Chapter  42:  Defining and Managing Chronic Fatigue Syndrome: Evidence Report/Technology Assessment Number 42

Objectives

Objectives of this evidence report are to summarize research evidence regarding the case definitions, prevalence, natural history and therapy of chronic fatigue syndrome (CFS).

Search Strategy

English and non-English citations were identified through July 2000 from four electronic bibliographic databases (MEDLINE, The Cochrane Library, PsycINFO, EMBASE), CFS Internet sites, the Journal of Chronic Fatigue Syndrome, references of pertinent articles, textbooks, and experts. The electronic search was updated through October 2000 using PubMed; experts identified relevant citations up to January 2001.

Selection Criteria

Published and unpublished studies that were conducted after 1980 and that involved adults with CFS were reviewed.

Data Collection and Analysis

Two reviewers (physician, psychometrician, research methodologist, and/or nurse) independently abstracted data from the selected studies. Data were synthesized descriptively, emphasizing the quality and methodologic design of studies. Meta-analyses were not done because of marked heterogeneity of study designs.

Main Results

• There are four well-recognized case definitions of CFS, and the Centers for Disease Control and Prevention (CDC) is spearheading the development of a fifth. Definitions, developed primarily by expert knowledge and consensus, have evolved over time. A few comparative research studies support the concept of a condition, characterized by prolonged fatigue and impaired ability to function, which is captured by the case definitions. The superiority of one case definition over another is not well established. The validity of any definition is difficult to establish because there are no clear biologic markers for CFS, and no effective treatments specific only to CFS have been identified.

• Findings from surveys show that the prevalence of CFS in community populations is probably less than 1% and in primary care populations less than 3%. The reliability of these estimates is limited, because surveys used different case definitions and varied assessment and reporting methods, and sometimes had poor response rates.

• Precise estimates of recovery, improvement, and/or relapse from CFS are not possible because there are few natural history studies and those that are available have involved selected referral populations or have used varying case definitions and followup methods.

• Thirty-eight controlled trials evaluating multiple treatment interventions show the following mixed results:

Immunologic Therapy

Evidence from 11 trials is scant and insufficient to conclude whether immunological therapies, such as immunoglobulin, Ampligen, Acyclovir, interferon, and transfer factor, are effective or ineffective.

Corticosteroids

Evidence from four trials suggests that there are no consistent benefits from mineralocorticoids (fludrocortisone), but that there are some improvements in fatigue and functional status with glucocorticoids (hydrocortisone). However, glucocorticoid therapy may severely suppress adrenal function.

Antidepressants

Evidence from five trials show that there is no consistent pattern of benefit from antidepressant therapy, though some participants in these trials experienced improved vigor and less anxiety.

Behavioral Therapies

Evidence from nine trials generally show that behavioral interventions that emphasize increased activity levels result in improvements in fatigue, overall well-being, quality of life and functional status.

Other

Evidence from trials is scant and insufficient to conclude whether complementary therapies, such as homeopathy, massage therapy and osteopathy are effective or ineffective. Evidence from trials is scant and insufficient to conclude whether multiple other therapies, such as nicotinamide adenine dinucleotide (NADH), galanthamine, growth hormone, essential fatty acids, and liver extract therapies, are effective or ineffective. Findings from one small trial suggest that magnesium therapy may improve energy, overall well-being, pain and distress in patients with CFS and magnesium deficiency.

Conclusions

Existing case definitions for CFS appear to characterize a group of people with prolonged fatigue and impaired ability to function. The validity and superiority of any particular case definition are not well established. Surveys suggest that the prevalence of CFS in community populations is less than 1%. Precise estimates of rates of recovery, improvement and/or relapse from CFS are not available. Although several therapies have been studied, potential benefits as well as harms of most therapies are not well established. Behavioral interventions that emphasize increasing activity levels may improve quality of life and function in some people with CFS.

This document is in the public domain and may be used and reprinted without permission except those copyrighted materials noted for which further reproduction is prohibited without the specific permission of copyright holders.

