• Decreased lung function in the first 10 years after the war.

  • Hospitalization for genitourinary diseases.


Virtually all the reports in the Gulf War and Health series have called for improved studies of Gulf War and other veterans. This committee reiterates that need but notes that at almost 20 years after the war, it is difficult, if not impossible, to reconstruct the exposures to which the veterans were subjected in theater. It is similarly difficult after so many years to establish the predeployment physical and mental health status of these veterans for comparison purposes. Nevertheless,

  • To date, while many studies have been conducted on Gulf War veterans, their quality is varied and many of them have substantial methodological limitations. As a result uncertainty remains concerning the relationship between deployment to the Gulf War and health outcomes. These limitations include:

    • Lack of representativeness of the Gulf War population in some studies, affecting external validity such that what we have learned from the samples studied cannot be easily extrapolated to all Gulf War veterans.

    • Low participation rates and, indeed, differential participation rates in many studies, affecting internal validity due to selection bias. For example, the significantly higher response rate among deployed veterans compared with nondeployed control groups observed in many studies may result from greater participation of deployed troops because they were already experiencing health problems.

    • Studies that may have been too narrow in their assessment of health status (for example, self-reported outcomes such as hypertension, diabetes, or cardiovascular disease), or use of data collection instruments that might have been too insensitive (or invalid) to detect abnormalities in deployed veterans (for example, death certificates or hospital-discharge diagnoses).

    • There is a particular problem with self-reported exposures, especially when respondents are aware of media reports linking outcomes with putative exposures.

    • Timing of the investigation relative to the latent period for some health outcomes (for example, cancer, and some neurologic outcomes, such as MS, ALS, or Parkinson’s disease, for which there might be a long time between exposure and disease onset).

    • Use of cross-sectional studies that limit assessment of temporality, symptom duration and chronicity, latency of onset, severity, and prognosis.

  • Future studies of Gulf War veterans, and indeed any veteran population, need to be adequately designed in order to:

    • Provide sufficient statistical power (precision).

    • Ensure validity, including the avoidance of bias, such as recall and response bias, meaning that efforts should be made to encourage equal participation among deployed and nondeployed individuals with the goal of avoiding participation rates linked to general health and symptom presence, and to ensure comparable assessment of the severity and frequency of reported symptoms.

The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement