Here we address a general issue raised by questions about the "Persian Gulf Syndrome" and toxic exposure to troops engaged in the Persian Gulf War: where and how does a new category of illness or disease become established, and how can that process be improved? Persian Gulf War veterans have reported health problems that many suggest are related to service in that war during the period 1990-1991. A variety of complaints have been recorded by health care providers throughout the country, but, for the most part, these have not resulted in a clear diagnosis or explanation except for a very small number of complaints which have led to clear diagnoses such as leishmaniasis or malaria. Several distinguished review panels have examined the evidence that service in the Gulf is related to alleged chronic health problems (1-6). While these reviews have caused participants to be impressed and concerned about the veterans' complaints, no review panel has arrived at a definitive understanding of the veterans' illnesses. The review panels generally agree that there is unlikely to be a single explanation for the veterans' complaints. Consequently, health scientists supported by both private and public funds have been engaged in laboratory, clinical, and epidemiologic studies to find the cause or causes of the illnesses.

Stress related to war, historically associated with health problems (7), as well as particular features of the Gulf War (major involvement of reserve troops, first major conflict with a volunteer force, substantial number of women soldiers) have been regularly cited as important contributors to complaints and disability among Gulf War veterans. An unknown number of complaints are simple those unexplained symptoms and illnesses experienced in any large population. (Almost 700,000 troops were present at some time during Operation Desert Shield/Storm.)

The absence of clear explanations for most of the illnesses has frustrated the veterans as well as health scientists. The Department of Defense's very late release of information about potential exposure of some of the troops to chemical or biologic weapons has tended to confirm the belief that the real causes of the illnesses are being covered up. Contributing to this view are the absence of adequate medical records about who received special immunizations that were not in general use (anthrax and botulinum), and about the use of pyridostigmine bromide pills to counter the anticipated effects of nerve gases (1). In addition, cutbacks in and limited use of field epidemiology staff appear to have resulted in inadequate collection and maintenance of objective information about the biologic or environmental risks experienced by service personnel in the Gulf region. Given this serious lack of information about risk factors, many believe that the causes of illnesses for most of the veterans will never be known.

Confusion With Stress

Despite these problems, some investigators and clinicians treating affected veterans have continued to seek explanations for those illnesses as well as effective means of treatment. The suggestion that stress plays an important role has been largely rejected by the veterans and some of their caregivers. Instead, suggestions have been made that the illnesses are the expression of one of several poorly understood syndromes, a result of exposure to agents with poorly understood health effects, or explained by unconfirmed hypotheses regarding altered metabolism. Some of the various explanations have included: syndromes such as chronic fatigue syndrome, fibromyalgia, and multiple chemical sensitivity, exposure to agents such as depleted uranium, smoke from oil-well fires, and dental amalgams, and hypotheses about altered metabolism described as oxidated phosphorylation disorder, brainstem dysregulation syndrome, and chemically induced porphyria (1).

It is unlikely that any of these hypotheses will provide a unifying explanation for the problems. Most will not stand the test of time. However, in this issue of the Journal, Haley(8) calls attention to the inadequacies in studies that have been cited in support of war-related stress as the major explanation of the illnesses. While epidemiologists may argue with the particular approach he takes to adjusting estimates of prevalence of post-traumatic stress syndrome, Haley's article calls attention to the fact that stress is at least as difficult to study as some of the other hypotheses suggested. He argues that investigators should not attribute the veterans' problems to stress simply because the evidence needed to study physical factors is not available.

The need for caution that Dr. Haley advises is well taken when one is trying to identify causal factors. The difficulty in studying stress, however, is no more reason to reject the role of stress in the veterans' illnesses than it is a reason to reject physical causes. Landrigan (9) summarized the reason to include stress in the study and treatment of veterans"

As a nation, we need to get beyond the fallacious idea that diseases of the mind either are not real or are shameful, and to better recognize that the mind and the body are inextricably linked (9. p.261)

The Core Issue

The many efforts to understand illnesses in Gulf War veterans call attention to a poorly recognized problem faced by all investigators who study any purported new condition or disease. Plaguing all such investigations is the fact that there is no readily accepted process by which medical science determines that a previously unrecognized disease should become "recognized."

