http://www.nytimes.com/2011/03/08/health/research/08fatigue.html By DAVID TULLER Published: March 4, 2011 When reports emerged 30 years ago that young gay men were suffering from rare forms of pneumonia and cancer, public health investigators scrambled to understand what appeared to be a deadly immune disorder: What were the symptoms? Who was most susceptible? What kinds of infections were markers of the disease? They were seeking the epidemiologist’s most essential tool — an accurate case definition, a set of criteria that simultaneously include people with the illness and exclude those without it. With AIDS, investigators soon recognized that injection-drug users, hemophiliacs and other demographic groups were also at risk, and the case definition evolved over time to incorporate lab evidence of immune dysfunction and other refinements based on scientific advances. “If you recognize something is happening, you need a case definition so you can count it,” said Andrew Moss, an emeritus professor of epidemiology at the University of California, San Francisco, and an early AIDS investigator. “You need to know whether the numbers are going up or down, or whether treatment and prevention work. And if you have a bad case definition, then it’s very difficult to figure out what’s going on.” Once a disease can be diagnosed reliably through lab tests, creating an accurate case definition becomes easier. But when an ailment has no known cause and its symptoms are subjective — as with chronic fatigue syndrome, fibromyalgia and other diseases whose characteristics and even existence have been contested — competing case definitions are almost inevitable. Now a new study of chronic fatigue syndrome has highlighted how competing case definitions can lead to an epidemiologic “Rashomon” — what you see depends on who’s doing the looking — and has stoked a fierce debate among researchers and patient advocates on both sides of the Atlantic. The study, published last month in The Lancet, reported that exercise and cognitive-behavioral therapy could help people with the illness. Advocates and some leading experts dismissed the findings and said the authors’ case definition was largely to blame. The British scientists who conducted the research identified study participants based largely on a single symptom: disabling and unexplained fatigue lasting at least six months. But many researchers, especially in the United States, say that definition takes in many patients whose real illness is not the syndrome but depression — which can often be eased with psychotherapy and exercise. The Lancet authors “have written their case definition to include both people with major depressive disorders and patients who clearly have received an insult to their immune systems and are depressed because they can no longer do things that they used to,” said Dr. Andreas Kogelnik, an infectious disease specialist in Mountain View, Calif., who treats many people with chronic fatigue syndrome. In studying the condition, he and other researchers exclude patients whose only symptom is fatigue, however disabling, and instead rely on a case definition that includes other cognitive, neurological and physiological symptoms. Those symptoms, they believe, indicate a complex immune system disorder possibly caused by a virus or another agent. Since 2009, studies have produced contradictory results over whether viruses related to mouse leukemia are associated with chronic fatigue syndrome, which is also called myalgic encephalomyelitis. A recent study found that people with the illness have distinct proteins in their spinal fluid, raising hope that a diagnostic test can someday be developed. No case definition is perfect; every disease has outliers. But whether a definition is broadly or narrowly drawn can profoundly affect the statistics vital for public health planning. A recent study of workers, for example, found that 2.5 percent to 11 percent suffered from carpal tunnel syndrome, depending on whether the case definition required reported symptoms, a physical exam, a nerve test or a combination of the three. Another study found that rates of acute gastroenteritis doubled when a looser case definition was used. If researchers filter their perceptions through different lenses — that is, case definitions that generate study populations varying in size and characteristics — it is hard to know whether they are studying the same phenomenon, overlapping ones or completely unrelated ailments. Determining whether findings from one study can be extrapolated to other patients becomes difficult at best. “You have to really define the characteristics, and everybody has to use the same criteria, because otherwise you’re calling something an apple and someone else is really looking at a peach and calling it an apple,” said Dr. Anita Belman, a neurologist at Stony Brook University who conducts research on pediatric multiple sclerosis. No one disputes that many people with chronic fatigue syndrome also suffer from depression. The question is which came first. Are patients depressed because a terrible disease has robbed them of their lives, or is the illness itself a somatic expression of an underlying depression? To researchers who believe that chronic fatigue syndrome is merely a psychological condition, that distinction may not seem important. But it matters deeply to those convinced it is a viral disease, who say the exercise therapy advised by the Lancet study can cause major relapses in people with chronic fatigue syndrome — a claim supported by some patient surveys. The single-symptom case definition used by the Lancet authors, known as the Oxford criteria, was developed in Britain in 1991. Like the team that conducted the current study, the 1991 group included prominent mental health professionals. But many scientists and clinicians view a multisymptom case definition published in 1994 by the Centers for Disease Control and Prevention in the United States as the international standard. In addition to six months of unexplained, disabling fatigue, the C.D.C. definition requires at least four of eight common symptoms: cognitive problems, sleep disorders, muscle pain, joint pain, headaches, tender lymph nodes, sore throat and what is called “postexertional malaise”— a relapse that occurs after even minimal activity. In 2005, the agency unveiled an “empirical” case definition that recommended specific screening questionnaires and cutoff scores for measuring fatigue, physical dysfunction and other symptoms. Critics challenged these newer guidelines on the same grounds as the Oxford criteria, arguing that the questionnaires and scoring methods were too ambiguous. In contrast, a 2003 case definition from Canada is considered the most restrictive and is preferred by many patients. It elevates postexertional malaise to a central role in the illness and requires a range of neurological, cognitive, endocrine or immunological symptoms. In 2009, researchers from DePaul University in Chicago reported that 38 percent of patients in a study sample suffering from depression alone were given misdiagnoses of chronic fatigue syndrome using the C.D.C. screening tools but not the narrower Canadian definition. The study suggests that the disease centers’ “empirical case definition has broadened the criteria such that some individuals with a purely psychiatric illness will be inappropriately diagnosed” with chronic fatigue syndrome, wrote Leonard A. Jason, a professor of community psychology at DePaul, and his colleagues. The authors also noted that using the new screening tools, the C.D.C. had greatly increased its estimate of the prevalence of the illness, to 2.5 percent of the population, or four million Americans. So the question remains: can therapy and exercise help patients with chronic fatigue syndrome, as the Lancet study reported? Yes, apparently — if the illness is identified with a case definition relying on fatigue alone. But does the evidence from that study prove that these strategies would help patients identified as having chronic fatigue syndrome through very different criteria? That is a much tougher argument to make.