Gulf War service is an uncertain trigger for ALS.

The suggestion that amyotrophic lateral sclerosis (ALS) was triggered by Gulf War service was first proposed when cases began to occur in young servicemen at an age when ALS is uncommon.1 This proposal gained widespread publicity following the formal recognition of ALS as a Gulf War–related illness by the Veterans Administration of the United States.2 This decision was made on humane and political grounds but has been taken as quasi-scientific validation for the claims. Two articles in this issue of Neurology, one by Haley and the other by Horner et al., allow readers to judge for themselves the scientific basis of claims of an excess risk of developing ALS in Gulf War veterans.3,4 These studies used both passive and active ascertainment of ALS cases using recruitment via military and health records and publicity campaigns among military personnel and their caregivers. Cases were then confirmed by case record review supplemented mostly by telephone and personal inquiries. Both studies almost certainly recruited from the same cohort of cases but Haley confined his study to 20 cases diagnosed under age 45 years, whereas Horner et al. confirmed 40 cases of all ages in their Gulf War–deployed group, which totaled nearly 700,000 service personnel. Any overlap of Gulf War– related ALS cases between these two articles is unknown. Haley calculated the expected incidence of ALS from national mortality statistics; Horner et al. determined the expected incidence from the incidence of ALS in a control group of co-serving military personnel not deployed to the Gulf War area. Both studies suggest a ballpark figure of twice the risk of developing ALS in Gulf War veterans and Haley suggests that the incidence figure has risen during each year of his study and therefore predicts an escalating incidence in the future. Although these researchers are to be congratulated for their thorough and painstaking work, there is still the concern that this degree of excess risk is not convincing, especially given the small number of ALS cases and a number of potential methodologic flaws. For both studies, under-reporting of ALS in the Gulf War population would have falsely lowered the excess risk figure. Confirmation of the diagnosis of ALS was not blinded in either study but nevertheless seems sound. Larger difficulties arise in trying to ascertain the true expected incidence of ALS because if this were falsely lowered the excess risk would be exaggerated. In using national mortality statistics for this purpose, Haley had to make certain assumptions (for example, that mortality statistics lag diagnosis by 2 years and that below age 45 years there is little rise in incidence of ALS with age), which may not be correct. Horner et al. have to contend with the possible problem of under-reporting of cases of ALS in their Gulf non-deployed control group. Both articles discuss some of these issues and no doubt there will follow a lively exchange of claims regarding the importance of these factors in influencing the results. There is a risk that this discussion will be polarized by vested interests and politics. Since these two studies were performed, mortality from ALS in Gulf War veterans has not risen as one would have expected and the age distribution of incidence of ALS has not shifted toward the younger age group as would be expected if an environmental agent triggered ALS prematurely.5 The issue of ALS as a Gulf War–related illness will probably remain under scrutiny until the passage of time and accumulation of further evidence allows a firmer scientific conclusion. How should the much larger number of service personnel, some with a variety of symptoms referred to as Gulf War syndrome, view this information? Haley mentions that 13 of his 20 patients with ALS had symptoms of Gulf War syndrome prior to their diagnosis of ALS but some of these symptoms seem unlikely to have been related to preclinical ALS. Identical symptoms have been described in service personnel who did not serve in the Gulf and presumably do not have an increased chance of getting ALS.6 Even if a twofold excess risk is correct, the baseline rarity of ALS means at worst an increase in