Several schemes have been proposed for evaluating whether some specified exposure (broadly defined) causes some specified illness (table 1). For example, the 1964 report of the Surgeon General's advisory committee on smoking and health (10) included a list of criteria for causality. Hill (11) published a frequently cited list in 1965; both reports focused on inferences of cause and effect from observational data. Much earlier, Koch published "postulates" about cause-effect inferences in the etiology of infectious disease (12, pp.22-23). Evans attempted to expand Hill and Koch's criteria to all types of disease by elaborating ten "Criteria for Causation" (13).

All of these efforts, however, address the relation between some identified exposure and some recognized health outcome. The present task is different: it is to determine whether there is another kind of outcome, either new or not previously recognized, that should be added to our understanding about human disease. If the answer to this last inquiry is "Yes," the criteria of Koch, Hill, and/or Evans may then be appropriate in efforts to further understand this newly described outcome.

In considering criteria that should inform decisions about new diseases, it is important to consider the different objectives that might be served by determining that a previously unrecognized condition should be added to the disease lexicon. For physicians, the primary objective is the need to arrive at a consensus on diagnostic criteria that will ultimately serve as the basis for determining appropriate clinical approaches to evaluation, treatment, or prevention of the condition in persons who may be affected. In addition, clear diagnostic criteria are needed to determine who suffers from the condition and sometimes in what degree, so that natural history of the condition can be studied successfully. These studies can identify relevant risk factors and etiologic agents that cause the condition and provide a basis for both treatment and prevention. Some would call this new nosologic entity. Proper diagnosis and identification of the causal agent (s) for each individual may also provide the basis (in the United States) for determining who is to pay for the care of the individual.

For the individual suffering from a candidate for a new nosologic identification, correct identification of the condition is important so that proper care or treatment for that condition can be sought. Other benefits of proper identification include: avoidance of further diagnostic tests that may be futile and expensive and meeting the need many people have for "diagnostic closure." In this regard, Eisenberg (14) has emphasized the important conceptual difference between what he refers to as an illness and as a disease:

To state if flatly, patients suffer "illnesses"; physicians diagnose and treat "disease". Let me make clear the distinction I intend: Illnesses are experiences of disvalued changes in states of being and in social function; diseases in the scientific paradigm of modern medicine, are abnormalities in the structure and function of body organs and systems...Illness and disease, so defined, do not stand in a one-to-one relationship. When disease is extreme...its persuasiveness makes illness inevitable. However, disease may occur in the absence of illness...[By contrast,] similar degrees of organ pathology may generate quite different reports of distress, differences determined by culture, expectation and setting (14 p.11).

Disease categories, therefore, should be understood to be a social constructions reflecting consensus not only among health professionals but among the lay public. People seeking explanations for their experiences of poor health achieve consensus among themselves that a set of symptoms and/or signs signifies something different from other known categories common in their culture. This in turn often precipitates their contact with health care professionals and other healers to confirm or reject the normality or abnormality of the phenomenon.

Health professionals attempt to classify each condition recognized as a disease within a standard framework or taxonomy of professional construction such as the International Classification of Diseases (ICD) (15) or the Diagnostic and Statistical Manual of Mental Disorders (DSM) (16) . When the classification does not seem to capture important aspects of the condition (unusual symptoms, presence or absence of recognized causative agents, response to treatment, etc.), clinicians commonly seek to explain what the condition represents and its possible etiology. This generally requires the development of a biologically plausible explanation for the condition compatible with a more general understanding or normal versus abnormal. This process can lead to the proposal and acceptance of new disease categories that emerge ultimately as the product of both lay (patient) and professional constrictions.

