n = 335) consisted of veterans living in North Texas who had registered with a VA clinic in Dallas or who were recruited by advertising. In comparison with the Seabee cohort, participants in the validation cohort were more likely to have served in the Army and in general to be more representative of those who served in the gulf with regard to racial and ethnic background, age, and wartime military status. In this study, the authors used more sophisticated questionnaires than in the earlier Seabee cohort in an effort to replicate the earlier findings. They undertook a series of analyses to test whether the latent syndrome structure they found in the earlier cohort could be replicated in the larger and more representative cohort. They allowed, by design, only four symptom scales per syndrome factor to load onto five models and compared the five models with their earlier findings by using structure equation models. The five models had either 12 or 16 measured variables, which loaded onto three first-order factors and zero or one higher-order factor. In two models, the four additional variables (or symptom factors) were allowed to load onto the primary or higher-order factors. The three primary syndrome factors were impaired cognition, confusion-ataxia, and central pain (termed arthromyoneuropathy in the original study); and the four additional variables or secondary symptom factors were chronic watery diarrhea, chronic fatigue involving excessive muscle weakness, chronic fever and night sweats, and middle and terminal insomnia. The higher-order factor was the presence of an underlying single Gulf War syndrome that could explain all variance and covariance among the three first-order factors. Overall, 29% of participants had one or more of the three first-order factors, defined by dichotomizing the syndrome factor scale at 1.5, as in the original study. They found that the apparent three-factor solution, originally demonstrated in the Seabee cohort, was also present in this new cohort (Model 1); that the three syndrome factors probably represented a higher-order syndrome, such as a single Gulf War syndrome (Model 2); and that some additional symptoms (the four secondary symptom factors) appeared in all three syndrome variants. They suggested that the confusion-ataxia syndrome may represent a more severe form of a single Gulf

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

War syndrome of which impaired cognition and central pain variants (the other two syndrome factors) were less severe forms.

The small sample may have limited exploration of less common symptoms, and the nonrandom sample may have limited the generalizability of some of the results, such as syndrome prevalence, to some degree; but the detailed questionnaires, the substantially more refined symptom measures, and the external validation of the findings through comparison with the Seabee cohort were strengths of the second study. Note that this study, by design, had no comparison group. The authors were seeking to validate the presence of a symptom complex in deployed veterans, rather than to examine its prevalence in deployed vs nondeployed forces. The authors concluded by recommending study of a national randomly selected sample of deployed and nondeployed Gulf War-era military populations with their methods of symptom measurement and syndrome definition.

Portland Area Veterans

Investigators studied clusters of unexplained symptoms in a population-based study of Portland area veterans by creating a new case definition of unexplained illness (Storzbach et al. 2000). Cases were identified through questionnaires as meeting a threshold number and combination of symptoms (cognitive and psychologic, and musculoskeletal) and on the duration of fatigue. Veterans whose symptom clusters remained unexplained at clinical examination (after exclusion of established diagnoses) were defined as constituting cases. Controls were those who at the time of clinical examination had no history of case-defining symptoms during or after their service in the Gulf War. In an analysis of the 241 cases and 113 controls, investigators found small but statistically significant deficits in cases on some neurobehavioral tests of memory, attention, and response speed. Cases also were statistically significantly more likely to report increased distress and psychiatric symptoms (Storzbach et al. 2000). Finally, more than half the veterans with unexplained musculoskeletal pain met symptom-based criteria for fibromyalgia, and a large proportion met symptom-based criteria for CFS (Bourdette et al. 2001). The study also undertook a factor analysis, which initially loaded 48 of the 69 symptoms and accounted for 21.6% of the variance. The researchers then re-examined the 48 symptoms in a second factor analysis. Three were retained for rotation and further analysis. These three factors loaded 35 symptoms; cognitive and psychologic, mixed somatic, and musculoskeletal accounted for 34.2% of the common variance. Rather than using those factors as a working case definition to explore symptom differences between deployed and nondeployed veterans, the authors used their three-factor solution to validate their a priori case definition composed of 35 symptoms encompassing musculoskeletal pain, cognitive and psychologic changes, gastrointestinal complaints, skin or mucous membrane lesions, or unexplained fatigue.

There were two major findings when the researchers compared the three-factor solution to the a priori case definition of Gulf War unexplained illnesses. First, the factor analysis did not include any symptoms related to the gastrointestinal system, the skin, or mucous membranes. Second, three symptoms that had loaded onto the musculoskeletal factor—numbness in fingers or toes, clumsiness, and dizziness—were not included in the case definition. They then used that information to assess the accuracy of their case definition. The three-factor solution identified 103 (91%) of the 113 controls and 189 (78%) of the 241 cases in their clinically evaluated subsample. They also tested their three-factor solution against a modification of the clinical case

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

definition-approach used by Fukuda and colleagues (1998)19 and found that the three-factor solution predicted 103 (91%) of 113 controls and 108 (94%) of 115 cases in their study. They concluded that their factor analysis confirmed the finding of a cognitive and psychologic factor found by Haley et al. (1997b), Fukuda et al. (1998), and Ismail et al. (1999), and the finding of a musculoskeletal factor reinforced the finding of a musculoskeletal factor by Haley et al. and Fukuda et al. This study, like others, had a relatively low response rate (64%), which introduced the possibility of selection bias. As a result of its case-control design, the study population by definition comprised only Gulf War veterans; this eliminated the possibility of examining differences between deployed and nondeployed veterans. Nonetheless, its careful clinical and psychologic examinations of a sample of the responders is a strength, and its use of factor analysis to validate its a priori case definition rather than to create one de novo is unique among the factor analysis studies.

Shapiro and colleagues (Shapiro et al. 2002) performed factor analysis to determine whether there was a unique syndrome among veterans in the Oregon cohort who witnessed the Khamisiyah demolition and might thereby have been exposed to nerve agents. They divided their population-based sample into three groups: witnesses to Khamisiyah demolitions, nonwitnesses who had been deployed to the Gulf War, and nondeployed veterans. Their analysis of 25 symptoms among three groups of 1,779 veterans identified three factors that accounted for 46.7-52.2% of the overall variance, depending on the exposure group. Three factors were common to all three groups: cognitive or psychologic (in the exposed group, this included unusual irritability or anger; mood swings; changes in memory; persistent fatigue, tiredness, or weakness; difficulty in concentrating; and depression), dysesthesia (in the exposed group, tingling, burning sensation of pins and needles, and numbness or lack of feeling), and vestibular dysfunction (in the exposed group, loss of balance or coordination and dizzy spells). There were slight differences in symptoms that loaded in the nonexposed and nondeployed groups. There were no differences in distributions of factor scores between the three groups in terms of the cognitive or psychologic and vestibular dysfunction subgroups, but in log-linear analysis and logistic regression analysis the dysesthesia factor was significantly associated with having witnessed the demolition.

Shapiro and colleagues provide two important cautions in the use of factor analysis. The first is that—because different investigators use different lists of symptoms, different samples, and different factor extraction and rotation techniques and factor-loading cutoffs—results from different studies are not directly comparable. The second is related to the use of factor analysis with dichotomous variables. As noted above, factor analysis traditionally has used interval-level data or at least ordinal data. However, most Gulf War syndrome investigators have used dichotomous data to assess symptom presence or absence. To examine the potential pitfalls inherent in that approach, Shapiro and her colleagues created a dataset of 19 dichotomous variables typically used in Gulf War syndrome research, randomly generated values for them whose row and column totals corresponded to the frequencies observed in the study, and applied factor analysis. They repeated that 500 times and found that their random datasets could result in five factors that explained 30% of variance and loading more than 95% of the time at the traditional 0.4 cutoff. That potentially has profound implications for the interpretation of data from factor analysis. As Shapiro et al. note, “in the absence of more robust decision rules for

19

In an initial clinical case definition, Fukuda et al. used any symptom reported for 6 months or longer by at least 25% of Gulf War veterans and at least 2.5 times more frequent among deployed veterans than among nondeployed veterans. For their analysis, Bourdette et al. modified the Fukuda et al. case definition to any symptom reported by 50% of Gulf War veterans reported for at least 1 month in the preceding 3 months.

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

these kinds of data, the resulting factor may be a rich mixture of randomness, which could lead investigators down uninformative paths.”

Department of Veterans Affairs Gulf War Health Registry

Hallman and colleagues (2003) examined patterns of reported symptoms among a sample of persons who participated in the VA Gulf War Health Registry. The study population consisted of a state-based random sample of 2,011 veteran registry members residing in Delaware, Illinois, New Jersey, New York, North Carolina, Ohio, and Pennsylvania who were not participating in other studies. Questionnaires included 48 symptoms, which were rated on a three-point ordinal scale, and were returned by 1,161 veterans (58% of the sample). The investigators divided the participants into two groups and conducted five factor analyses in each group to ensure consistency. They identified four factors that accounted for 50.2% of the variance. The factors were mood-memory-fatigue (depression, anxiety, sudden mood changes, problems concentrating and remembering, unexplained weakness, sleep problems, and unexplained fatigue), musculoskeletal (pain or numbness in joints or muscles), gastrointestinal (abdominal pain and gas, diarrhea, nausea, and vomiting) and throat-breathing (difficulty in swallowing, swollen glands, nose or sinus problems, coughing, difficulty in breathing, and difficulty in tasting). Like Cherry et al. (2001b), they also conducted a cluster analysis (see below) to examine consistency between the two different statistical methods. The principal limitation of the study is the lack of a nondeployed control group, which limits its ability to identify factors that may have been peculiar to exposure to the Gulf War. However, by starting with presumably the most symptomatic subset of Gulf War veterans (those who had left the service and registered with the Gulf War Health Registry), it also had power to identify clusters peculiar to symptomatic Gulf War veterans. However, the four factors it identified were largely similar to factors identified both by other Gulf War investigators and in civilian populations (Gillespie et al. 1999; Nisenbaum et al. 1998).

Cluster Analysis

A somewhat related technique, cluster analysis (see Chapter 3), has been used in three cohorts to determine how groups of patients with particular symptoms may be related to one another (Cherry et al. 2001b; Everitt et al. 2002; Hallman et al. 2003).

University of Manchester Cohort (UK)

Cherry and colleagues (Cherry et al. 2001b) sequentially partitioned members of the three cohorts, using scores from the 95 symptoms reported. Convergence was reached within 200 iterations. Participants divided into six clusters, which the authors then compared with their seven-factor solution and the standardized mean factor scores. Cluster 1 was composed primarily of well people and had a smaller proportion of Gulf War veterans (36.4%) than nondeployed veterans (48.5%). Clusters 2 and 3 had similar prevalences in both groups. The final three clusters accounted for 23.8% of Gulf War veterans but only 9.8% of nondeployed veterans and included clusters with high scores on respiratory and gastrointestinal illnesses (cluster 4), on psychologic ill health (cluster 5), and both overall and especially on neurologic symptoms (cluster 6). Thus, there was an excess of 14% of Gulf War veterans in the three least healthy clusters.

Guy’s, King’s, St. Thomas’s Schools of Medicine

Everitt and colleagues (2002) randomly sampled 500 participants from among the three cohorts (Gulf War veterans, Bosnian veterans and nondeployed Gulf War-era controls). They

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

regrouped the original 50 ordinal-scale symptoms into 10 categories, retaining the same four-point severity score. They also used a technique, known as the gap statistic, that can be used to suggest the number of clusters that best describe the data (Tibshirani 2001). They identified five clusters by using conventional cluster analysis. Cluster 1 had low scores for all symptoms, cluster 2 had the highest scores for musculoskeletal symptoms and high scores for neuropsychologic, cluster 3 had high scores for neuropsychologic and higher scores for the remaining nine symptom groups, cluster 4 had high scores only for musculoskeletal symptoms, and cluster 5 had high scores in all 10 symptom groups, especially musculoskeletal and neuropsychologic. Gulf War veterans were 3-4 times more likely to fall into cluster 2 and 11 times more likely to fall into cluster 5, although cluster 5 contained only 26 people from all three cohorts combined. With the gap statistic, two clusters were identified, one with low scores in each symptom group and another with higher mean scores for the musculoskeletal and neuropsychologic groups. Some 72% percent of Gulf War veterans, 87% of Bosnian veterans, and 94% of era-deployed veterans were classified in cluster 1. The authors interpreted their findings to mean that there was no convincing evidence of a new unique Gulf War syndrome. The study’s strengths included a random sample and standardized symptom measurements. However, the authors caution that the finding of nonspecific symptoms may have been distorted by reporting bias.