Suggested Citation

Mulrow CD, Ramirez G, Cornell JE, et al. Defining and Managing Chronic Fatigue Syndrome. Evidence Report/Technology Assessment No. 42 (Prepared by San Antonio Evidence-based Practice Center at The University of Texas Health Science Center at San Antonio under Contract No. 290-97-0012). AHRQ Publication No. 02-E001. Rockville (MD): Agency for Healthcare Research and Quality; October 2001.

Overview

This evidence report is a systematic review that summarizes scientific literature about the following aspects of chronic fatigue syndrome (CFS) in adults: case definitions, prevalence and natural history, and treatment. In an effort to organize and clarify this body of research knowledge, the Agency for Healthcare Research and Quality (AHRQ) contracted for this evidence-based review with the San Antonio Evidence-based Practice Center (EPC). The National Institute of Allergy and Infectious Diseases nominated the topic for an evidence-based review.

Reporting the Evidence

Initially, a broad-ranging list of more than 20 questions was considered for the evidence report. Seventeen technical experts from the United States, Canada, Australia and the United Kingdom used a Delphi consensus process to prioritize questions that the evidence report could realistically address, given the enormity of the data, and the limits on time and resources. The scope of the report was narrowed to the following high priority questions:

1. What are the existing case definitions of CFS in adults?

2. Which case definitions, if any, have been substantiated and/or validated with reliably discriminating constellations of symptoms in adults?

3. What are the prevalence and natural history of CFS in adults?

4. Do controlled studies in adults show that particular therapies improve clinical symptoms of CFS when compared to placebo, no therapy, or each other?

Methodology

Findings

• There are four well-recognized case definitions of CFS exist. The Centers for Disease Control and Prevention (CDC) is currently spearheading the development of a fifth definition. Definitions have been developed primarily by expert knowledge and consensus processes, and have evolved over time. A few comparative research studies support the concept of a condition, characterized by prolonged fatigue and impaired ability to function, which is captured by the case definitions. The superiority of one case definition over another is not well established. The validity of any definition is difficult to establish because there are no clear biologic markers for CFS and no effective treatments that are specific only to CFS have been identified.

• Findings from surveys that have involved at least 100 adult participants suggest the prevalence of CFS in community populations is less than 1 percent. Prevalence rates reported in surveys conducted in primary care settings range from approximately 0.04 percent to 2.6 percent. Ability to interpret the reliability of these estimates is limited by the following factors: use of different case definitions, variability in assessment and reporting methods, and poor response rates in some studies.

• Prospective natural history studies have varied findings. Precise estimates of recovery, improvement, and/or relapse are not possible because there are few natural history studies and those that are available have involved selected referral populations or have used varying case definitions and followup methods. Rates of self-reported global improvement in symptoms at 12 to 18 months range from 11 percent to 64 percent. Rates of self-reported worsening of symptoms at 12 to 18 months range from 15 percent to 20 percent. Investigators from one study estimate that the cumulative probability of recovery from CFS at 5 years is approximately 30 percent.

• Thirty-eight controlled trials evaluated a heterogeneous mix of interventions and had mixed results.

Future Research

There is no shortage of questions for future CFS research. Our technical experts and peer reviewers identified the following questions as high priority.

Report's Purpose and Scope

Several thousand references to chronic fatigue syndrome (CFS) appear in MEDLINE and other medical, bibliographic research databases. In an effort to organize and clarify this body of research knowledge, the Agency for Healthcare Research and Quality (AHRQ) contracted for this evidence-based review of the CFS medical literature with the San Antonio Evidence-based Practice Center (EPC). The National Institute of Allergy and Infectious Diseases nominated the topic and partnered with the San Antonio EPC to utilize the report in their research agendas.

Initially, a broad-ranging list of more than 20 questions was suggested for the evidence report. Seventeen technical experts from the U.S., Canada, and the U.K used a Delphi consensus process to prioritize questions that the evidence report could realistically address, given the enormity of the data, and the limits on time and resources. They narrowed the initial scope of the report to the following high priority questions:

1. What are the existing case definitions for CFS in adults?

2. Which case definitions, if any, have been substantiated and/or validated with reliably constellations of symptoms in adults?