The setting of diagnostic criteria for a condition is ultimately a social process, and as such is influenced by multiple social forces. Among these are: 1) belief systems prevalent in the culture/society about the phenomenon (i.e., the categories we have chosen to accept as legitimate) (17, 18); 2) power relationships (who has the power to declare a phenomenon a disease, or abnormal condition) (19); 3) consequences of the diagnosis (insurance coverage, social ostracization); 4) dominant frames of reference for explaining the diagnosis (e.g., clinician does or does not get paid; patient does or does not have access to effective therapy).

When considered together, there are risks and benefits for patients , for physicians, and for society when a new diagnostic category is invoked and acted on. These were well summarized in the case of mitral valve prolapse, when Quill et al. (21) examined the border between normal variants in the population and the developing interest in characterizing a variant as a disease. Here the risks and benefits they described have been supplemented by adding consideration of not acting on a new diagnostic category.

Just as there are risks and benefits associated with diagnosing a disease, there are also risks and benefits of not making a diagnosis. For patients, not being assigned a diagnosis may imply that their complaints are not valid or serious, or forestall treatment while an incorrect diagnosis may lead to treatment that is inappropriate or harmful. For clinicians, the risk of not making a diagnosis include losing the respect of one's patient (and potential loss of income), but benefits include avoiding premature labeling and consequent initiation of inappropriate therapy. Risks to the society of no diagnosis may, for example, stigmatize individuals as malingering or feigning illness, or allow an infectious condition to spread to others. Benefits of making no diagnosis may include cost savings associated with no treatment and availability of a larger segment of the population to contribute to the work of society (because they are not excused by a diagnosis a problem may provide a beneficial impetus for study to enlarge our understanding of what the problem is, and development of accounts to predict and explain its occurrence.

Diagnostic categories have evolved over the years, resulting at intervals of roughly ten years in a revision of formal codes for disease classification such as the International Classification of Diseases (ICD) or the Diagnostic and Statistical Manual of Mental Disorders (DSM). For the most part, changes in these codifications represent refinements that result from improved understanding of physio- or psycho pathology. Entirely new disease categories generally emerge when there is recognition of a new clustering of symptoms in a population, usually with replication of their presentation across a series of patients. Once this happens, a provisional nomenclature is applied to the phenomenon. Clinicians' confidence about differentiating the new phenomenon from other known phenomena increases most commonly when there are technological developments or an objective marker (e.g., an etiologic test or a biopsy finding) that has a high probability of differentiating those with, from those without, the condition. Although an objective indicator is desirable, it may not be necessary and may not even exist. Use of these indicators reflects clinicians' working models of causation or of pathologic processes. Finally, as a consensus grows among professionals, formal diagnostic criteria emerge and modification in a taxonomy such as the ICD or DSM ensues.

While the above describes the general progression followed as a new disease category becomes accepted, it does not indicate how this acceptance actually occurs.

In order to address the general concern that Persian Gulf veterans have some new disease or diseases, we need to develop a way to group events or symptoms into one or several classes that will eventually allow generalizations regarding the events and their various causes. This requires that ill persons be grouped into categories so that members of each category can be distinguished from those of every other. The categories, in turn, need to be distinguished from already existing categories.

The difficulty in accomplishing this goal is described by Susser (12):

The criteria of determinants and effects do not correspond with each other, first because determinants have many effects and second because effects have many determinants....A single experience can give rise to multiple and diverse manifestations;...A single manifestation, on the other hand, can arise as a consequence of multiple and diverse experiences....

(12, p. 45).

For example, there are many causes of inflammation, while cigarette smoking has many health effects.

MacMahon and Pugh (22) provide some guidance by noting that two different types of criteria, manifestational and causal, have been used to group ill persons:

1. Manifestational criteria: The ill persons are grouped according to their similarity with respect to symptoms, signs, changes in body fluids or tissues, physiologic function, behavior, prognosis, or some combination of these features. Examples of diseases defined by manifestational criteria are fracture of the femur, diabetes mellitus, mental retardation, the common cold, schizophrenia, and cervical cancer.

