Paper
When Do Symptoms Become a Disease?
Published May 1, 2001 · R. Aronowitz
Annals of Internal Medicine
148
Citations
7
Influential Citations
Abstract
Symptoms to Disease: The Historical Context When does or should a collection of symptoms constitute a disease? The investigator or clinician who asks this question (the symptoms-to-disease question) usually wants to know whether there are rules or norms, currently or in the past, that say whether a particular collection of largely symptom-based criteria has enough specificity, utility, or plausibility to justify the appellation disease. I answer this question by first examining how the meanings and contexts of the words symptom and disease have changed over the past 200 years. I then look at examples of past and present symptom complexes that have been treated as specific entities in western medicine, and I draw some conclusions about our norms for naming and classifying them. My goal is not to judge whether these norms are good or bad but to make the norms explicit so that we understand what is at stake in the many controversies over the definition and legitimacy of contemporary symptom-based diagnoses. A partial answer to the symptoms-to-disease question can be found in the differences between older and modern entries for the word symptom (from the Greek, combining roots for to fall and together) in the Oxford English Dictionary (1). In the dictionary's citations from the Middle Ages to the late 19th century, symptoms were generally the bodily or mental phenomena that a person experienced and that arose from, accompanied, or (often) constituted a particular illness. An 1804 citation, for example, states that His skin was hot, and his pulse strong. These symptoms could be attributed to inflammation of the brain. Changes experienced by people or observed by physicians (such as pulse strong) could be called symptoms. By the latter half of the 19th century, symptom had taken on other meanings. An 1869 citation in the Oxford English Dictionary states that Diseases are distinguished from each other either by such alterations in the organs themselves, or their secretions, as can be ascertained by the senses of the observer (physical signs); or by changes in the functions of the parts affected (symptoms). Symptoms had become subjective, explicitly and consciously opposed to signs, the objective alterations noted by the physician. Symptoms by themselves were less likely to constitute disease because they did not arise directly from the alterations of organs but rather from the organs' less precise functions. Symptoms were defined in the context of the increasingly prominent scientific, medical, and social act of distinguishing specific diseases from each other. In this act, symptoms were to play an increasingly subordinate role to signs. This new understanding of ill health as a result of specific disease, increasingly defined more by signs than symptoms, was accompanied by a relentless push to define disease at ever deeper levels (from subjective experience to organ derangements and then to cellular derangements). As a result, many patients lost their symptom-based and clinically based diagnoses and became medical orphans. In the beginning of the 20th century, for example, a physicianon the basis of a history and physical examinationmight offer the diagnosis peptic ulcer disease to patients who later in the century would lose or never receive this diagnosis if the characteristic objective abnormality, the visualized gastric or duodenal ulcer, was not detected by an upper gastrointestinal series or a specialist's endoscope. But neither these patients nor their symptoms disappeared. A common medical response has been to create new, increasingly nonspecific symptom-based diagnoses for these patients, such as the negatively named nonulcer dyspepsia. Just as both lay and medical people have generally viewed the shift from symptom-based to mechanism- or anatomy-based diagnosis as progress, they have often stigmatized the resulting nonspecific diagnoses, the patients who have received them, and even the physicians who have cared for these patients. This stigmacombined with many other cultural developments, such as a more widespread consumer ethos, lay advocacy for specific diseases, and the power of special interests to influence specific health legislationhas led to intense pressure to find mechanisms for, redefine, or simply rename nonspecific diagnoses so that they seem more specific and imply putative underlying mechanisms. (One example is the call to replace the term chronic fatigue syndrome with the term postviral syndrome.) These developments, in turn, have contributed to a medical backlash against affected patients and their symptom-based, nonspecific diagnoses, resulting in a cycle of medical and political controversies over disease definition. Thus, the contemporary context for the symptoms- to-disease question is typically a controversy over the legitimacy of a residual, negatively defined disease entity that has little clinical or laboratory specificity. Knowing this context, we can guess that these entities will have a difficult time meeting any proposed normative