3. What are the prevalence and natural history of CFS in adults?

4. Do controlled studies in adults show that particular therapies improve clinical symptoms of CFS, when compared to placebo, no therapy, or each other?

Defining Chronic Fatigue and Chronic Fatigue Syndrome

The symptom of fatigue may be defined by a pervasive sense of tiredness or lack of energy that is not related exclusively to exertion.¹ This type of fatigue is not alleviated by rest and must be distinguished from weakness, malaise, and temporary tiredness that occur as a direct result of excessive physical or mental exertion. In order to qualify fatigue as pathological, researchers have described the symptom as a continuum, much like blood pressure or obesity.² Although no obvious cutoff exists between normal and abnormal fatigue, assessment of fatigue using unique qualifiers, such as duration and degree of functional impairment, may help distinguish between the transient, mild states and more severe, prolonged disorders.

Researchers and clinicians distinguish prolonged fatigue from both chronic fatigue and CFS.³ Prolonged fatigue is defined as disabling fatigue that lasts at least one month. If this degree of fatigue lasts at least six months, it is called chronic. If the fatigue syndrome is disabling, lasts at least six months, is unexplained by other medical or psychological conditions, and is associated with a requisite number of other symptoms, it is considered CFS. Scientists have proposed other more specific definitions of CFS. These are discussed in the Results Section of the report entitled, “What are the existing case definitions for chronic fatigue syndrome?”

Burden of Disease

Fatigue is commonly reported by people in community settings and by patients attending primary care practices. For example, as many as 9% to 13% of people in community settings, and as many as 10% to 25% of people visiting general practitioners complain of prolonged and disabling fatigue.^4,56,7 Most such people do not have CFS. They do not have accompanying symptoms (e.g., sore throat, impaired memory or concentration, tender lymph nodes, myalgia, arthralgia, headaches, sleep disturbance, fever, muscle weakness) that are required for current definitions of CFS, and they may have medical or psychiatric conditions that explain their prolonged fatigue. The actual prevalence of prolonged fatigue states that meet CFS definitions is described later in this report under the Results section entitled, “What is the prevalence and natural history of CFS?”

People with CFS have described their burden of illness in various manners.^8-10 For example, some people describe CFS as a state of illness that is experienced as “an integrated whole encompassing body, mind, and spirit.”^9,10 They may experience fatigue as a flu-like illness or a sensation of heaviness or being weighted down or of being drained, groggy, foggy-minded, or drugged.^8,10 Others describe a lack of energy or stamina, or multiple cognitive problems such as an inability to complete short, routine tasks or difficulty concentrating.^8,10 Some people describe fluctuating courses of illness characterized by sudden and unexpected changes in symptoms.^8,9 They may feel marginalized, socially isolated, embarrassed, or stigmatized.^8,9 Some people find CFS extremely debilitating, such that they lose their social and occupational functioning.

History and Terms Used to Describe Chronic Fatigue Syndromes

Disorders characterized by severe chronic fatigue have been reported in the literature as early as the 18^th Century.¹¹ In 1750, Manningham described a syndrome of severe fatigue with symptoms of low-grade fever, weariness throughout the body, and pains.¹² In 1869, George Beard, a New York physician, used the term neurasthenia to describe a chronic fatigue syndrome.² The diagnosis of neurasthenia was given to patients with general nervousness, hysteria, or minor depression, as well as to patients with fatigue. Neurasthenia is still listed in the World Health Organization's International Classification of Diseases as a syndrome of mental and physical fatigue of at least three months' duration.

In the early 1900s, the concept that chronic fatigue followed infections, such as influenza and typhoid, was established and widely accepted.² For example, myalgic encephalomyelitis, also known as Royal Free Disease, was originally thought to represent a post-viral fatigue syndrome. In 1955, an epidemic of fatigue with myalgia, and sensory motor symptoms occurred at several Royal Free Hospital branches in London. An infectious viral etiology that affected the brain and muscle was initially considered.²