2. Causal criteria: Here the grouping depends on the similarity of individuals with respect to a specifies experience believed to be a cause of their illness. Examples of diseases defined by causal criteria are birth trauma, silicosis, syphilis and leading poisoning (22, p. 48).

MacMahon and Pugh describe several additional important features related to the task of defining a new disease. First, they call attention to the fact that "Polymorphous effects of newly isolated causal agents may be understood, indeed expected" (22, p. 49). In this regard, they refer to the many different illnesses associated with cigarette smoking. Equally, they note that a causal agent may not be associated with all those who are ill with the particular set of manifestations.

Finally they note that:

In the absence of knowledge of causal factors, manifestational criteria provide the only basis for...classification. Here the setting of the limits of disease entities appears to be a highly intuitive process, having as a governing principle the assumption that the greater the similarity of the manifestations of illness, the more properly the illness of the persons exhibiting the manifestations may be considered an entity (22, p. 51).

It is rare that a first effort at collecting a set of manifestations will prove to be the final characterization of a disease. Increased attention to and study of the condition may lead to more precise criteria for diagnosis as well as some redefinition of the condition. Indeed, it has been suggested that the patient described by Alzheimer (23) did not have what we now call Alzheimer's disease (16). Such diagnostic "drift" is not rare. Moreover, the more heterogeneous the presentations and potential causes, the more difficult it will be to define a new disease or syndrome. The knowledge that accrues after the "disease" is first defined, however, allows, amendments that contribute to better specification of the condition. Ultimately, the "disease" may be grouped with others or separated into more than one category as the newly defined condition comes under systematic study.

For example, after evidence had been gathered on two separately defined disease categories spinabifida and anencophaly, the natural history of the two disease in a variety of populations suggested they could be reasonably considered as one entity now known as "neural tube defect." Conversely, there was a time when two patients, one with typhoid and one with typhus infections but each presenting with a high fever and rash, would have been diagnosed as having the same disease. Isolation of causative agents and better specification of the disease course led to recognition that these were two quite distinct diseases requiring separate diagnostic categories. In similar situations today, further study of an initial disease definition generally leads to improved understanding of the disease and its determinants, which in turn leads to a change over time in the manifestations of the disease that are considered as a central to its definition (24). Sometimes the definition of disease may shift entirely from being based on clinical history and physical examination to the related physiologic or laboratory observations, as has occurred with hypertension and osteoporosis.

Proposed Steps In The Analysis

Recognizing that the process of defining a new disease is necessarily iterative, we propose five criteria to initiate this process and evaluate the evidence that a new disease, or a new category of diseases, should be recognized. No single step is indispensable; they are meant, rather, to guide analysis and to provide some structure to deal with a problem that may be difficult to resolve.

1. Difference from known conditions

The complex of signs, symptoms, and/or other findings, must be sufficiently different from recognized disease entities or syndromes to merit a new designation at some level of the diagnostic hierarchy. There may be a conceptual problem in determining that some condition is a separate entity versus a variant of something already recognized. There will always be boundary problems between illness and health. An answer of "Yes" requires a complete and accurate description of the new entity and of how it differs from entities already recognized.

An example of a currently evolving syndrome that may ultimately be accepted as a news disease is chronic fatigue syndrome. While variants of this syndrome have been traced back possibly as early as Hippocrates, the assignment of causal mechanisms over the course of the 19th and 20th centuries seems to reflect more the emerging understanding of more general disease processes than well-studied specific causal links (20). In the 19th century, a focus on the nervous system was associated with a condition neurasthenia, while the mid-20th century's attention to infection characterized a similar condition as chronic brucellosis. In the second half of this century, new emphasis on endocrinology led to the suggestion that hypoglycemia was the explanation for chronic fatigue. Most recently, interest in immuno-toxicity has led to syndromes labeled "total allergy syndrome" and "chronic candidiasis." On careful examination, each of these explanations has been dismissed, yet health practitioners continue to believe that there may be an explainable syndrome that is characterized most prominently by chronic fatigue. As noted earlier, efforts to understand and study chronic fatigue syndrome have resulted in different sets of criteria being proposed for defining the condition. Recently, investigators from Australian the United Kingdom, and the United States have proposed a revision to an older definition explicitly for use in carefully designed studies of chronic fatigue syndrome (25).