Department of Veterans Affairs Gulf War Health Registry

Hallman and colleagues (2003) conducted cluster analysis in their examination of 1,161 veterans participating in the VA Gulf War Health Registry. They used the mean factor scores from their factor analysis to group respondents on the basis of severity of symptoms. Examining the two randomly divided subsamples a total of five times each but using cluster analysis, they identified two stable clusters. Cluster 1, making up 60.4% of the sample, consisted of veterans who reported no or mild symptoms in each of the four factors. Cluster 2, the remaining 39.6% of participants, consisted of veterans with moderate-to-severe factor scores in the mood-memory-fatigue and musculoskeletal factors and mild-to-moderate scores in the gastrointestinal and throat-breathing factors. People classified in cluster 2 reported twice as many symptoms, reported more severe problems, were in poorer health, and had a greater reduction in mean activity as people in cluster 1 (37.2% vs 17.8%).

Summary and Conclusion

Factor analysis has been performed on all the major cohorts covered in Chapter 4, and cluster analysis has been performed on three cohorts, one of which was studied by both methods. The findings, despite methodologic differences, are quite similar. There seem to be similar groups of symptoms that fall roughly into factors that describe neurocognitive symptoms, musculoskeletal symptoms, and peripheral nervous system symptoms. Less commonly reported are factors that involve gastrointestinal and respiratory symptoms. Well-conducted factor analysis starts with representative samples with high participation rates. Several studies fall short on those two criteria, for instance, by including members of only one branch of the service (e.g., Fukuda et al. 1998; Haley et al. 1997b; Knoke et al. 2000), small samples (e.g., Haley et al. 1997b), or largely symptomatic groups of veterans (Hallman et al. 2003). Another problem is the lack of a comparison group in some of the studies, which limits investigators’ ability to compare the presence of factors in deployed and nondeployed groups (e.g., Bourdette et al. 2001; Haley et al. 1997b). Although results of the studies are valuable and add rich detail to the epidemiologic literature surrounding Gulf War veterans, other studies are more representative and hence more

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

generalizable (Ismail et al. 1999; Cherry et al. 2001b; Doebbeling et al. 2000; Kang et al. 2002). In those studies, the findings were quite similar, broadly describing neurologic, psychologic, cognitive, fatigue, and musculoskeletal symptoms. One exception is the Kang et al. study’s last three factors—gastrointestinal, pulmonary, and upper respiratory—which may have been the result of using eigenvalues less than 1.0 in constructing the models. The three studies that compared the factors that emerged in the most representative deployed and nondeployed groups (Cherry et al. 2001b; Doebbeling et al. 2000; Ismail et al. 1999) found factors that were remarkably similar between the deployed and nondeployed groups and did not suggest a unique complex of symptoms that existed in the deployed group but not in the nondeployed group. However, in each of those studies, as in many of the less generalizable studies, symptoms were more severe in the deployed than in the nondeployed groups.

The three studies that used cluster analysis had broadly similar findings. The two studies that included nondeployed comparison groups (Cherry et al. 2001b; Everitt et al. 2002) failed to identify a unique cluster of symptoms that might represent a unique Gulf War syndrome, but each did identify a highly symptomatic cluster of patients that had a statistically significantly higher proportion of Gulf War veterans than non-Gulf War veterans, ranging from 14% in Cherry et al. (Cherry et al. 2001b) to 22.2% in Everitt et al. (Everitt et al. 2002) (two-cluster solution comparing Gulf War with Gulf War-era veterans). The cluster analyses confirm the finding from the most representative factor-analysis studies: although there is not a unique symptom complex among Gulf War veterans, they are clearly more symptomatic than their nondeployed counterparts.

In the end, studies using those designs have their limitations as noted by Shapiro (2002), and should be viewed in the context of their inherent limitations. They demonstrate that deployed veterans report more symptoms and more severe symptoms than their nondeployed counterparts. However, there is no symptom complex peculiar to deployed Gulf War veterans. The primary and secondary studies are summarized in Table 5.19.

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

TABLE 5.19 Factor Analyses of Gulf War Veteran Cohorts (Primary and Secondary Studies)

Reference

Population

Type of Data

Method

Rotation

Factor Loading Cutoff and Eigenvalue Cutoff

No. Factors Isolated

Percentage of Variance Explained

Factors Identified

Unique Factors in Deployed Veterans?

Cherry et al. 2001b (primary)

Active and retired n = 11,914

Interval (21 points)

Principal components

Orthogonal

> 0.40, Not stated

7

48

Psychologic, peripheral, neurologic, respiratory, gastrointestinal, concentration, appetite

All present; mean factor scores higher in GWVs for psychologic, peripheral, respiratory, gastrointestinal, concentration; lower for appetite

Kang et al. 2002 (primary)

Active and retired n = 19,383

Ordinal

Iterated principal factors

Orthogonal

> 0.30, extracted factors had to have at least two symptoms >0.40; 1 0.57 for GWVs, 0.53 for nondeployed

6

 

Fatigue or depression, neurologic, musculoskeletal/rheumatologic, gastrointestinal, pulmonary, upper respiratory

Factors similar but 4 neurologic symptoms loaded on neurologic factor for deployed but not for non-deployed

Doebbeling et al. 2000 (primary)

Active and reserve n = 3,695

Ordinal and dichotomous

Unknown

Varimax and promax

1 0.35 Not stated

3

35 in both samples of deployed, 30 in non-deployed

Somatic distress; Psychological distress; Panic

Correlation between derivative and validation samples, same factors in non-deployed. Prevalence not stated.

Ismail et al. 1999 (primary)

Active n = 3,214

Ordinal and dichotomous

Principal factors

Orthagonal

> 0.40

>1.0

3

~20

Mood-cognition; Respiratory system; Peripheral nervous system

No but 3-factor solution fit less well in Bosnian cohort than

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

Reference

Population

Type of Data

Method

Rotation

Factor Loading Cutoff and Eigenvalue Cutoff

No. Factors Isolated

Percentage of Variance Explained

Factors Identified

Unique Factors in Deployed Veterans?

 

 

 

 

 

 

 

 

 

inGW deployed and less well in non- deployed than in Bosnian cohort.

Prevalence not mentioned

Forbes et al. 2004 (primary)

Active and retired n = 2,781

Ordinal

Unknown

Orthagonal and oblique

> 0.40

> 1.0

3

47.1

Psycho-physiological distress; cognitive distress; arthro-neuromuscular distress

No Prevalence similar but severity higher in GWV

Fukuda et al. 1998 (secondary)

Active and reserve (Air Force) n = 3,255

Ordinal

Principal components

Oblique (Promax, in subsample 1 and Procrustes in subsample 2)

> 0.40

> 1.0

3

39.1

Fatigue; mood-cognition; musculoskeletal pain

NA

45% of deployed met factor score-based case definition of chronic multisymptom illness vs 15% of non-deployed

Knoke et al. 2000 (secondary)

Active (Navy) n = 1,459

Ordinal and dichotomous

Principal factors

Orthagonal

> 0.40

> 1.0

5

80-93

Insecurity; somatization; depression; obsessive-compulsive; malaise

Somatization, depression, obsessive-compulsive. About 3 times as common

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

Reference

Population

Type of Data

Method

Rotation

Factor Loading Cutoff and Eigenvalue Cutoff

No. Factors Isolated

Percentage of Variance Explained

Factors Identified

Unique Factors in Deployed Veterans?

Haley et al. 1997b (secondary)

Active and retired (Navy) n = 249

Interval

Principal axes

Orthagonal

> 0. 40

1.0

6

71

Impaired cognition; Confusion-ataxia; Arthromyoneuropathy; Phobia-apraxia; Fever-adenopathy; Weakness-incontinence

NA

Haley et al. 2001 (secondary)

Active, reserve, retired n = 335

Continuous

Principal factors (in developmental sample)

Orthagonal and oblique

> 0.40

1.0 (in developmental sample)

Forced into five models with 3 syndrome factors

29 of any of the three syndrome factors

Impaired cognition; Confusion-ataxia; Central pain

Compared fit of factor analysis with that found in earlier Seabee study (Haley 1997b) using structural estimating equations. Some models also fitted higher order factor “Gulf War syndrome” and loaded four additional symptoms (chronic fatigue involving excessive muscle weakness, chronic fever and night sweats, middle and terminal insomnia, chronic watery diarrhea) onto

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

Reference

Population

Type of Data

Method

Rotation

Factor Loading Cutoff and Eigenvalue Cutoff

No. Factors Isolated

Percentage of Variance Explained

Factors Identified

Unique Factors in Deployed Veterans?

 

 

 

 

 

 

 

 

 

higher-order factor

Bourdette et al. 2001 (secondary)

Active and reserve n = 443

Dichotomous

Principal components

Orthagonal

0.30

> 1.0

3

34.2

Cognitive/psychologic; Mixed somatic; Musculoskeletal

NA

Shapiro et al. 2002 (secondary)

Active and retired n = 1779

Dichotomous

Principal components

Orthagonal

> 1.0

> 0.60

3

46.7 among exposed, 49.8 among non-exposed deployed, 52.2 among non-deployed

Cognitive-psychologic; Dysesthesia; Vestibular dysfunction

Higher odds of having witnessed demolition among those reporting dysesthesia

Hallman et al. 2003 (secondary)

Retired, participants in VA Gulf War Health Registry n = 1,161

Ordinal

Principal axis

Oblique

Not stated

4

50.2

Mood-memory-fatigue; musculoskeletal; gastrointestinal; throat-breathing

NA

NOTE: GW = Gulf War, GWV = Gulf War Veteran; VA = Department of Veterans Affairs.

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

INJURY AND EXTERNAL CAUSES OF MORBIDITY AND MORTALITY (ICD-10 S00-Y98)

Primary Studies

The first large mortality study of nearly all Gulf War-deployed veterans (n = 695,516) identified no excess postwar mortality, from all causes combined, compared with nondeployed veterans (n = 746,291). One particular cause of mortality, from motor-vehicle accidents, was somewhat higher (RR 1.31, 95% CI 1.14-1.49), but the risk was lower than the expected rate based on overall US mortality (SMR 0.82, 95% CI 0.75-0.89) after adjustment for age, sex, race, and year of death (Kang and Bullman 1996). The study examined mortality patterns from 1991 through 1993 by using two databases: the VA Beneficiary Identification and Records Locator Subsystem (BIRLS) and deaths reported to the Social Security Administration.20 It compared deployed veterans with a cohort of similar size of veterans who did not serve in the Gulf War. Adjustments were made for age, race, marital status, branch of service, and type of unit. It also found no increase in suicide or homicide among Gulf War veterans. The second publication by the authors found that by 1994, the excess mortality risk from motor-vehicle accidents had disappeared. That finding is consistent with the mortality pattern after the Vietnam War (CDC 1987; Thomas et al. 1991; Watanabe and Kang 1995). The study found no overall differences in mortality between deployed and nondeployed veterans; additionally the mortality risk in both the deployed and nondeployed was less than half of what was expected in their civilian counterparts (Kang and Bullman 2001).

A study of all UK veterans of the Gulf War (n = 53,462) in relation to contemporaneous controls found no increase in mortality other than a small and nonsignificant increase in accidental death (RR 1.18, 95% CI 0.98-1.42) (Macfarlane et al. 2000). In that study, controls (n = 53,450) were matched by sex, age, branch of service, and level of fitness in an attempt to control for the healthy-warrior effect. Accidental-death increases were due primarily to motor-vehicle accidents (RR 1.25, 95% CI 0.91-1.72), air and space accidents (RR 1.77, 95% CI 0.86-3.81), and accidents caused by submersion, suffocation, or ingestion of foreign bodies (RR 3.25, 95% CI 1.00-13.69). There were no increases in suicide or homicide. The study covered the years 1991-1999 and did not address changes over time in excess external cause (or motor-vehicle) mortality.