criteria for a bona fide specific disease. Understanding the history of symptom-based diagnoses, we can make sense of what is at stake in these controversies. In the early 20th century, symptom-based diagnoses were not typically enshrouded in the current kinds of controversies. The medical meanings of such diagnoses were not very different from those in use today, but they were put to different clinical and social uses. To take one example, some U.S. physicians in the early decades of the 20th century debated whether angina pectoris should be defined as a freestanding, symptom-based, functional disordera characteristic pattern of chest pain symptoms not connected to one organic lesionor as a symptom of a specific, anatomically defined disease: coronary artery obstruction leading to myocardial ischemia. Although the debates were heated, pitting generalists against specialists and clinical skills against new technology, the underlying assumptions and the stakes differed from those in contemporary debates over symptom-based diagnoses, such as the chronic fatigue syndrome (2). The term functional disorder still had something of its earlier 19th-century meaningphysiologic, diffuse, and hiddenas opposed to its more modern absence of mechanism meaning. While functional diagnoses, such as the symptom-based angina pectoris, might have been held in lower regard than other diagnoses, they were not always strictly divided from real diagnoses. For the purpose of diagnosis, distinctions between mind and body, anatomic and functional, were less clearly demarcated. Many aspects of medical practice in earlier eras were supporting characters in this looser nosology. The reality of limited effective treatments and the weak resolution of diagnostic technology contributed to less hubris about the power of diagnosis and more accommodation of uncertainty by both physician and patient. The greater relative number of general practitioners, who frequently used clinical strategies such as watchful waiting and placebo prescription (practices often derided by specialists), was also part of a less confrontational, more flexible, more individualized style of diagnosis. Economic relations between patients and physicians were different too: There were fewer middle-class consumers of medical care, and payment was made or bartered directly between patient and physician rather than provided by third parties on the basis of specific diagnoses and treatments. In sum, less was riding on the specificity of diagnoses, symptom-based or otherwise. Although we cannot and would not want to set back the clock, this historical overview reminds us that contemporary interest in and controversies over symptom-based diagnoses reflect the increased expectations placed on specific diseases. Paths to Specificity: Historical Accident and False Appearances What paths have symptom complexes taken en route to acquiring the status of specific diseases? The history of many well-accepted rheumatologic diagnoses shows much about how symptom clusters have come to be accepted as having some specificity in clinical practice. These entities are frequently used by clinicians today, and their utility and specificity are arguably sustained by effective clinical strategies and by the later identification of specific, mediating immunologic mechanisms. Many of these entities, such as Still disease (1897), the Reiter syndrome (1916), and the Behet syndrome (1937), had their origins in the work of astute clinicians who carefully examined and reported on small numbers of patients seen in private practice (3-5). Although these syndromes were not immediately recognized as new entities and their names were not coined by the men after whom the syndromes were later named, these symptom clusters owe their existence as specific diseases to the norms and practices of an older era much different from our own. They have been grandfathered into our current nosology. I strongly doubt whether these diagnoses could be newly coined and promoted today. Experienced clinicians have less authority with which to get such work accepted and published and to garner the grant support necessary for protected time to do and write up research. Most prestigious medical journals do not publish small, uncontrolled case series. What is new and exciting today, with few exceptions, are diseases defined by new agents and (preferably molecular) mechanisms, not by purely clinical criteria. It is also important to recognize that many contemporary symptom-based diseases owe their existence to an initial and often erroneous suspicion of this more conventional, mechanism-based specificity. For example, what we now call the chronic fatigue syndrome began as chronic EpsteinBarr virus infection (2). The fact that social factors largely explain which symptom clusters get accepted as diseases has direct relevance to the care of patients. To patients whose symptoms ar
Highly CitedSymptoms have historically been defined more by signs than symptoms, and modern medicine has shifted from symptom-based diagnoses to sign-based diagnoses, with symptoms playing a subordinate role in disease definition.
Full text analysis coming soon...