2. Appropriate limits of a definition

The description, parameters, and severity of a new condition must not be so broad as to exclude a common etiology, or so narrow as to suggest either bias (and possibly mutual redefinition of symptoms) or censoring (of possible atypical cases). For example, for years, health scientists have noted a common late afternoon wave of fatigue and have associated it with circadian patterns of reduced blood sugar in that period of an individual's day. During the 1960s and 1970s, low blood sugar became an increasingly popular explanation for persistent fatigue. Along with the emergence of the label hypoglycemia to explain protracted exhaustion, some caregivers began to argue for the need to take an elaborate dietary history and to introduce behavioral modification to correct the condition. Systematic population studies, however, have demonstrated that the notion includes numerous asymptomatic people with fasting or postprandial blood sugar levels in the range of 40 to 50 mg/dL (24). Subsequent to these studies, the term hypoglycemia has gradually returned to more limited use as an explanation of afternoon fatigue.

3. Replication by appropriate range of practitioners

The condition must have been observed and confirmed by a significant number of qualifies medical practitioners, drawn from a broad panel of subdisciplines and places of practice. There would be some suspicion, for example, if a new disease entity could be recognized by surgeons, but not by internists, or by doctors of osteopathy but not doctors of medicine, or by physicians practicing in a specific locality, but not by qualified peers brought in to evaluate the matter.

Enthusiasm for chronic brucellosis as an explanation for chronic fatigue syndrome accompanied Evans' original publication (26). Failure to replicate the findings, and the demonstration that antibodies to brucella were widespread and unrelated to fatigue in general were widespread and unrelated fatigue in general population studies, resulted in the ultimate rejection of chronic brucellosis as a distinct condition characterized predominantly by chronic fatigue.

In contrast, reports that stomach ulcers could be related to infection with Helicobactor pylori were initially met with disbelief by many mainstream physicians and research scientists (27, 28). Yet, widespread replication of findings associating infection with stomach ulcers and gastritis has changed these attitudes and treatment of this infection now plays an important role in therapy of stomach ulcers.

4. Common features and causes

The designation of a new disease category is not entirely independent of considerations of cause and effect, and adequate demonstration of a cause-effect linkage can be persuasive that the effect should be recognized. For example, do affected persons tend to have something in common, e.g. common exposure(s), common susceptibilities, or common demographic features? Does it seem that all categories of persons with such features are susceptible? An answer of "No" is not conclusive, because knowledge may be incomplete, or because relevant exposure may be so ubiquitous as to make the condition commonplace, but we would expect such exceptions to be uncommon.

5. Context in which condition occurs

When evaluating the evidence that a new disease or a new category of disease might deserve to be recognized, it is important to consider whether there are undue pressures (including social, financial, and political) that promote or discourage the acceptance of a new disease category. Such pressures may be generated, for example, by perceived governmental needs to reduce medical coverage, by desire for incline by physicians or hospitals for services delivered, by patient-oriented concerns with compensation, by concerns about political vicissitudes, by incomplete or biased new exposures, or by a sense of solidarity among members of some possibly affected group. It should be stressed that such pressures may not be overt, they may be unrecognized by those they affect, and they may at times even be recognized and seen as laudable by persons who are dedicated to a cause. These types of pressures must be prevented from serving as the primary determinant for whether a new disease category should be established.