As part of continuing mortality surveillance, the UK Defence Analytical Services Agency periodically publishes its cumulative mortality figures for deployed veterans vs Gulf War-era controls. From 1991 to June 30, 2005, there was no increase in mortality other than a small and nonsignificant increase in transportation accidents, which include land, water, and air accidents (SMR 1.21, 95% CI 0.96-1.51). “Other external causes of accidental injury”—a category including falls, drowning, and poisoning—also showed a small and nonsignificant increase (SMR 1.07, 95%CI 0.74-1.54) (Defence Analytical Services Agency 2005). Compared with earlier surveillance, those data show that the differences between deployed veterans and Gulf War-era controls in deaths from external causes disappeared about 10 years after the war. There also were a small and statistically nonsignificant increase in intentional self-harm (SMR 1.08,

20

The degree of completeness of these record systems was assessed with a validation study that used state vital-statistics data. Ascertainment was estimated at 89% of all deaths in the Gulf War cohort and comparison group.

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

95% CI 0.85-1.39) and a small and statistically nonsignificant reduction in risk of death from assault (SMR 0.46, 95% CI 0.15-1.38).

Finally, one hospitalization study addresses the question of transportation-related injuries. A study of armed services personnel on active duty during the Gulf War was conducted after the war (1991-1994) at DOD, VA, and nonfederal hospitals in California (Gray et al. 2000). The purpose of including the latter type of hospitalization was to eliminate potential bias related to veterans’ seeking care outside DOD and VA facilities. The authors found increased rates of hospitalization for the category “injury and poisoning” in DOD hospitals (PMR 1.03, 95% CI 1.01-1.05) and California hospitals (PMR 1.11, 95% CI 1.04-1.18), but not in VA hospitals. Table 5.20 summarizes the results of the primary mortality studies.

Secondary Studies

A mortality study of US active-duty military personnel focused exclusively on the Gulf War period of Operations Desert Storm and Desert Shield (1990-1991). It compared noncombat mortality among troops stationed in the Gulf War and troops on active duty elsewhere. There was no excess noncombat mortality in deployed veterans except for unintentional injury due to vehicle accidents and other causes (Writer et al. 1996). A similar study of noncombat injuries in navy and marine personnel during the Gulf War found that most patient visits were for injuries and poisonings21 (Shaw et al. 1991).

A post-Gulf War population-based study of Iowa veterans (1995-1996) found that self-reported traumatic injuries were more likely in deployed than in nondeployed veterans (Zwerling et al. 2000). However, in a large UK study, self-reported “accidental injuries” were lower in Gulf War veterans than in nondeployed veterans, but the category was very broad (Simmons et al. 2004); the authors interpreted this unexpected result as due to inclusion of both major and minor injuries and to the possibility that Gulf War veterans are more inclined to report illnesses than injuries because of the belief that illnesses are associated with deployment.

Gackstetter et al. (2002) performed a nested case-control study of the large Gulf War-deployed and nondeployed population assembled by Kang and colleagues (1996). Deployed veterans who died in motor-vehicle accidents through 1995 (n = 1,343) were more likely to be male, younger, less educated, and never married than nondeployed controls (10 controls/case). They were also more likely to be enlisted, have combat occupations, and be in the National Guard or reserves and not in the Air Force. One of the datasets used by Gackstetter et al. (2002) examined prior morbidity patterns to determine underlying physical and mental health among 980 deployed veterans and nondeployed veterans who died in motor-vehicle accidents. After adjustment for demographic factors and military characteristics, the authors found that prior treatment for mental-health problems, particularly drug or alcohol abuse was strongly associated with such deaths, particularly among nondeployed veterans. The only predictor of motor-vehicle deaths among deployed veterans was prior motor-vehicle injury.

Summary and Conclusion

The committee found that various studies have looked at mortality in Gulf War veterans but have numerous limitations. The principal limitation is the short duration of their followup

21

As noted earlier, poisoning is one of the standard codes used on death certificates under “other external causes of accidental injury”. The category includes falls, drowning, and poisoning.

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

observation period. More time must elapse before investigators will be able to assess increased mortality that would result from illnesses with long latency, such as cancer, or with a gradually deteriorating course, such as cardiovascular disease. Another potential limitation in comparing deployed and nondeployed personnel is the healthy-warrior effect. That might result in selection bias, insofar as chronically ill or less fit members of the armed forces might be less likely than more fit members to have been deployed. Thus, there might have been nonrandom assignment of those selected for deployment and not selected for deployment. That is demonstrated by the excess number of deaths from HIV infection among nondeployed veterans reported by Kang and Bullman (2001).

Some studies provide evidence of a modest increase in transportation-related injuries among deployed compared with nondeployed Gulf War veterans in the decade immediately after deployment. That increase in mortality appears to have been restricted to the first several years after the war.

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

TABLE 5.20 Mortality and Injury Studies

Study

Population

Outcomes

Results

Adjustment

Comments or Limitations

Kang and Bullman 1996; Kang and Bullman 2001

695,516 Gulf War veterans vs 746,291 non-Gulf War veterans

Mortality 1991-1997; Cox proportional hazards models

Increased deaths from motor-vehicle accidents in Kang and Bullman 1996 (RR 1.31, 95% CI 1.14-1.49)

RRs became nonsignificant in Kang and Bullman 2001 (RR 1.17, 95% CI 0.98-1.4) in 1994-1995;

Increased HIV deaths in non-Gulf War veterans; no difference in potential nerve gas exposure; no homicide or suicide increase

Sex, age, race, marital status, branch of service, type of unit

Short duration of followup; healthy warrior effect may obscure difference

Macfarlane et al. 2000

53,462 Gulf War veterans vs 53,450 Gulf War-era cohort, UK

Mortality 1991-1999

Higher mortality in Gulf War veterans from external causes (RR 1.18, 95% CI 0.98-1.42); no increase in homicide or suicide

Matching by sex, age, branch, fitness for service

 

Defence Analytical Services Agency 2005

53,409 Gulf War veterans vs 53,143 Gulf War-era cohort, UK

Mortality 1991-June 2005

No increase in mortality except small and nonsignificant increase in “transport accidents” (SMR 1.21, 95% CI 0.96-1.51); “other external causes of accidental injury” (SMR 1.07, 95% CI 0.74-1.54); higher deaths from external causes disappeared about 10 years after Gulf War

Matching by sex, age, branch

 

Gray et al. 2000

652,979 Gulf War veterans vs 652,922 randomly selected nondeployed veterans

Morbidity 1991 to 1994

Increased rates of hospitalizations for the “injury and poisoning” in DOD hospitals (PMR 1.03, 95% CI 1.01-1.05) and California (PMR 1.11, 95% CI 1.04-1.18); decreased rates for VA hospitals (PMR 0.89, 95% CI 0.83-0.96)

DOD hospitals adjusted for age, sex, and race; California and VA hospitals adjusted for age and sex

 

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

ALL-CAUSE HOSPITALIZATION STUDIES

This section concentrates on all-cause hospitalization to determine whether there is an excess risk of hospitalization among Gulf War veterans. Hospitalizations for specific causes, although noted here, are discussed in more detail throughout the report. The primary studies are summarized in Table 5.21.

Primary Studies

Studies of differences in rates of hospitalization between deployed and nondeployed populations can indicate excess morbidity associated with Gulf War service. Although they are less able to detect subtle differences than studies that measure morbidity directly or examine outpatient morbidity, they are less crude than studies of differential mortality. Overall hospitalization (that is, for all causes) and cause-specific hospitalization were the subject of several large studies, mostly of active-duty personnel discharged from DOD hospitals.

The first study (1991-1993) compared the hospitalizations of almost 550,000 Gulf War veterans and almost 620,000 nondeployed veterans and found no consistent differences over time in all-cause hospitalizations after the war (Gray et al. 1996). There were increased rates of hospitalization of Gulf War veterans in some diagnostic categories in some years (for example, neoplasms in 1991 and diseases of the blood in 1992), but the rates were not consistently increased, except rates of hospitalization for mental illness in 1992 and 1993. The study also found increased hospitalization, in 1991 only, for the broad category “genitourinary system diseases”. The authors found, more specifically, that the increase was due to female veterans being hospitalized for inflammatory diseases of the ovary, fallopian tube, pelvic cellular tissue, and peritoneum. Those increases could be explained by deferral of care, postwar pregnancies, and some psychiatric disorders (alcohol dependence, nondependent drug abuse, and adjustment reactions). The study also examined reasons for separation from the armed services in 1991-1993. Contrary to expectations, the study found that deployed veterans were less likely than nondeployed veterans to have separated for reasons of medical disqualification, dependence or hardship, entry into officer programs, retirement, or behavior or performance failure.

A second hospitalization study extended the study period (1991-1996) and re-examined the dataset to search for excess hospital admissions for unexplained illnesses (Knoke and Gray 1998). The authors reasoned that the first study might have missed hospitalizations for a new or poorly recognized syndrome.

In the third study, Gray et al. (2000) examined hospitalizations (1991-1994) of active-duty, reserve, and former military personnel who had been deployed to the Gulf War. The study examined hospitalizations at DOD, VA, and nonfederal California hospitals to eliminate potential bias related to veterans’ seeking care outside DOD and VA facilities. Because of the unreliability of state-of-residence data in DOD and VA datasets, the authors could not directly compare rates of hospitalization among the three sources. Rather, they compared PMRs of hospitalization-discharge diagnoses (14 diagnostic categories from ICD-9) in Gulf War-deployed and nondeployed veterans. PMRs of most disease categories were not increased. However, four categories were increased in VA patients (but not in active-duty military or California veterans): respiratory (PMR 1.19, 95% CI 1.10-1.29), digestive (PMR 1.12, 95% CI 1.05-1.18), skin (PMR 1.14, 95% CI 1.00-1.27), and ill-defined diseases (PMR 1.24, 95% CI 1.16-1.33). Among

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

respiratory diseases, the authors reported increases in asthma, but no data were shown. The study is antecedent to a more detailed study of respiratory hospitalizations in relation to exposure to smoke from oil-well fires (Smith et al. 2002). The authors also found increased rates of hospitalization for the category “injury and poisoning” in DOD (PMR 1.03, 95% CI 1.01-1.05) and California hospitals (PMR 1.11, 95% CI 1.04-1.18). Each finding is discussed in the relevant section of this chapter.

Other hospitalization studies are reviewed in this chapter and address specific causes of hospitalizations, for example, in relation to exposure to oil-well fires and respiratory outcomes (Smith et al. 2002) or exposure to nerve agents and specific hospitalizations (Gray et al. 1999b; Smith et al. 2003). For the most part, the studies did not find increased hospitalization in relation to the exposures. A study that did find an increase in hospitalizations for acute psychiatric disorders (Dlugosz et al. 1999) is discussed in the psychiatric section. Table 5.22 summarizes the results of the all-cause hospitalization studies.

Summary and Conclusion

The all-cause hospitalization studies provide some reassurance that excess hospitalizations did not occur among veterans of the Gulf War who remained on active duty through 1994. The studies, however, have several limitations, including that they were largely of active-duty personnel and cannot be generalized to the entire cohort of Gulf War veterans, inasmuch as it been noted that Gulf War veterans who left the military reported worse health outcomes than those who remained (Ismail et al. 2000). As is the case for mortality studies, it is too soon to capture hospitalizations from illnesses that might have longer latency, such as some cancers. In addition, hospitalization data might be incomplete on people separated from the military and admitted to nonmilitary (VA and civilian) hospitals. The studies did not measure the use of outpatient treatment and thus detected only illnesses that required hospitalization.