In the mid-1980s, associations between Epstein-Barr virus and chronic fatigue were reported.^13-17 The proposed Epstein-Barr etiology received enormous media and medical attention and led to a consensus conference organized in 1985 by the National Institute of Allergic and Infectious Diseases. Numerous CFS groups were developed and began lobbying for answers to their questions about the condition. Publications addressing chronic mononucleosis, and another U.S. epidemic in the Lake Tahoe, Nevada area convinced researchers that methods to study the nosology and causality of CFS were needed.^16,18,19

In the last two decades, several case definitions for CFS have been proposed, and there have been multiple studies aimed at describing particular manifestations of patients with chronic fatigue syndrome. Investigators have conducted surveys aimed at describing the prevalence of CFS in community and clinical settings, as well as longitudinal studies describing the natural history of CFS. Several trials have evaluated whether various therapies improve symptoms and outcomes. Finally, hundreds of studies examining potential etiologies and diagnostic markers for CFS have been conducted. The purpose of this evidence report is: a) to organize and categorize research related to CFS; b) describe existing case definitions of CFS and how they are substantiated; c) assess the prevalence and natural history of CFS in adults; and d) summarize trials of therapies for adults with CFS.

Expert Input

This evidence report was produced by a multidisciplinary group that included: Cynthia Mulrow, MD, MSc, a general internist, Gilbert Ramirez, DrPH, a research methodologist, and John Cornell, PhD, a psychometrician. Martha Harris, AHIP, MLS, MA, a medical librarian, conducted literature searches and maintained the reference databases. Christine Aguilar, MD, MPH, a physician research associate, and Jodi Sapp, RN, a nurse research associate, abstracted information from the articles; David Mullins developed electronic databases for the project and provided data entry/data validation support. Drs. Ramirez and Cornell calculated effect size statistics for comparing study outcomes. Drs. Mulrow, Ramirez and Aguilar coordinated the project administratively. Karen Stamm edited and formatted the report, and coordinated communication with technical experts and peer reviewers. Drs. Ramirez, Cornell, and Mulrow wrote the report.

An international expert panel provided input regarding: the plan and format for the evidence report; interpretation and presentation of findings; critique of drafts; and future research priorities. Members of the panel from the United States included: Dedra Buchwald, MD, University of Washington-Seattle; Mark Demitrack, MD, Eli Lilly and Company; Sudhir Gupta, MD, PhD, University of California-Irvine; Leonard A. Jason, Ph.D., DePaul University; James F. Jones, MD, National Jewish Medical and Research Center; Nancy G. Klimas, MD, University of Miami; Kathy Rabin, JD, a CFS patient; John H. Renner, MD, National Council for Reliable Health Information (until his death in September 2000); Joan Shaver, PhD, RN, University of Illinois-Chicago; Eng M. Tan, MD, Scripps Research Institute; and Vicki C. Walker, BA, Research and Public Policy Project Manager, The CFIDS Association of America. Non-U.S. members of the technical expert panel included: Susan Abbey, MD, University of Toronto, Canada; Peter Campion, PhD, FRCGP, MRCP (UK), DCCH, University of Hull, United Kingdom; Robert H. Loblay, MD, University of Sydney, Australia; and Peter D. White, MD, St. Bartholomew's and the Royal London School of Medicine and Dentistry, Queen Mary and Westfield College, London, United Kingdom. Ex-officio members included: David Morens, MD, National Institute for Allergy and Infectious Diseases, and William Reeves, MD, MSPH, Centers for Disease Control and Prevention. In addition, the report's penultimate draft was peer-reviewed by 15 individuals selected for their expertise in CFS (see Appendix A for a full list of external peer reviewers).

Questions Addressed in Evidence Report

The agency that nominated the topic initially suggested a broad-ranging list of more than 20 questions for our report to address. We enlisted the help of the technical expert panel to determine which questions we could realistically address, given the enormity of the data, and the limits on time and resources. Using a consensus Delphi process, the technical experts narrowed our initial scope to the following highest priority questions:

1. What are the existing case definitions for CFS?

2. Which case definitions, if any, have been substantiated and/or validated with reliably discriminating constellations of symptoms in adults?

3. What are the prevalence and natural history of CFS in adults?

4. Do controlled studies in adults show that particular therapies improve clinical symptoms of CFS, when compared to placebo, no therapy, or each other?