It is difficult to differentiate when pressure is legitimate (the medical community is properly reluctant to recognized as "new" every condition that does not fit readily into existing understanding of disease etiology) from when pressure is undue (because it derived from illegitimate or misguided interests of affected parties). When there is disagreement between lay and professional groups and/or among professionals, the search for explanatory models propels both groups to act. The lay group considers alternative attribution for the cause of the disease, as well as for the cause of the professionals' disagreement of failure to name the experience. Development of criteria for human immunodeficiency (HIV-AIDS) exemplifies the processes in which lay people organized to prod the professionals to improve their understanding of a new phenomenon in which there had been little previous interest or attention. Past breeches of trust between patients and professionals (e.g., over the effects of the use of diethylstilbestrol (DES) and the Dalkon shield, and in regard to the effects of black lung and byssinosis) have created a sociopolitical context against which the public makes inferences about patient/professional disagreements.

The five criteria proposed for use when exploring the possibility that a new condition exists are neither exhaustive nor fail-safe. They are proposed in an effort to promote a more systematic approach to the consideration of suggestions that an new condition such as a Gulf War syndrome exists. Determination that a new disease exists is potentially important with respect to differential diagnosis, treatment decisions, and properly targeted prevention efforts, as well as the design of research programs. The emergence of any new disease process presents a serious and difficult challenge to medical science. Understandably, physicians are comforted by the known, and troubled by conditions that they cannot readily explain. All of use need help in finding ways to see out environment in a new light, and learn how that new light changes what we can subsequently observe.

In the end, we are doubtful that the symptoms linked with Persian Gulf syndrome can meet these or other reasonable criteria for the diagnosis of a new condition. However, it is possible that new research findings will shift the totality of evidence in the other direction. Much research on Gulf War syndrome is already in progress, and more will surely be undertaken as both our understanding of the matter and the needed research tools develop.

Acknowledgments

A portion of this commentary was prepared during the authors' work while serving on the Institute of Medicine's Committee to Review the Health Consequences of Service During the Persian Gulf War (2). Helpful comments were received from several other members of the committee, as well as from Dr. Myron E. Wegnam.

References

1. Hyams KC, Wignall ES, Roswell R. War syndromes and their evaluation: from the US Civil War to the Persian Gulf War. Ann Intern Med. 1996; 125; 398-405.

2. Institute of Medicine , Health consequences of service during the Persian Gulf War: recommendations for research and information systems. Washington, DC.: National Academy Press, 1996.

3. Defense Science Board. Final Report; Defense Science Board Task Force on Persian Gulf War Health Effects. Washington, DC.; Office of the Under Secretary of Defense for Acquisitions and Technology, 1994.

4. NIH Technology Assessment Workshop Panel. The Persian Gulf experience and health. JAMA 1994; 272: 391-6

5. Persian Gulf Veterans Coordinating Board. Summary of the issues impacting upon the health of Persian Gulf Veterans. Version 3.0. Washington, DC.: Support Office of the Persian Gulf Veterans Coordinating Board, 1994.

6. Persian Gulf Veterans Coordinating Board. Unexplained illnesses among Desert Storm Veterans. Arch Intern Med. 1995; 155: 262-8.

7. Presidential Advisory Committee on Gulf War Veterans' Illnesses. Final report. Washington, DC.: US GPO, December, 1996.

8. Haley RW. Is the Gulf War syndrome due to stress? The evidence reexamined. Am J Epidemiolo 1997; 146;695 703.

9. Landrigan PJ. Illnesses in Gulf War Veterans: causes and consequences. JAMA 1997; 277; 259-61.

10. Surgeon General, Advisory Committee of the US Public Health Service. Smoking and health. (PHS publication no. 1103). Washington, DC.: US GPO, 1964.