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

TABLE 5.21 All-Cause Hospitalization Studies

Study

Population

Outcomes

Results

Adjustment

Comments orLimitations

Gray et al. 1996

579,931 US Gulf War veterans who were on regular active duty vs about 700,000 randomly selected controls; of these, 1,165,411 had complete data for 1991 (with losses ineach later year)

Hospitalization records: DOD only, 1991-1993

Increased rates of hospitalizations for respiratory disease (1991); genitourinary disease (1991), neoplasm (1991), blood (1992), mental disorders (1992-1993); decrease drates for infection (1991-1992), endocrine or metabolic (1991-1993), circulatory (1992), digestive (1991-1993), musculoskeletal (1991-1992), ill-defined conditions (1991-1992)

Gulf War service, sex, age, race, marital status, branch and length of service, occupation, rank, salary

Active duty only, no assessment of outpatient treatment, respiratory findingsremoved after adjustment for VA screening-program attendance

Knoke and Gray 1998

552,111 deployed vs 1,479,751 nondeployed service members in service during Gulf War and remaining there through 1996

Hospitalization records: DOD only, 1991-1996, ICD 799.9 (unexplained illness)

No excess in hospitalizations in this period when effect of CCEP was eliminated

Race, rank, salary, military branch, occupation, prewar hospitalization, sex

Active duty only, no assessment of outpatient treatment, respiratory findingsremoved after adjustment for VA screening-program attendance

Gray et al. 2000

652,979 veterans vs 652,922 random controls; active duty plus National Guard

Hospitalization records: DOD plus VA plus California hospitals, 1991-1994

Increased PMRs: DOD, injury and poisoning; VA, respiratory, digestive, ill-defined conditions; California, injury and poisonings

Stratified by age, sex, and ethnicity

Could not identify multiple admissions for same subject across databases

Gray et al. 1999b

124,487 Army Gulf War veterans(active duty and National Guard) with possible low exposure to chemical munitions vs 224,804 other Army Gulf War veterans deployed at same time

Postwar hospitalizations, 1991-1995

Groupwith low modeled exposure to nerve agent had highest rate of hospitalizations for all causes and for neoplasms

Prewar hospitalization, reserve, sex, age group, marital status, race, pay grade, occupation

Modeling of exposures, short time of followup

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

Study

Population

Outcomes

Results

Adjustment

Comments or Limitations

Smith et al. 2003

431,762 regular and reserve personnel in Army and Air Force in gulf during Khamisiyah demolitions

Postwar morbidity, 1991-2000

Cardiac arrythmias for nerve-agent exposure (RR 1.23, CI 1.04-1.44)

No increase for other agents

Sex, age, status, prewar hospitalization, pay grade, race, branch, days deployed, marital status, occupation

NOTE: CCEP = Comprehensive Clinical Evaluation Program; DOD =Department of Defense; VA = Department of Veterans Affairs.

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

MULTIPLE CHEMICAL SENSITIVITY

Multiple chemical sensitivity22 is a controversial condition that can be loosely defined by a person’s inability to tolerate multiple chemically unrelated compounds. Although it has been described by physicians since the 1950s, major medical associations have questioned the existence of MCS (American Academy of Allergy 1999; American College of Physicians 1989; American Medical Association 1992). In contrast, a recent evaluation of the biomedical literature commissioned at the request of the UK Health and Safety Executive, found “suggestive” evidence that MCS exists (Graveling et al. 1999). Still, there are no pathologic or laboratory tests and it is often a diagnosis reached by exclusion when no other cause for the symptoms can be identified. There is no validated questionnaire for this symptom complex.

Researchers have developed a set of criteria for the diagnosis of MCS (for example, the Cullen criteria23). There are an array of symptoms such as fatigue, cognitive impairment, and headaches) that might be elicited by relatively low concentrations of chemicals with diverse structures and mechanisms of action. For example, symptomatic individuals often report that their symptoms are caused and later triggered by exposure to pesticides, fuels, combustion products, perfumes and other chemical agents (Caress et al. 2002; Kipen and Fiedler 2002). People, including Gulf War veterans, who have MCS symptoms report functional impairment and disability (Black et al. 1999; Fiedler et al. 1996; Jason et al. 2000). About 2-6% of the US population reports having MCS according to various definitions used in population-based studies (Caress and Steinemann 2003; Caress et al. 2002; Kreutzer et al. 1999).

Primary Studies

There is no validated model for MCS and no universally adopted definition of it; therefore, for the purposes of this section, a primary study is one that includes a comparison group, an appropriate questionnaire—that is, a study that uses generally recognized criteria for MCS. Primary studies are summarized in Table 5.22.

Using previously collected symptom reporting from their population-based UK cohort, Reid and colleagues (2001) estimated the prevalence of MCS and CFS and their relationships to Gulf War exposures. (The findings on CFS are presented earlier in this chapter.) One control group was veterans deployed to Bosnia, and the other was a group of Gulf War-era veterans who were deployed elsewhere. In contrast with some studies, the nondeployed control groups were recruited from among the subset of nondeployed service members who were fit for combat duty; this avoided selection bias from the healthy-warrior effect. A case of MCS was defined by using symptom criteria of Simon and colleagues (1993). The prevalence of MCS in deployed veterans was 1.3% (95% CI 1.0-1.7)—a higher figure than that in the two comparison groups, the Bosnia

22

Multiple chemical sensitivity is not listed as a condition in the ICD-10.

23

Cullen's definition is widely used. It includes four elements: (1) the syndrome is acquired after a documented environmental exposure that might have caused objective evidence of health effects; (2) the symptoms are referable to multiple organ systems and vary predictably in response to environmental stimuli; (3) the symptoms occur in relation to measurable concentrations of chemicals, but the concentrations are below those known to harm health; and (4) no objective evidence of organ damage can be found (Cullen 1987).

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

and Gulf War-era cohorts. The OR of MCS in Gulf War-deployed vs Bosnia-deployed was 4.5 (95% CI 1.7-11.8), and in Gulf War-deployed vs Gulf War era-deployed was 7.2 (95% CI 2.8-18.2). MCS was associated with the majority of the exposures, but self-reported pesticide exposure was among the strongest. Limitations of the study were self-reported symptoms and exposures. The study on which the data are based (Unwin et al. 1999) estimated the prevalence of MCS at 0.8% in Gulf War veterans, a prevalence similar to that in Bosnia veterans (0.4%) and Gulf War-era veterans (0.3%). The differences between the three cohorts were not significant. The Unwin study did not construct a case of MCS from symptom criteria, like that of Reid et al., but rather asked respondents to self-report their medical disorders, one of which was “multiple chemical sensitivity”.

As part of the large population-based Iowa study, Black and colleagues (2000) sought to determine symptom prevalence of and risk factors for MCS. The case criteria were developed by expert consensus. A total of 3,695 veterans were surveyed with structured telephone interviews in 1995-1996. The details of the study are described in Chapter 4. The prevalence of MCS was 5.4% among deployed and 2.6% among nondeployed. Determined through multivariate analysis, the independent risk factors for developing MCS were deployment to the Gulf War, numerous sociodemographic factors (such as, age, male sex, marital status, and education), psychiatric history, and current psychiatric conditions. The sample size was a strength of the study, as was the use of an expert consensus working case definition as described in the study.

As part of the same study, Black and colleagues (1999) surveyed veterans to determine the effect of MCS on disability. Of the total sample of 169 subjects who met their case definition of MCS, the authors found high levels of disability in comparison to those (n = 3,526) who did not meet their case criteria. MCS cases reported more than 12 days in bed due to disability (OR 3.2, 95% CI 1.7-6.3), receipt of VA disability status (OR 3.5, 95% CI 2.1-5.9), receipt of VA disability compensation (OR 3.9, 95% CI 1.9-7.8), receipt of medical disability status (OR 7.3, 95% CI 1.0-50.9), and unemployment (OR 9.8, 95% CI 4.8-20.1). Adjustments were made for age, sex, branch of military, rank, and whether a veteran had regular military or reserve status. A study limitation is that it did not have external validation of disability status.

Canada deployed more than 3,000 sea, land, and air forces to the gulf region. A large proportion of them participated in a naval blockade and they were responsible for one-fourth of enemy interceptions in the gulf. A survey of the entire cohort found that deployed veterans were 4 times as likely as nondeployed veterans (OR 4.01, 95% CI 2.43-6.62) to report symptoms of MCS (Goss Gilroy Inc. 1998).

Secondary Studies

Studying an Army cohort at Fort Devens (n = 180) in 1995, Proctor and colleagues (2001) conducted in-person interviews to determine the prevalence of presumptive MCS, chemical sensitivity, and CFS. The comparison population (n = 46) was an air ambulance company deployed to Germany during the Gulf War. During an environmental interview, if subjects reported health symptoms triggered by chemical odors, they were questioned to determine whether they met MCS case criteria adapted from Cullen (1987). The subjects were not excluded for having self-reported asthma, as well as concurrent psychiatric diagnosis (by psychiatric diagnostic interviews via the SCID). Among deployed veterans, there was a nonsignificant increased prevalence of presumptive MCS of 2.9% vs 0% among nondeployed. An important limitation of the study is the small sample; a strength was that it adopted the Cullen criteria for MCS.

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

In 1999, Gray and colleagues (2002) surveyed all Seabees (n = 18,945) who had been on active duty during the time of the Gulf War regardless of whether they remained on active duty, were in the reserve, or had separated from the service. There were 11,868 respondents, who were divided into three groups: 3,831 Seabees deployed to the Gulf War, 4,933 Seabees deployed elsewhere, and 3,104 Seabees not deployed. MCS was included in a checklist of about 20 physician-diagnosed conditions about which veterans were asked. In comparison with nondeployed Seabees, the Gulf War-deployed Seabees reported being more than 4 times as likely to have been given an MCS diagnosis (OR 4.47, 95% CI 2.30-8.69). The odds were similarly increased when Gulf War-deployed veterans were compared with those deployed elsewhere (OR 4.08, 95% CI 2.29-7.24). The strength of this study was the large and homogeneous population.

Summary and Conclusion

Overall, the rates of MCS are similar in deployed and civilian populations. MCS or MCS-like symptoms have, as noted earlier, neither a validated questionnaire nor a standard definition. Diagnosis is often by exclusion. Several large or population-based studies of Gulf War veterans found, by questionnaire, that the prevalence of MCS-like symptoms ranged from 2-6%. Most studies found that the prevalence in Gulf War veterans was about 2-4 times higher than that in nondeployed veterans. However, none of the primary studies used the same definition, so it is difficult to compare them. Furthermore, none performed medical evaluations. Although direct physician interaction with patients is always preferable, it should be recognized that there are no physical findings with MCS.

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

TABLE 5.22 Multiple Chemical Sensitivity (MCS)

Study

Design

Population

Outcomes

Results

Adjustments

Comments

Reid et al. 2001

Population-based, cross-sectional, prevalence

3,531 Gulf war-deployed vs 2,050 Bosnia-deployed vs 2,614 Gulf War-era deployed

Symptoms meeting criteria of Simon et al. 1993 by questionnaire

Gulfvs Bosnia: OR 4.5 (95% CI 1.7-11.8); Gulf vs Gulf War-era: OR 7.2 (95% CI 2.8-18.2)

Sex, age, marital status, education, rank, employment status on followup

Self-reported symptoms and functioning, unvalidated case definition

Unwin et al. 1999

Population-based, cross-sectional, prevalence

4,248 Gulf war-deployed vs 4,250 Bosnia-deployed vs 4,246 Gulf War-era deployed

Self-reported medical condition on questionnaire

Gulf vs Bosnia: OR 1.9 (95% CI 0.8-4.4); Gulf vs Gulf War-era: OR 2.2 (95% CI 1.0-4.9)

Age, smoking, alcohol consumption, marital status, educational attainment, officer or other rank, employment status, civilian or military status

 

Black et al. 2000

Population-based, cross-sectional, prevalence

1,896 deployed vs 1,799 nondeployed

Structured telephonesurvey, case criteria for MCS by expert consensus

Independent risk factors for MCS by multivariate analysis, significant at 5%: deployment to Gulf War (OR 1.94); age, male sex, rank, branch of service, previous psychiatric treatment (OR 2.31); current mental illness

Age, sex, branch of military, rank

Strengths are large sample, use of an expert consensus, working case definition

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

Study

Design

Population

Outcomes

Results

Adjustments

Comments

Black et al. 1999

Population-based, cross-sectional, prevalence

169 with MCS vs 3,526 with out MCS

Structured telephone survey, case criteria for MCS by expert consensus, SF-36

>12 days in bed to disability (OR 3.2 95%, CI 1.7-6.3); VA disability status (OR 3.5, 95% CI 2.1-5.9); VA disability compensation (OR3.9, 95% CI 1.9-7.8); medical disability (OR 7.3, 95% CI 1.0-50.9); unemployment (OR 9.8, 95% CI 4.8-20.1); greater health services use

Age, sex, branch of military, rank, regular military or reserve status

No external validation of disability status

Goss Gilroy Inc. 1998

Population-based, cross-sectional, prevalence

3,113 deployed vs 3,439 nondeployed

Algorithm requiring physical illness with routine exposure to several substances, atleast two substances making person ill, avoidance of atleast one activity, positive response to two different sets of eight systemic symptoms, cognitive dysfunction

OR 4.01 (95% CI2.43-6.62)

Rank, income, age

Self-reported symptoms, unvalidated case definition

NOTE: VA = Department of Veterans Affairs.