Literature Search and Selection Methods

Selection Process

Figure 1. Literature search results flow chart

   Figure 1. Literature search results flow chart

Figure 1 depicts the selection process. Of 5,221 identified records from the various electronic databases and sources, 1,111 were eliminated because they were duplicate citations cited in more than one database (i.e., same authors, title, journal source information). Two reviewers independently screened the titles and abstracts of the remaining 4,110 records to select articles that were relevant to evidence questions one through four. Of the 4,110 records that were screened, 518 were considered potentially relevant to the specified evidence report questions. Of 124 that addressed case definitions for CFS, 41 met selection criteria. Of 83 articles that addressed prevalence or natural history, 20 met the selection criteria. Of 311 articles that addressed therapies, 38 were controlled trials that met selection criteria.

Data Abstraction Process

Two reviewers (physician, psychometrician, methodologist and/or nurse) independently abstracted data from the selected studies. None of the abstractors were blinded either to study title or to author's names. Disagreements in abstractions were uncommon (less than 1% of items) and were resolved by consensus. No formal reliability testing was done.

We critically appraised the quality or rigor of studies by assessing several parameters. For studies addressing prevalence, we assessed: the method and time course of ascertainment, the sampling method, whether the sample was representative, whether an established case definition was used, and the response rate. For studies addressing natural history, we assessed: sampling methods, whether the sample was representative, whether an established case definition was used, methods of followup, and followup rates. For studies addressing substantiation of case definitions, we assessed: incorporation and selection biases, whether the study sample was representative, how the diagnosis was verified and by whom, whether compared groups were equivalent with respect to key sociodemographic or clinical variables that could affect the manifestations being studied, and whether groups were matched on such confounding variables. For studies addressing therapy, we assessed: whether the controlled study was randomized, the adequacy of randomization (method and concealment of assignment); whether the trial was single or double-blind; numbers of dropouts; and appropriateness of analyses methods.

Data Synthesis Process

We synthesized data descriptively, emphasizing methodological characteristics of the studies, such as sources of populations enrolled, CFS case definitions used to select study participants, sample sizes, adequacy of randomization process, interventions and comparisons. We examined relationships between clinical outcomes, participant characteristics, and methodological characteristics in evidence tables and graphical summaries, such as forest plots. We grouped outcomes according to a list that had been developed specifically to assess symptoms in patients with CFS.²⁰

To help compare outcomes across studies, we computed standardized mean differences between treatment and comparison group scores as a measure of effect size for each study. These estimates were adjusted for between-group differences at baseline and for small sample bias.²¹ We adjusted for baseline differences by calculating an “effect size” at baseline; by definition, it should be zero if study groups were well matched. When we found a non-zero “effect size” at baseline, we adjusted outcome effect sizes were by subtracting the baseline effect size.

Question 1. What are the Existing Case Definitions for Chronic Fatigue Syndrome?

In 1988 Holmes et al. published the first case definition for CFS in the mainstream peer reviewed medical literature. The Centers for Disease Control and Prevention (CDC) adopted this definition. To meet this definition, patients had to present with new onset of persistent or relapsing, debilitating fatigue or easy fatigability with no previous history of similar symptoms. The fatigue could not be alleviated by bed rest, and had to impair daily activity below 50% of premorbid activity for at least 6 months. Eight or more of 11 minor symptom criteria had to be met. Other medical and psychiatric conditions that produced similar symptoms by proper history, exam, and laboratory analysis had to be excluded. Symptoms must have occurred at or after the onset of fatigue and have continued or been recurrent for at least 6 months.

Symptom criteria included: 1) mild fever or chills, 2) sore throat, 3) painful lymph nodes, 4) unexplained generalized muscle weakness, 5) myalgia, 6) post-exertional fatigue (lasting more than 24 hrs), 7) new onset generalized headaches, 8) migratory arthralgia without joint swelling or redness, 9) neuropsychiatric complaints (one or more of the following: photophobia, forgetfulness, excessive irritability, confusion, difficulty in thinking or concentrating, depression), 10) sleep disturbance (hypersomnia or insomnia), and 11) acute onset (over a few hours to a few days) of main symptom complex. To count, physical criteria had to be documented by a physician on at least two occasions, at least one month apart. Physical criteria included: 1) low grade fever, 2) nonexudative pharyngitis, and 3) palpable or tender anterior posterior cervical or axillary lymph nodes.