11. Hill AB. The environment and disease: association or causation? Proc. R Soc Med 1965; 58; 295-300.

12. Susser M. Causal thinking in the health sciences. New York: Oxford University Press, 1973.

13. Evans AS. Causation and disease: the Henle Koch postulates revisited. Yale J Biol Med 1976; 49; 175-95.

14. Eisenberg I., Disease and illness; distinctions between professional and popular ideas of sickness, Cult Med Psychiatry 1977; 1; 9-23.

15. World Health Organization. The international statistical classification of diseases and health related problems. 10th Revision. Geneva: World Health Organization, 1992.

16. American Psychiatric Association. Diagnostic and statistical manual of mental disorders (DSM IV), 4th ed. Washington, DC.: American Psychiatric Association, 1994.

17. Chrisman N, Kleinman A. Popular health care, social networks, and cultural meanings: the orientation of medical anthropology. In: Mechanic D, ed. Handbook of health, health care, and the health professions. New York: Free Press, 1983; 569-90.

18. Mechanic D. The concept of illness behavior. J Chronic Dis 1962; 15: 189-94.

19. Zola I. Medicine as an institution of social control. In: Ehrenreich J. ed. The cultural crisis of modern medicine. New York: Monthly Review Press, 1978: 80-100.

20. Straus SE. History of chronic fatigue syndrome. Rev Infect Dis 1991; 13 (Suppl 1): S2 7.

21. Quill TE. Lipkin M Jr, Greenland P. The medicalization al variants: the case of mitral valve prolapse. J Gen Intern Med 1988; 3; 267-76.

22. MacMahon B, Pugh TF. Epidemiology: principles and practices, Boston: Little, Brown & Co. 1970.

23. Alzheimer A. Uber cine eigenartigie Ekkrankung der Himrinde. Allgen Zeit Psychiat Psych Gerich Med 1907; 64: 146-8.

24. Yager J, Young RT. Non-hypoglycemia is an epidemic condition. N Engl J Med 1974; 291: 907-9/

25. Fukuda K, Straus SE. Hickie I, et al. The chronic fatigue syndrome: a comprehensive approach to its definition and study. International Chronic Fatigue Syndrome Study Group. Ann A Intern Med 1994; 37; 139-51.

27. Marshall BJ, Armstrong JA, McGechie DB, et al. Attempt to fulfill Koch's postulates for pyloric Campylobacer. Med J Aust 1985; 142: 436-9.

28. Marshall BJ, McGechie DB, Rogers PA, et al. Pyloric Campy-lobacter infection and gastroduodenal disease. Med J Aust 1985; 142: 439-44.

Dr. Haley Replies to Wegman et al's Invited Commentary of Gulf War Syndrome

Robert W. Haley

In their invited commentary on my article (1), Wegman et al. (2) acknowledge the inadequacies in the studies that have been cited in support of war-related stress as the explanation of persisting illnesses in Persian Gulf War veterans. However, they subtly attempt to leave open the door to the stress theory by reasoning, "The difficulty in studying stress...is no...reason to reject the role of stress in the veterans' illnesses...." (2, p. 705).

On the contrary, the problem with the stress theory of the Gulf War syndrome is not a difficulty in detecting post-traumatic stress disorder (PTSD). When PTSD is present, the diagnosis is readily made with a structured interview by a trained psychiatrist or psychologist in which the official criteria for PTSD are satisfied and other conditions with which it ca be confused (e.g., other psychiatric, neurologic, or neurotoxic disorders) are separated out. As I demonstrate in my review, the problem in regard to the investigations into Gulf War syndrome arose from an ill-advised over reliance on nonspecific diagnostic tools (namely, psychometric PTSD seales) in epidemiologic surveys and a failure to employ the definitive diagnostic interview technique until after public positions in favor of the stress etiology had become entrenched.

References

1. Haley RW, Is the Gulf War syndrome due to stress? The evidence reexamined. Am J Epidemiol 1997; 146; 695-703.

2. Wegman DH, Woods NF, Bailar JC. Invited commentary: How would we know a Gulf War syndrome if we saw one? Am J Epidemiol 1997; 146;704-11.