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

REFERENCES

Abul AT, Nair PC, Behbehanei NA, Sharma PN. 2001. Hospital admissions and death rates from asthma in Kuwait during pre- and post-Gulf War periods. Annals of Allergy, Asthma, and Immunology 86(4):465-468.

Abushaban L, Al-Hay A, Uthaman B, Salama A, Selvan J. 2004. Impact of the Gulf war on congenital heart diseases in Kuwait. International Journal of Cardiology 93(2-3):157-162.

ACS (American Cancer Society). 2006. Statistics for 2006. [Online]. Available: http://www.cancer.org/docroot/stt/stt_0.asp [accessed March 21, 2006].

Al-Khalaf B. 1998. Pilot study: The onset of asthma among the Kuwaiti population during the burning of oil wells after the Gulf War. Environment International 24(1-2):221-225.

The ALS Association. 2006. The ALS Association. [Online]. Available: http://www.alsa.org/; [accessed January 17, 2006].

Amato AA, McVey A, Cha C, Matthews EC, Jackson CE, Kleingunther R, Worley L, Cornman E, Kagan-Hallet K. 1997. Evaluation of neuromuscular symptoms in veterans of the Persian Gulf War. Neurology 48(1):4-12.

American Academy of Allergy, Asthma, and Immunology (AAAAI). 1999. Idiopathic environmental intolerances. Journal of Allergy and Clinical Immunology 103(1 Pt 1):36-40.

American College of Physicians. 1989. Clinical ecology. Annals of Internal Medicine 111 (2):168-178.

American Medical Association. 1992. Clinical ecology Journal of the American Medical Association 268(24):3465-3467.

Anger WK. 2003. Neurobehavioural tests and systems to assess neurotoxic exposures in the workplace and community. Occupational and Environmental Medicine 60(7):531-538, 474.

Anger WK, Storzbach D, Binder LM, et al. 1999. Neurobehavioral deficits in Persian Gulf veterans: Evidence from a population-based study. Journal of the International Neuropsychological Society. 5(3):203-212.

Annegers JF, Appel SH, Perkins P, Lee J. 1991. Amyotrophic lateral sclerosis mortality rates in Harris County, Texas. Advances in Neurology 56:239-243.

Araneta MR, Moore CA, Olney RS, Edmonds LD, Karcher JA, McDonough C, Hiliopoulos KM, Schlangen KM, Gray GC. 1997. Goldenhar syndrome among infants born in military hospitals to Gulf War veterans. Teratology 56(4):244-251.

Araneta MR, Destiche DA, Schlangen KM, Merz RD, Forrester MB, Gray GC. 2000. Birth defects prevalence among infants of Persian Gulf War veterans born in Hawaii, 1989-1993. Teratology 62(4):195-204.

Araneta MR, Schlangen KM, Edmonds LD, Destiche DA, Merz RD, Hobbs CA, Flood TJ, Harris JA, Krishnamurti D, Gray GC. 2003. Prevalence of birth defects among infants of Gulf War veterans in Arkansas, Arizona, California, Georgia, Hawaii, and Iowa, 1989-1993. Birth Defects Research 67(4):246-260.

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

Araneta MR, Kamens DR, Zau AC, Gastanaga VM, Schlangen KM, Hiliopoulos KM, Gray GC. 2004. Conception and pregnancy during the Persian Gulf War: The risk to women veterans. Annals of Epidemiology 14(2):109-116.

Armon C. 2003. An evidence-based medicine approach to the evaluation of the role of exogenous risk factors in sporadic amyotrophic lateral sclerosis. Neuroepidemiology 22(4):217-228.

Armon C. 2004a. Amyotrophic Lateral Sclerosis. In: Nelson LM, Tanner CM, Van Den Eeden SK, McGuire VM, Editors. Neuroepidemiology: From Principles to Practice. New York: Oxford University Press. Pp. 162-187.

Armon C. 2004b. Occurrence of amyotrophic lateral sclerosis among Gulf War veterans. Neurology 62(6):1027.

Axelrod BN, Milner IB. 1997. Neuropsychological findings in a sample of Operation Desert Storm veterans. Journal of Neuropsychiatry and Clinical Neurosciences 9(1):23-28.

Barrett DH, Doebbeling CC, Schwartz DA, Voelker MD, Falter KH, Woolson RF, Doebbeling BN. 2002. Posttraumatic stress disorder and self-reported physical health status among US Military personnel serving during the Gulf War period: A population-based study. Psychosomatics 43(3):195-205.

Binder LM, Storzbach D, Anger WK, Campbell KA, Rohlman DS. 1999. Subjective cognitive complaints, affective distress, and objective cognitive performance in Persian Gulf war veterans. Archives of Clinical Neuropsychology 14(6):531-536.

Black DW, Carney CP, Forman-Hoffman VL, Letuchy E, Peloso P, Woolson RF, Doebbeling BN. 2004a. Depression in veterans of the first gulf war and comparable military controls. Annals of Clinical Psychiatry 16(2):53-61.

Black DW, Carney CP, Peloso PM, Woolson RF, Schwartz DA, Voelker MD, Barrett DH, Doebbeling BN. 2004b. Gulf War veterans with anxiety: Prevalence, comorbidity, and risk factors. Epidemiology 15(2):135-142.

Black DW, Doebbeling BN, Voelker MD, Clarke WR, Woolson RF, Barrett DH, Schwartz DA. 1999. Quality of life and health-services utilization in a population-based sample of military personnel reporting multiple chemical sensitivities. Journal of Occupational and Environmental Medicine 41(10):928-933.

Black DW, Doebbeling BN, Voelker MD, Clarke WR, Woolson RF, Barrett DH, Schwartz DA. 2000. Multiple chemical sensitivity syndrome: Symptom prevalence and risk factors in a military population. Archives of Internal Medicine 160(8):1169-1176.

Blanchard EB, Jones-Alexander J, Buckley TC, Forneris CA. 1996. Psychometric properties of the PTSD Checklist (PCL). Behaviour Research and Therapy 34(8):669-673.

Blanchard MS, Eisen SA, Alpern R, Karlinsky J, Toomey R, Reda DJ, Murphy FM, Jackson LW, Kang HK. 2006. Chronic Multisymptom Illness Complex in Gulf War I Veterans 10 Years Later. American Journal of Epidemiology 163(1):66-75.

Bombardier CH, Buchwald D. 1996. Chronic fatigue, chronic fatigue syndrome, and fibromyalgia. Disability and health-care use. Medical Care 34(9):924-930.

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

Bourdette DN, McCauley LA, Barkhuizen A, Johnston W, Wynn M, Joos SK, Storzbach D, Shuell T, Sticker D. 2001. Symptom factor analysis, clinical findings, and functional status in a population-based case control study of Gulf War unexplained illness. Journal of Occupational and Environmental Medicine 43(12):1026-1040.

Brailey K, Vasterling JJ, Sutker PB. 1998. Psychological Aftermath of Participation in the Persian Gulf War. Lundberg A, Editor. The Environment and Mental Health: A Guide for Clinicians. London: Lawrence Erlbaum Associates. Pp. 83-101.

Brown RG, Scott LC, Bench CJ, Dolan RJ. 1994. Cognitive function in depression: Its relationship to the presence and severity of intellectual decline. Psychological Medicine 24(4):829-847.

Buchwald D, Garrity D. 1994. Comparison of patients with chronic fatigue syndrome, fibromyalgia, and multiple chemical sensitivities. Archives of Internal Medicine 154(18):2049-2053.

Buchwald D, Pearlman T, Umali J, Schmaling K, Katon W. 1996. Functional status in patients with chronic fatigue syndrome, other fatiguing illnesses, and healthy individuals. American Journal of Medicine 101(4):364-370.

Bullman TA, Mahan CM, Kang HK, Page WF. 2005. Mortality in US Army Gulf War Veterans Exposed to 1991 Khamisiyah Chemical Munitions Destruction. American Journal of Public Health 95(8):1382-1388.

Bunegin L, Mitzel HC, Miller CS, Gelineau JF, Tolstykh GP. 2001. Cognitive performance and cerebrohemodynamics associated with the Persian Gulf Syndrome. Toxicology and Industrial Health 17(4):128-137.

California Birth Defects Monitoring Program. 2006. Discoveries and Data Center. [Online]. Available: http://www.cbdmp.org/bd_intro.htm [accessed April 10, 2006].

Caress SM, Steinemann AC. 2003. A review of a two-phase population study of multiple chemical sensitivities. Environmental Health Perspectives 111(12):1490-1497.

Caress SM, Steinemann AC, Waddick C. 2002. Symptomatology and etiology of multiple chemical sensitivities in the southeastern United States. Archives of Environmental Health 57(5):429-436.

CDC (Centers for Disease Control). 1987. Postservice mortality among Vietnam veterans. Morbidity and Mortality Weekly Report 36(5):61-64.

CDC. 1995. Unexplained illness among Persian Gulf War veterans in an Air National Guard Unit: Preliminary report--August 1990-March 1995. Journal of the American Medical Association 274(1):16-17.

Chalder T, Berelowitz G, Pawlikowska T, Watts L, Wessely S, Wright D, Wallace EP. 1993. Development of a fatigue scale. Journal of Psychosomatic Research 37(2):147-153.

Cherry N, Creed F, Silman A, Dunn G, Baxter D, Smedley J, Taylor S, Macfarlane GJ. 2001a. Health and exposures of United Kingdom Gulf war veterans. Part II: The relation of health to exposure. Occupational and Environmental Medicine 58(5):299-306.

Cherry N, Creed F, Silman A, Dunn G, Baxter D, Smedley J, Taylor S, Macfarlane GJ. 2001b. Health and exposures of United Kingdom Gulf war veterans. Part I: The pattern and extent of ill health. Occupational and Environmental Medicine 58(5):291-298.

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

Chio A, Benzi G, Dossena M, Mutani R, Mora G. 2005. Severely increased risk of amyotrophic lateral sclerosis among Italian professional football players. Brain 128(Pt 3):472-476.

Coffman CJ, Horner RD, Grambow SC, Lindquist J. 2005. Estimating the occurrence of amyotrophic lateral sclerosis among Gulf War (1990-1991) veterans using capture-recapture methods. Neuroepidemiology 24(3):141-150.

Cohen J. 1992. A power primer. Psychological Bulletin 112:155-159.

Coombe MD, Drysdale SF. 1993. Assessment of the effects of atmospheric oil pollution in post war Kuwait. Journal of the Royal Army Medical Corps 139(3):95-97.

Cowan DN, DeFraites RF, Gray GC, Goldenbaum MB, Wishik SM. 1997. The risk of birth defects among children of Persian Gulf War veterans. New England Journal of Medicine 336(23):1650-1656.

Cowan DN, Lange JL, Heller J, Kirkpatrick J, DeBakey S. 2002. A case-control study of asthma among U.S. Army Gulf War veterans and modeled exposure to oil well fire smoke. Military Medicine 167(9):777-782.

Cullen MR. 1987. The worker with multiple chemical sensitivities: An overview. Occupational Medicine 2(4):655-661.

David AS, Farrin L, Hull L, Unwin C, Wessely S, Wykes T. 2002. Cognitive functioning and disturbances of mood in UK veterans of the Persian Gulf War: A comparative study. Psychological Medicine 32(8):1357-1370.

Davis LE, Eisen SA, Murphy FM, Alpern R, Parks BJ, Blanchard M, Reda DJ, King MK, Mithen FA, Kang HK. 2004. Clinical and laboratory assessment of distal peripheral nerves in Gulf War veterans and spouses. Neurology 63(6):1070-1077.

DeBakey S, Paxton M, Weaver R, Lange J, Cowan D, Kang H, Hooper T, Gackstetter G. 2002. #52 Risk of motor vehicle fatality associated with prior morbidity among gulf war era veterans. Annals of Epidemiology 12(7):509.

Defence Analytical Services Agency. 2005. 1990/1991 Gulf Conflict—UK Gulf Veterans Mortality Data: Causes of Death. National Statistics.

Department of Health and Human Services. 1999. Mental Health: A Report of the Surgeon General. [Online]. Available: http://www.surgeongeneral.gov/library/mentalhealth/home.html [accessed April 10, 2006].

Dlugosz LJ, Hocter WJ, Kaiser KS, Knoke JD, Heller JM, Hamid NA, Reed RJ, Kendler KS, Gray GC. 1999. Risk factors for mental disorder hospitalization after the Persian Gulf War: U.S. Armed Forces, June 1, 1991-September 30, 1993. Journal of Clinical Epidemiology 52(12):1267-1278.