Australian and British working case definitions for CFS were developed in the early 1990s. Several Australian definitions were proposed. One of the most widely referenced definitions was used and reported by Lloyd et al.²² These criteria required the following for a diagnosis of CFS: 1) chronic persisting or relapsing fatigue of a generalized nature for greater than 6 months that is exacerbated by minor exercise and that causes significant disruption of usual daily activities, 2) neuropsychiatric dysfunction including impairment of concentration evidenced by difficulty in completing mental tasks which were easily accomplished before the onset of the syndrome and new onset of short-term memory impairment, and 3) no alternative diagnosis found by history or physical exam over a six-month period. Persons with a previous history of psychosis were excluded from the accepted cases.

In 1991, Sharpe, Wessely and White along with a multidisciplinary group of CFS researchers proposed a set of criteria known as the Oxford or British criteria.²³ Their case definition required a minimum duration of six months of fatigue that had a definite onset and that was associated with functional impairment. As with the Australian definition, psychiatric illness was not an exclusion criterion, except in the instance of either substance abuse or psychosis. They defined multiple associated symptoms, such as myalgia, mood disturbance, and sleep disturbance. Post-infectious fatigue was listed as a subcategory of CFS with identical features occurring after a known infective episode.

Schluederberg et al. suggested a refinement to the 1988 CDC definition in 1992.²⁴ The proposed refinement was based on proposed changes discussed at a 1991 National Institute of Allergy and Infectious Diseases/National Institute of Mental Health workshop. The proposed changes were that a diagnosis of CFS could be made only excluding: 1) specific psychiatric disorders (psychotic depression, bipolar disorder, and schizophrenia); 2) substance abuse; and 3) specific infectious diseases (e.g., chronic active hepatitis, HIV infection). Additionally, the proposed changes allowed for a diagnosis of CFS co-existing with the following: 1) fibromyalgia; 2) etiologically known, treatable infectious diseases that have resolved in terms of their acute clinical and laboratory features, but in whom fatigue and other subjective complaints persist despite appropriate therapy; 3) non-psychotic depression, either concurrent, 1 month post-onset or 6 months or more before onset; and 4) panic, generalized anxiety, or somatoform disorders. Many of these suggestions were incorporated into the 1994 CDC definition that is described below.

In 1994, a second CDC definition was proposed by an international collaborative group that included authors of the 1988 CDC, 1990 Australian, and 1991 British definitions.²⁵ This definition was developed in response to critiques that the original Holmes' CDC definition limited the number of CFS cases diagnosed because of its strict exclusion of patients with histories of psychiatric complaints. The revised 1994 definition did not consider all psychiatric disorders, such as anxiety disorders and less severe forms of depression, as exclusionary diagnoses for CFS. Specifically, the 1994 definition does not exclude anxiety disorders and less severe forms of depression, but does exclude substance abuse, eating disorders, dementia and psychosis including psychotic depression. Other changes included eliminating all physical signs from the inclusion criteria, decreasing the number of required symptoms from 8 to 4, and limiting the list of symptoms to 8 from 11.

Table 1. Case definitions of CFS

Table 1. Case definitions of CFS

Criteria	CDC-1988	CDC-1994	Australian 1990	UK/Oxford-1991
Minimum duration (months)	6	6	6	6
Functional impairment	50% decrease in activity	Substantial	Significant disruption of usual activities	Disabling
Cognitive or neuropsychiatric symptoms	May be present	May be present	Required	Mental fatigue required
Other symptoms	8 required	4 required	Not specified	Not required, but other symptoms may be present, particularly myalgia, mood and sleep disturbance
	Neuropsychologic complaints (one or more of the following: photophobia, transient visual scotomata, forgetfulness, excessive irritability, confusion, difficulty thinking, inability to concentrate, depression)	Substantial impairment in short-term memory or concentration
	Sore throat	Sore throat
	Painful lymph nodes in the anterior or posterior cervical or axillary distribution	Tender lymph nodes
	Muscle discomfort or myalgia	Muscle pain
	Migratory arthralgia without joint swelling or redness	Multi-joint pain without swelling or redness
	Generalized headaches (of a type, severity, or pattern that is different from headaches the patient may have had in the premorbid state)	Headaches of a new type, pattern or severity
	Sleep disturbance	Unrefreshing sleep
	Prolonged (24 hours or greater) generalized fatigue from levels of exercise that would have been easily tolerated in the patient's premorbid state	Post-exertional malaise lasting more than 24 hours.
	Mild fever or chills
	Unexplained generalized muscle weakness
	Description of the main symptom complex as initially developing over a few hours to a few days
New onset	Required	Required	Not required	Required
Medical exclusions	Extensive list of known physical causes	Clinically important	Known physical causes	Known physical causes
Psychiatric exclusions	New onset or history of: anxiety disorder, depressive disorder including endogenous depression and bipolar disorder, schizophrenia, substance abuse	Melancholic (severe, major) or psychotic depression, substance abuse, bipolar disorders, eating disorder, schizophrenia, dementia, delusional disorders	Psychosis, bipolar disorder, substance abuse, eating disorder	Psychosis, bipolar disorder, eating disorder, organic brain disease

Studies that have evaluated the validity of different definitions and manifestations of fatigue are described under the question entitled, “Which case definitions, if any, have been substantiated or validated with reliably discriminating constellations of symptoms?” Of note, existing case definitions have been developed primarily using expert opinion and/or expert consensus processes. None have been well validated. The CDC is spearheading ongoing efforts to continuously refine the definition of CFS; a newly refined case definition is expected in 2001 (personal communication).

Question 2. Which Case Definitions, if any, Have Been Substantiated or Validated with Reliably Discriminating Constellations of Symptoms?

Simple Group Comparison Studies

Table 4. Examples of reported prevalence of chronic (more...)

Table 4. Examples of reported prevalence of chronic fatigue syndrome among large community and primary care populations

Study	Sample	Criteria	Ascertainment	Prevalence
Jason61 1999 Chicago N = 28,673	Community Response rate: 65% for initial telephone screen; 41% of persons with CFS-like symptoms completed physician examination	CDC 1994	Self-reported symptoms followed by physician examination	0.42 %
Steele81 1998 San Francisco N = 16,970	Community Response rate: 87%	CDC 1994	Self-report	0.08 to 0.2%
Kawakami82 1998 Japan N = 508	Community Response rate: 27%	CDC 1988 Oxford Australian	Self-report	0% 1.5% 1.5%
Wessely83 1997 Britain N = 2,376	Primary care Response rate: 83%	CDC 1988 CDC 1994 Oxford Australian	Self-report**	1.2% 2.6% 2.2% 1.4%
Lawrie84 1997 Scotland N = 695	Community Response rate: 78%	Oxford	Self-report	0.7% (0.4% 1yr incidence)
Lawrie85 1995 Scotland N= 1,039	Community Response rate: 71%	Oxford	Self-report	0.6%
Jason86 1995 Chicago n = 1,031	Community Response rate: 89%	CDC 1988 CDC 1994 Oxford Australian	Interviews and medical records	0.1% 0.1% 0.2% 0.2%
Buchwald87 1995 Seattle N = 4,000	Health maintenance organization roster Response rate: 77%	CDC 1988	Self-report	0.1 to 0.3%
McDonald88 1993 Britain N = 686	Primary care Response rate: not reported; 1071 questionnaires completed from all adults of a 7500 patient population	Oxford	Self-report	2.5%
Gunn79 1993 US N = 590	Surveillance referrals to CDC Response rate: 57%	CDC 1988	Referral by physicians	0.004 to 0.01%
Bates5 1993 US N = 1,000	Primary care Response rate: 99%	CDC 1988 Oxford Australian	Self-report	0.3% 0.4% 1%
Price80 1992 US N = 13,538	Community (ECA) Response rate: 75%	CDC 1988*	Self-report	0.007%
Lloyd22 1990 Australia N = 104 practitioners N = 500 CFS patients	Surveillance referrals from general practitioners Response rate: 98%	Australian	Practitioner-report	0.04%

* Retrospective assessment of only 7 of 11 minor symptoms

** Physical findings were not included for CDC 1988 definitioN

The 30 studies that compared frequencies of specific symptoms among different groups of patients are summarized in Table 4. The table lists studies according to the definition of CFS that was used. It tabulates study results regarding the predominant symptoms that are listed in the various case definitions, and shows comparisons between CFS patients and other groups.