Dobie DJ, Kivlahan DR, Maynard C, Bush KR, McFall M, Epler AJ, Bradley KA. 2002. Screening for post-traumatic stress disorder in female Veteran's Affairs patients: Validation of the PTSD checklist. General Hospital Psychiatry 24(6):367-374.

Doebbeling BN, Clarke WR, Watson D, Torner JC, Woolson RF, Voelker MD, Barrett DH, Schwartz DA. 2000. Is there a Persian Gulf War syndrome? Evidence from a large population-based survey of veterans and nondeployed controls. American Journal of Medicine 108(9):695-704.

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

Doyle P, Maconochie N, Davies G, Maconochie I, Pelerin M, Prior S, Lewis S. 2004. Miscarriage, stillbirth and congenital malformation in the offspring of UK veterans of the first Gulf war. International Journal of Epidemiology 33(1):74-86.

Draxler RR, McQueen JT, Stunder BJB. 1994. An evaluation of air pollutant exposures due to the 1991 Kuwait oil fires using a Lagrangian model. Atmospheric Environment 28(13):2197-2210.

Eisen SA, Kang HK, Murphy FM, Blanchard MS, Reda DJ, Henderson WG, Toomey R, Jackson LW, Alpern R, Parks BJ, Klimas N, Hall C, Pak HS, Hunter J, Karlinsky J, Battistone MJ, Lyons MJ. 2005. Gulf War veterans' health: Medical evaluation of a US cohort. Annals of Internal Medicine 142(11):881-890.

Epstein KR. 1995. The chronically fatigued patient. Medical Clinics of North America 79(2):315-327.

Everitt B, Ismail K, David AS, Wessely S. 2002. Searching for a Gulf War syndrome using cluster analysis.Psychological Medicine 32(8):1371-1378.

Fiedler N, Kipen H, Natelson B, Ottenweller J. 1996. Chemical sensitivities and the Gulf War: Department of Veterans Affairs Research Center in basic and clinical science studies of environmental hazards. Regulatory Toxicology and Pharmacology 24(1 Pt 2):S129-S138.

Forbes AB, McKenzie DP, Mackinnon AJ, Kelsall HL, McFarlane AC, Ikin JF, Glass DC, Sim MR. 2004. The health of Australian veterans of the 1991 Gulf War: Factor analysis of self-reported symptoms. Occupational and Environmental Medicine 61(12):1014-1020.

Forbes D, Creamer M, Biddle D. 2001. The validity of the PTSD checklist as a measure of symptomatic change in combat-related PTSD. Behaviour Research and Therapy 39(8):977-986.

Franse LV, Valk GD, Dekker JH, Heine RJ, van Eijk JT. 2000. 'Numbness of the feet' is a poor indicator for polyneuropathy in Type 2 diabetic patients. Diabetic Medicine 17(2):105-110.

Fukuda K, Straus SE, Hickie I, Sharpe MC, Dobbins JG, Komaroff A. 1994. The chronic fatigue syndrome: A comprehensive approach to its definition and study. International Chronic Fatigue Syndrome Study Group. Annals of Internal Medicine 121(12):953-959.

Fukuda K, Nisenbaum R, Stewart G, Thompson WW, Robin L, Washko RM, Noah DL, Barrett DH, Randall B, Herwaldt BL, Mawle AC, Reeves WC. 1998. Chronic multisymptom illness affecting Air Force veterans of the Gulf War. Journal of the American Medical Association 280(11):981-988.

Gackstetter G, DeBakey S, Cowan D, Paxton M, Weaver R, Lange J, Kang H, Bullman T, Lincoln A, Hooper T. 2002. Fatal motor vehicle crashes among veterans of the gulf war era: A nested case-control study. Annals of Epidemiology 12(7):509.

Gillespie N, Kirk KM, Heath AC, Martin NG, Hickie I. 1999. Somatic distress as a distinct psychological dimension. Social Psychiatry and Psychiatric Epidemiology 34(9):451-458.

Goldenberg DL. 1999. Fibromyalgia syndrome a decade later: What have we learned? Archives of Internal Medicine 159(8):777-785.

Goldstein G, Beers SR, Morrow LA, Shemansky WJ, Steinhauer SR. 1996. A preliminary neuropsychological study of Persian Gulf veterans. Journal of the International Neuropsychological Society 2(4):368-371.

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

Goshorn RK. 1998. Chronic fatigue syndrome: A review for clinicians. Seminars in Neurology 18(2):237-242.

Goss Gilroy Inc. 1998. Health Study of Canadian Forces Personnel Involved in the 1991 Conflict in the Persian Gulf. Ottawa, Canada: Goss Gilroy Inc. Department of National Defence.

Graveling RA, Pilkington A, George JP, Butler MP, Tannahill SN. 1999. A review of multiple chemical sensitivity. Occupational and Environmental Medicine 56(2):73-85.

Gray GC, Coate BD, Anderson CM, Kang HK, Berg SW, Wignall FS, Knoke JD, Barrett-Connor E. 1996. The postwar hospitalization experience of U.S. veterans of the Persian Gulf War. New England Journal of Medicine 335(20):1505-1513.

Gray GC, Kaiser KS, Hawksworth AW, Hall FW, Barrett-Connor E. 1999a. Increased postwar symptoms and psychological morbidity among U.S. Navy Gulf War veterans. American Journal of Tropical Medicine and Hygiene 60(5):758-766.

Gray GC, Smith TC, Knoke JD, Heller JM. 1999b. The postwar hospitalization experience of Gulf War Veterans possibly exposed to chemical munitions destruction at Khamisiyah, Iraq. American Journal of Epidemiology 150(5):532-540.

Gray GC, Smith TC, Kang HK, Knoke JD. 2000. Are Gulf War veterans suffering war-related illnesses? Federal and civilian hospitalizations examined, June 1991 to December 1994. American Journal of Epidemiology 151(1):63-71.

Gray GC, Reed RJ, Kaiser KS, Smith TC, Gastanaga VM. 2002. Self-reported symptoms and medical conditions among 11,868 Gulf War-era veterans: The Seabee Health Study. American Journal of Epidemiology 155(11):1033-1044.

Haley RW. 2003. Excess incidence of ALS in young Gulf War veterans. Neurology 61(6):750-756.

Haley RW, Hom J, Roland PS, Bryan WW, Van Ness PC, Bonte FJ, Devous MD Sr, Mathews D, Fleckenstein JL, Wians FH Jr, Wolfe GI, Kurt TL. 1997a. Evaluation of neurologic function in Gulf War veterans. A blinded case-control study. Journal of the American Medical Association 277(3):223-230.

Haley RW, Kurt TL, Hom J. 1997b. Is there a Gulf War Syndrome? Searching for syndromes by factor analysis of symptoms. Journal of the American Medical Association 277(3):215-222.

Haley RW, Billecke S, La Du BN. 1999. Association of low PON1 type Q (type A) arylesterase activity with neurologic symptom complexes in Gulf War veterans. Toxicology and Applied Pharmacology 157(3):227-233.

Haley RW, Luk GD, Petty F. 2001. Use of structural equation modeling to test the construct validity of a case definition of Gulf War syndrome: Invariance over developmental and validation samples, service branches and publicity. Psychiatry Research 102(2):175-200.

Haley RW, Fleckenstein JL, Marshall WW, McDonald GG, Kramer GL, Petty F. 2000a. Effect of basal ganglia injury on central dopamine activity in Gulf War syndrome: Correlation of proton magnetic resonance spectroscopy and plasma homovanillic acid levels. Archives of Neurology 57(9):1280-1285.

Haley RW, Marshall WW, McDonald GG, Daugherty MA, Petty F, Fleckenstein JL. 2000b. Brain abnormalities in Gulf War syndrome: Evaluation with 1H MR spectroscopy. Radiology 215(3):807-817.

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

Haley RW, Vongpatanasin W, Wolfe GI, Bryan WW, Armitage R, Hoffmann RF, Petty F, Callahan TS, Charuvastra E, Shell WE, Marshall WW, Victor RG. 2004. Blunted circadian variation in autonomic regulation of sinus node function in veterans with Gulf War syndrome. American Journal of Medicine 117(7):469-478.

Hallman WK, Kipen HM, Diefenbach M, Boyd K, Kang H, Leventhal H, Wartenberg D. 2003. Symptom patterns among Gulf War registry veterans. American Journal of Public Health 93(4):624-630.

Hardt J, Buchwald D, Wilks D, Sharpe M, Nix WA, Egle UT. 2001. Health-related quality of life in patients with chronic fatigue syndrome: An international study. Journal of Psychosomatic Research 51(2):431-434.

Higgins EM, Ismail K, Kant K, Harman K, Mellerio J, Du Vivier AW, Wessely S. 2002. Skin disease in Gulf war veterans. QJM 95(10):671-676.

Holmes DT, Tariot PN, Cox C. 1998. Preliminary evidence of psychological distress among reservists in the Persian Gulf War. Journal of Nervous and Mental Disease 186(3):166-173.

Holmes GP, Kaplan JE, Gantz NM, Komaroff AL, Schonberger LB, Straus SE, Jones JF, Dubois RE, Cunningham-Rundles C, Pahwa S, et al. 1988. Chronic fatigue syndrome: A working case definition. Annals of Internal Medicine 108(3):387-389.

Hom J, Haley RW, Kurt TL. 1997. Neuropsychological correlates of Gulf War syndrome. Archives of Clinical Neuropsychology 12(6):531-544.

Horner RD, Kamins KG, Feussner JR, Grambow SC, Hoff-Lindquist J, Harati Y, Mitsumoto H, Pascuzzi R, Spencer PS, Tim R, Howard D, Smith TC, Ryan MA, Coffman CJ, Kasarskis EJ. 2003. Occurrence of amyotrophic lateral sclerosis among Gulf War veterans. Neurology 61(6):742-749.

Hotopf M, Mackness MI, Nikolaou V, Collier DA, Curtis C, David A, Durrington P, Hull L, Ismail K, Peakman M, Unwin C, Wessely S, Mackness B. 2003. Paraoxonase in Persian Gulf War veterans. Journal of Occupational and Environmental Medicine 45(7):668-675.

Hotopf M, David A, Hull L, Nikalaou V, Unwin C, Wessely S. 2004. Risk factors for continued illness among Gulf War veterans: A cohort study. Psychological Medicine 34(4):747-754.

Hull L, Farrin L, Unwin C, Everitt B, Wykes T, David AS. 2003. Anger, psychopathology and cognitive inhibition: A study of UK servicemen. Personality and Individual Differences 35(5):1211.

Ikin JF, Sim MR, Creamer MC, Forbes AB, McKenzie DP, Kelsall HL, Glass DC, McFarlane AC, Abramson MJ, Ittak P, Dwyer T, Blizzard L, Delaney KR, Horsley KWA, Harrex WK, Schwarz H. 2004. War-related psychological stressors and risk of psychological disorders in Australian veterans of the 1991 Gulf War. British Journal of Psychiatry 185:116-126.

Iowa Persian Gulf Study Group. 1997. Self-reported illness and health status among Gulf War veterans: A population-based study. Journal of the American Medical Association 277(3):238-245.

Ishoy T, Suadicani P, Guldager B, Appleyard M, Gyntelberg F. 1999a. Risk factors for gastrointestinal symptoms. The Danish Gulf War Study. Danish Medical Bulletin 46(5):420-423.

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

Ishoy T, Suadicani P, Guldager B, Appleyard M, Hein HO, Gyntelberg F. 1999b. State of health after deployment in the Persian Gulf. The Danish Gulf War Study. Danish Medical Bulletin 46(5):416-419.

Ishoy T, Andersson AM, Suadicani P, Guldager B, Appleyard M, Gyntelberg F, Skakkebaek NE, Danish Gulf War Study. 2001a. Major reproductive health characteristics in male Gulf War Veterans. The Danish Gulf War Study. Danish Medical Bulletin 48(1):29-32.

Ishoy T, Suadicani P, Andersson A-M, Guldager B, Appleyard M, Skakkebaek N, Gyntelberg F. 2001b. Prevalence of male sexual problems in the Danish Gulf War Study. Scandinavian Journal of Sexology 4(1):43-55.

Ismail K, Blatchley N, Hotopf M, Hull L, Palmer I, Unwin C, David A, Wessely S. 2000. Occupational risk factors for ill health in Gulf veterans of the United Kingdom. Journal of Epidemiology and Community Health 54(11):834-838.

Ismail K, Everitt B, Blatchley N, Hull L, Unwin C, David A, Wessely S. 1999. Is there a Gulf War syndrome? Lancet 353(9148):179-182.

Jason LA, Taylor RR, Kennedy CL. 2000. Chronic fatigue syndrome, fibromyalgia, and multiple chemical sensitivities in a community-based sample of persons with chronic fatigue syndrome-like symptoms. Psychosomatic Medicine 62(5):655-663.

Joseph TK, Foster L, Pasquina PF. 2004. Decreased prevalence of peripheral nerve pathology by electrodiagnostic testing in Gulf War veterans. Military Medicine 169(11):868-871.

Kang HK, Bullman TA. 1996. Mortality among US veterans of the Persian Gulf War. New England Journal of Medicine 335(20):1498-1504.

Kang HK, Bullman TA. 2001. Mortality among US veterans of the Persian Gulf War: 7-year follow-up. American Journal of Epidemiology 154(5):399-405.

Kang HK, Mahan CM, Lee KY, Magee CA, Murphy FM. 2000. Illnesses among United States veterans of the Gulf War: A population-based survey of 30,000 veterans. Journal of Occupational and Environmental Medicine 42(5):491-501.

Kang H, Magee C, Mahan C, Lee K, Murphy F, Jackson L, Matanoski G. 2001. Pregnancy outcomes among US Gulf War veterans: A population-based survey of 30,000 veterans. Annals of Epidemiology 11(7):504-511.

Kang HK, Mahan CM, Lee KY, Murphy FM, Simmens SJ, Young HA, Levine PH. 2002. Evidence for a deployment-related Gulf War syndrome by factor analysis. Archives of Environmental Health 57(1):61-68.

Kang HK, Natelson BH, Mahan CM, Lee KY, Murphy FM. 2003. Post-traumatic stress disorder and chronic fatigue syndrome-like illness among Gulf War veterans: A population-based survey of 30,000 veterans. American Journal of Epidemiology 157(2):141-148.

Karlinsky JB, Blanchard M, Alpern R, Eisen SA, Kang H, Murphy FM, Reda DJ. 2004. Late prevalence of respiratory symptoms and pulmonary function abnormalities in Gulf War I Veterans. Archives of Internal Medicine 164(22):2488-2491.

Kelsall HL, Sim MR, Forbes AB, Glass DC, McKenzie DP, Ikin JF, Abramson MJ, Blizzard L, Ittak P. 2004a. Symptoms and medical conditions in Australian veterans of the 1991 Gulf War: Relation to immunisations and other Gulf War exposures. Occupational and Environmental Medicine 61(12):1006-1013.

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

Kelsall HL, Sim MR, Forbes AB, McKenzie DP, Glass DC, Ikin JF, Ittak P, Abramson MJ. 2004b. Respiratory health status of Australian veterans of the 1991 Gulf War and the effects of exposure to oil fire smoke and dust storms. Thorax 59(10):897-903.

Kessler RC, Sonnega A, Bromet E, Hughes M, Nelson CB. 1995. Posttraumatic stress disorder in the National Comorbidity Survey. Archives of General Psychiatry 52(12):1048-1060.

Kessler RC, Chiu WT, Demler O, Merikangas KR, Walters EE. 2005. Prevalence, severity, and comorbidity of 12-month DSM-IV disorders in the National Comorbidity Survey Replication. Archives of General Psychiatry 62(6):617-627.

Kipen HM, Fiedler N. 2002. Environmental factors in medically unexplained symptoms and related syndromes: The evidence and the challenge. Environmental Health Perspectives 110 (S4):597-599.

Knoke JD, Gray GC. 1998. Hospitalizations for unexplained illnesses among US veterans of the Persian Gulf War. Emerging Infectious Diseases 4(2):211-219.

Knoke JD, Gray GC, Garland FC. 1998. Testicular cancer and Persian Gulf War service. Epidemiology 9(6):648-653.

Knoke JD, Smith TC, Gray GC, Kaiser KS, Hawksworth AW. 2000. Factor analysis of self-reported symptoms: Does it identify a Gulf War syndrome? American Journal of Epidemiology 152(4):379-388.

Komaroff AL, Fagioli LR, Doolittle TH, Gandek B, Gleit MA, Guerriero RT, Kornish RJ 2nd, Ware NC, Ware JE Jr, Bates DW. 1996. Health status in patients with chronic fatigue syndrome and in general population and disease comparison groups. American Journal of Medicine 101(3):281-90.

Kreutzer R, Neutra RR, Lashuay N. 1999. Prevalence of people reporting sensitivities to chemicals in a population-based survey. American Journal of Epidemiology 150(1):1-12.

Kroenke K, Koslowe P, Roy M. 1998. Symptoms in 18,495 Persian Gulf War veterans. Latency of onset and lack of association with self-reported exposures. Journal of Occupational and Environmental Medicine 40(6):520-528.

Lange G, Tiersky LA, Scharer JB, Policastro T, Fiedler N, Morgan TE, Natelson BH. 2001. Cognitive functioning in Gulf War Illness. Journal of Clinical and Experimental Neuropsychology 23(2):240-249.

Lange JL, Schwartz DA, Doebbeling BN, Heller JM, Thorne PS. 2002. Exposures to the Kuwait oil fires and their association with asthma and bronchitis among gulf war veterans. Environmental Health Perspectives 110(11):1141-1146.

Levine PH, Young HA, Simmens SJ, Rentz D, Kofie VE, Mahan CM, Kang HK. 2005. Is testicular cancer related to Gulf War deployment? Evidence from a pilot population-based study of Gulf War era veterans and cancer registries. Military Medicine 170(2):149-153.

Lezak M, Loring D, Howieson D. 2004. Neuropsychological Assessment. New York: Oxford University Press.

Lindem K, Heeren T, White RF, Proctor SP, Krengel M, Vasterling J, Sutker PB, Wolfe J, Keane TM. 2003a. Neuropsychological performance in Gulf War era veterans: Traumatic stress symptomatology and exposure to chemical-biological warfare agents. Journal of Psychopathology and Behavioral Assessment 25(2):105-119.

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

Lindem K, Proctor SP, Heeren T, Krengel M, Vasterling J, Sutker PB, Wolfe J, Keane TM, White RF. 2003b. Neuropsychological performance in Gulf War era veterans: Neuropsychological symptom reporting. Journal of Psychopathology and Behavioral Assessment 25(2):121-127.

Lindem K, White RF, Heeren T, Proctor SP, Krengel M, Vasterling J, Wolfe J, Sutker PB, Kirkley S, Keane TM. 2003c. Neuropsychological performance in Gulf War era veterans: Motivational factors and effort. Journal of Psychopathology and Behavioral Assessment 25 (2):129-138.

Lucchini R, Albini E, Benedetti L, Alessio L. 2005. Neurobehavioral science in hazard identification and risk assessment of neurotoxic agents—what are the requirements for further development? International Archives of Occupational and Environmental Health 78(6):427-437.

Macfarlane GJ, Thomas E, Cherry N. 2000. Mortality among UK Gulf War veterans. Lancet 356(9223):17-21.

Macfarlane GJ, Biggs AM, Maconochie N, Hotopf M, Doyle P, Lunt M. 2003. Incidence of cancer among UK Gulf war veterans: Cohort study. British Medical Journal 327 (7428):1373-1375.

Mackness B, Mackness MI, Arrol S, Turkie W, Durrington PN. 1997. Effect of the molecular polymorphisms of human paraoxonase (PON1) on the rate of hydrolysis of paraoxon. British Journal of Pharmacology 122(2):265-268.

Maconochie N, Doyle P, Carson C. 2004. Infertility among male UK veterans of the 1990-1 Gulf war: Reproductive cohort study. British Medical Journal 329(7459):196-201.

Magruder KM, Frueh BC, Knapp RG, Davis L, Hamner MB, Martin RH, Gold PB, Arana GW. 2005. Prevalence of posttraumatic stress disorder in Veterans Affairs primary care clinics. General Hospital Psychiatry 27(3):169-179.

Matarazzo JD. 1972. Wechsler's Measurement and Appraisal of Adult Intelligence. 5th ed. Baltimore, Md: Williams and Wilkins

McCauley LA, Joos SK, Lasarev MR, Storzbach D, Bourdette DN. 1999. Gulf War unexplained illnesses: Persistence and unexplained nature of self-reported symptoms. Environmental Research 81(3):215-223.

McCauley LA, Lasarev M, Sticker D, Rischitelli DG, Spencer PS. 2002. Illness experience of Gulf War veterans possibly exposed to chemical warfare agents. American Journal of Preventive Medicine 23(3):200-206.

McGuire V, Longstreth WT Jr, Koepsell TD, van Belle G. 1996a. Incidence of amyotrophic lateral sclerosis in three counties in western Washington state. Neurology 47(2):571-573.

McKenzie DP, Ikin JF, McFarlane AC, Creamer M, Forbes AB, Kelsall HL, Glass DC, Ittak P, Sim MR. 2004. Psychological health of Australian veterans of the 1991 Gulf War: An assessment using the SF-12, GHQ-12 and PCL-S. Psychological Medicine 34(8):1419-1430.

McQueen JT, Draxler RR. 1994. Evaluation of model back trajectories of the Kuwait oil fires smoke plume using digital satellite data. Atmospheric Environment 28(13):2159-2174.

Merck Manuals Online Medical Library. Polyneuropathy: Peripheral Nerve Disorders. [Online]. Available: http://www.merck.com/mmhe/sec06/ch095/ch095h.html [accessed July 11, 2006].

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

Nelson LM, McGuire V, Longstreth WT Jr, Matkin C. 2000. Population-based case-control study of amyotrophic lateral sclerosis in western Washington State. I. Cigarette smoking and alcohol consumption. American Journal of Epidemiology 151(2):156-163.

Nicolson GL, Nasralla MY, Haier J, Pomfret J. 2002. High frequency of systemic mycoplasmal infections in Gulf War veterans and civilians with Amyotrophic Lateral Sclerosis (ALS). Journal of Clinical Neuroscience 9(5):525-529.

Nimnuan C, Rabe-Hesketh S, Wessely S, Hotopf M. 2001. How many functional somatic syndromes? Journal of Psychosomatic Research 51(4):549-557.

NINDS (National Institute of Neurological Disorders and Stroke). 2006. National Institute of Neurological Disorders and Stroke. [Online]. Available: http://www.ninds.nih.gov/ [accessed January 17, 2006].

Nisenbaum R, Reyes M, Mawle AC, Reeves WC. 1998. Factor analysis of unexplained severe fatigue and interrelated symptoms: Overlap with criteria for chronic fatigue syndrome. American Journal of Epidemiology 148(1):72-77.

Nisenbaum R, Ismail K, Wessely S, Unwin C, Hull L, Reeves WC. 2004. Dichotomous factor analysis of symptoms reported by UK and US veterans of the 1991 Gulf War. Population Health Metrics 2(1):8.

O'Toole BI, Marshall RP, Grayson DA, Schureck RJ, Dobson M, Ffrench M, Pulvertaft B, Meldrum L, Bolton J, Vennard J. 1996. The Australian Vietnam Veterans Health Study: III. psychological health of Australian Vietnam veterans and its relationship to combat. International Journal of Epidemiology 25(2):331-340.

Pearce JM. 2004. Myofascial pain, fibromyalgia or fibrositis? European Neurology 52(2):67-72.

Penman AD, Tarver RS, Currier MM. 1996. No evidence of increase in birth defects and health problems among children born to Persian Gulf War Veterans in Mississippi. Military Medicine 161(1):1-6.

Petruccelli BP, Goldenbaum M, Scott B, Lachiver R, Kanjarpane D, Elliott E, Francis M, McDiarmid MA, Deeter D. 1999. Health effects of the 1991 Kuwait oil fires: A survey of US army troops. Journal of Occupational and Environmental Medicine 41(6):433-439.

Pizarro J, Silver RC, Prause J. 2006. Physical and mental health costs of traumatic war experiences among Civil War veterans. Archives of General Psychiatry 63(2):193-200.

Proctor SP, Heeren T, White RF, Wolfe J, Borgos MS, Davis JD, Pepper L, Clapp R, Sutker PB, Vasterling JJ, Ozonoff D. 1998. Health status of Persian Gulf War veterans: Self-reported symptoms, environmental exposures and the effect of stress. International Journal of Epidemiology 27(6):1000-1010.

Proctor SP, Heaton KJ, White RF, Wolfe J. 2001. Chemical sensitivity and chronic fatigue in Gulf War veterans: A brief report. Journal of Occupational and Environmental Medicine 43(3):259-264.

Proctor SP, White RF, Heeren T, et al. 2003. Neuropsychological Functioning in Danish Gulf War Veterans. Journal of Psychopathology and Behavioral Assessment 25(2):85-93.

Reid S, Hotopf M, Hull L, Ismail K, Unwin C, Wessely S. 2001. Multiple chemical sensitivity and chronic fatigue syndrome in British Gulf War veterans. American Journal of Epidemiology 153(6):604-609.

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

Ries, LAG, Eisner, MP, Kosary, CL, Hankey, BF, Miller, BA, Clegg, L, Mariotto, A, Feuer, EJ, and Edwards, BK, eds. 2005. SEER Cancer Statistics Review, 1975-2002. [Online]. Available: http://seer.cancer.gov/csr/1975_2002/ [accessed 2006].

Rivera-Zayas J, Arroyo M, Mejias E. 2001. Evaluation of Persian Gulf veterans with symptoms of peripheral neuropathy. Military Medicine 166(5):449-451.

Roland PS, Haley RW, Yellin W, Owens K, Shoup AG. 2000. Vestibular dysfunction in Gulf War syndrome. Otolaryngology—Head and Neck Surgery 122(3):319-329.

Rose MR, Sharief MK, Priddin J, Nikolaou V, Hull L, Unwin C, Ajmal-Ali R, Sherwood RA, Spellman A, David A, Wessely S. 2004. Evaluation of neuromuscular symptoms in UK Gulf War veterans: A controlled study. Neurology 63(9):1681-1687.

Rostker, B. 2000. US Demolition Operations at Khamisiyah. [Online]. Available: http://www.gulflink.osd.mil/khamisiyah_ii/ [accessed August 6, 2004].

Rowland LP. 2000. Hereditary and Acquired Motor Neuron Diseases. In: Rowland LP, Editor. Merritt's Neurology. 10th ed. Philadelphia, PA: Lippincott Williams and Wilkins. Pp. 708-714.

Roy-Byrne P, Arguelles L, Vitek ME, Goldberg J, Keane TM, True WR, Pitman RK. 2004. Persistence and change of PTSD symptomatology—a longitudinal co-twin control analysis of the Vietnam Era Twin Registry. Social Psychiatry and Psychiatric Epidemiology 39(9):681-685.

Shapiro SE, Lasarev MR, McCauley L. 2002. Factor analysis of Gulf War illness: What does it add to our understanding of possible health effects of deployment? American Journal of Epidemiology 156(6):578-585.

Sharief MK, Priddin J, Delamont RS, Unwin C, Rose MR, David A, Wessely S. 2002. Neurophysiologic analysis of neuromuscular symptoms in UK Gulf War veterans: A controlled study. Neurology 59(10):1518-1525.

Shaw E, Hermansen L, Pugh W, White M. 1991. Disease and Non-Battle Injuries Among Navy and Marine Corps Personnel During Operation Desert Shield/Desert Storm. US: US Naval Health Research Center.

Siddique N, Sufit R, Siddique T. 1999. Degenerative Motor, Sensory, and Autonomic Disorders. In: Goetz CG, Pappert EJ, Editors. Textbook of Clinical Neurology. 1st ed. Philadelphia, PA: W.B. Saunders Company. Pp. 695-717.

Sillanpaa MC, Agar LM, Milner IB, Podany EC, Axelrod BN, Brown GG. 1997. Gulf War veterans: A neuropsychological examination. Journal of Clinical and Experimental Neuropsychology 19(2):211-219.

Simmons R, Maconochie N, Doyle P. 2004. Self-reported ill health in male UK Gulf War veterans: A retrospective cohort study. BMC Public Health 4(1):27.

Simon GE, Daniell W, Stockbridge H, Claypoole K, Rosenstock L. 1993. Immunologic, psychological, and neuropsychological factors in multiple chemical sensitivity: A controlled study. Annals of Internal Medicine 119(2):97-103.

Smith TC, Gray GC, Knoke JD. 2000. Is systemic lupus erythematosus, amyotrophic lateral sclerosis, or fibromyalgia associated with Persian Gulf War service? An examination of Department of Defense hospitalization data. American Journal of Epidemiology 151(11):1053-1059.

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

Smith TC, Heller JM, Hooper TI, Gackstetter GD, Gray GC. 2002. Are Gulf War veterans experiencing illness due to exposure to smoke from Kuwaiti oil well fires? Examination of Department of Defense hospitalization data. American Journal of Epidemiology 155(10):908-917.

Smith TC, Gray GC, Weir JC, Heller JM, Ryan MA. 2003. Gulf War veterans and Iraqi nerve agents at Khamisiyah: Postwar hospitalization data revisited. American Journal of Epidemiology 158(5):457-467.

Sostek MB, Jackson S, Linevsky JK, Schimmel EM, Fincke BG. 1996. High prevalence of chronic gastrointestinal symptoms in a National Guard Unit of Persian Gulf veterans. American Journal of Gastroenterology 91(12):2494-2497.

Steele L. 2000. Prevalence and patterns of Gulf War illness in Kansas veterans: Association of symptoms with characteristics of person, place, and time of military service. American Journal of Epidemiology 152(10):992-1002.

Sterne JA, Davey Smith G. 2001. Sifting the evidence-what's wrong with significance tests? British Medical Journal 322(7280):226-231.

Storzbach D, Campbell KA, Binder LM, McCauley L, Anger WK, Rohlman DS, Kovera CA. 2000. Psychological differences between veterans with and without Gulf War unexplained symptoms. Psychosomatic Medicine 62(5):726-735.

Storzbach D, Rohlman DS, Anger WK, Binder LM, Campbell KA. 2001. Neurobehavioral deficits in Persian Gulf veterans: Additional evidence from a population-based study. Environmental Research 85(1):1-13.

Straus SE. 1991. History of chronic fatigue syndrome. Reviews of Infectious Diseases 13(S1):S2-S7.

Stretch RH, Bliese PD, Marlowe DH, Wright KM, Knudson KH, Hoover CH. 1996a. Psychological health of Gulf War-era military personnel. Military Medicine 161(5):257-261.

Stretch RH, Marlowe DH, Wright KM, Bliese PD, Knudson KH, Hoover CH. 1996b. Post-traumatic stress disorder symptoms among Gulf War veterans. Military Medicine 161(7):407-410.

Sutker PB, Davis JM, Uddo M, Ditta SR. 1995. War zone stress, personal resources, and PTSD in Persian Gulf War returnees. Journal of Abnormal Psychology 104(3):444-452.

Tasman A, Lieberman J, Kay J. 2003. Psychiatry. 2nd ed. London, UK: Wiley

Thomas TL, Kang HK, Dalager NA. 1991. Mortality among women Vietnam veterans, 1973-1987. American Journal of Epidemiology 134(9):973-980.

Tibshirani R WGHT. 2001. Estimating the number of clusters in a data set via the gap statistic. Journal of the Royal Statistical Society: Series B (Statistical Methodology) 63(2):411-423.

Unwin C, Blatchley N, Coker W, Ferry S, Hotopf M, Hull L, Ismail K, Palmer I, David A, Wessely S. 1999. Health of UK servicemen who served in Persian Gulf War. Lancet 353(9148):169-178.

Valenti M, Pontieri FE, Conti F, Altobelli E, Manzoni T, Frati L. 2005. Amyotrophic lateral sclerosis and sports: A case-control study. European Journal of Neurology 12(3):223-225.

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

Vasterling JJ, Brailey K, Tomlin H, Rice J, Sutker PB. 2003. Olfactory functioning in Gulf War-era veterans: Relationships to war-zone duty, self-reported hazards exposures, and psychological distress. Journal of the International Neuropsychological Society 9(3):407-418.

Watanabe KK, Kang HK. 1995. Military service in Vietnam and the risk of death from trauma and selected cancers. Annals of Epidemiology 5(5):407-412.

Weathers F, Ford J. 1996. Psychometric properties of the PTSD Checklist (PCL-C, PCL-S; PCL-M; PCL-PR). In: Stamm BH, Editor. Measurement of Stress Trauma and Adaptation. Lutherville, Maryland: Sidron Press.

Werler MM, Sheehan JE, Mitchell AA. 2005. Gulf war veterans and hemifacial microsomia. Birth Defects Research. Part A, Clinical and Molecular Teratology 73(1):50-52.

Wessely S. 1998. The epidemiology of chronic fatigue syndrome. Epidemiologia e Psichiatria Sociale 7(1):10-24.

Wessely S. 2005. Risk, psychiatry and the military. British Journal of Psychiatry 186:459-466.

White RF, Proctor SP, Heeren T, Wolfe J, Krengel M, Vasterling J, Lindem K, Heaton KJ, Sutker P, Ozonoff DM. 2001. Neuropsychological function in Gulf War veterans: Relationships to self-reported toxicant exposures. American Journal of Industrial Medicine 40(1):42-54.

Wikipedia. Polyneuropathy. [Online]. Available: http://en.wikipedia.org/wiki/Polyneuropathy; [accessed July 11, 2006].

Winkenwerder W. 2002. US Demolition Operations at Khamisiyah. Washington, DC: Department of Defense.

Wolfe F, Smythe HA, Yunus MB, Bennett RM, Bombardier C, Goldenberg DL, Tugwell P, Campbell SM, Abeles M, Clark P, et al. 1990. The American College of Rheumatology 1990 Criteria for the Classification of Fibromyalgia. Report of the Multicenter Criteria Committee. Arthritis and Rheumatism 33(2):160-172.

Wolfe F, Ross K, Anderson J, Russell IJ, Hebert L. 1995. The prevalence and characteristics of fibromyalgia in the general population. Arthritis and Rheumatism 38(1):19-28.

Wolfe F, Anderson J, Harkness D, Bennett RM, Caro XJ, Goldenberg DL, Russell IJ, Yunus MB. 1997. Health status and disease severity in fibromyalgia: Results of a six-center longitudinal study. Arthritis and Rheumatism 40(9):1571-1579.

Wolfe J, Erickson DJ, Sharkansky EJ, King DW, King LA. 1999a. Course and predictors of posttraumatic stress disorder among Gulf War veterans: A prospective analysis. Journal of Consulting and Clinical Psychology 67(4):520-528.

Wolfe J, Proctor SP, Erickson DJ, Heeren T, Friedman MJ, Huang MT, Sutker PB, Vasterling JJ, White RF. 1999b. Relationship of psychiatric status to Gulf War veterans' health problems. Psychosomatic Medicine 61(4):532-540.

Writer J, DeFraites R, Brundage J. 1996. Comparative Mortality among US Military Personnel in the Persian Gulf and Worldwide during Operations Desert Shield and Desert Storm. Journal of the American Medical Association 275(2):118-121.

Zwerling C, Torner JC, Clarke WR, Voelker MD, Doebbeling BN, Barrett DH, Merchant JA, Woolson RF, Schwartz DA. 2000. Self-reported postwar injuries among Gulf War veterans. Public Health Reports 115(4):346-349.

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×

This page intentionally left blank.

Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 115
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 116
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 117
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 118
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 119
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 120
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 121
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 122
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 123
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 124
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 125
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 126
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 127
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 128
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 129
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 130
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 131
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 132
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 133
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 134
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 135
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 136
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 137
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 138
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 139
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 140
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 141
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 142
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 143
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 144
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 145
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 146
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 147
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 148
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 149
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 150
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 151
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 152
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 153
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 154
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 155
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 156
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 157
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 158
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 159
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 160
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 161
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 162
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 163
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 164
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 165
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 166
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 167
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 168
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 169
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 170
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 171
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 172
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 173
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 174
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 175
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 176
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 177
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 178
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 179
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 180
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 181
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 182
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 183
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 184
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 185
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 186
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 187
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 188
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×
Page 189
Suggested Citation:"5 Health Outcomes." Institute of Medicine. 2006. Gulf War and Health: Volume 4: Health Effects of Serving in the Gulf War. Washington, DC: The National Academies Press. doi: 10.17226/11729.
×