There are several problems with the studies that compare symptoms among groups of patients. First, participants with CFS in these studies were chosen because they met criteria for one of the four definitions of CFS. This results in incorporation bias, whereby “common symptoms” of CFS are bound to occur at a much higher frequency than in unselected samples. Second, many of the studies involved only referred patients who were followed in specialty clinics. Third, studies used different definitions and instruments to measure the same symptom(s). Fourth, not all studies evaluated the same symptoms. Even if measured, the authors may have only reported the symptoms that showed significant differences between groups.

The authors most commonly reported the following: functional status, fatigue, concentration ability, and depressive symptoms. Compared to healthy participants as well as patients with multiple sclerosis, participants with CFS almost always had greater impaired function, more problems with concentration, more depressive symptoms and greater fatigue. Compared to depressed participants, study participants with CFS generally had greater impaired function and more somatic symptoms, but fewer depressive symptoms.

In one study, Komaroff examined the adequacy of the CDC 1988 case definition and the frequency of 11 minor symptoms in patients with CFS compared to patients with multiple sclerosis or depression.⁷² He found that 92% of the patients with CFS had at least 8 minor symptoms. Symptoms that commonly occurred in all three groups of patients were muscle weakness, arthralgia and sleep disturbance. Symptoms that appeared more unique to patients with CFS that were not listed in the CDC 1988 definition were anorexia and nausea.

Multivariate Classification Studies

Table 5. Prospective studies with one-year or greater (more...)

Table 5. Prospective studies with one-year or greater followup data on outcomes of CFS

Study	Participants (Age and illness duration in years)	Diagnostic Criteria and Followup	Baseline Measures	Followup Measures
Reyes106 1999 N = 155	155 adults from CDC surveillance system Women: 85% White: 97% Median illness: 4.4	CDC 1988 Followup: 91% Median followup time: 5.9 yrs		Cumulative probability of recovery at 5 yrs: 31% and at 10 yrs 48% 57% of patients reporting recovery continued to feel recovered 6 months later
Russo108,119 1998 US N = 98	98 adults from referral CFS clinic Women: not given Avg. age: 40 Avg. illness: 5.5	CDC 1988 Followup: 80% Average followup time: 2.5 yrs	Health perceptions 6.37 + 12.84 Physical functioning 36.98 + 30.21 Role functioning 23.16 + 35.17 Mental health 51.14 + 21.74 General health questionnaire 12.67 + 7.86 Modified symptom questionnaire 4.22 + 4.47	12.68 + 18.59* 39.08 + 33.17 27.57 + 38.67 62.63 + 21.95* 7.98 + 7.71* 2.74 + 2.90** *Significant improvement, p < .001 **p < .01 Predictors of improvement: younger age, fewer physical signs, and lower (worse) initial scores on mental health and general health questionnaire. 41% improved or recovered 69% no change or worse
Ray110,111 1997,1995 Britain N = 137	137 adults from a hospital out-patient clinic Women: 70% Avg. age: 37 Mean illness: 3.4	Oxford plus fatigue precipitated by minimal exertion Followup: 88% Followup time: 1 yr	Fatigue score 4.03 + 1.14 Impairment score 22.47 + 5.83	3.34 + 1.35* 20.12 + 7.25**Significant improvement, p < 0.001 64% improved 15% worsened 21% unchanged Poorer outcomes were predicted by illness duration, subjective cognitive impairment, and somatic symptoms and not by anxiety, depression, or general emotional distress.
Vercoulen90 1996 Netherlands N = 246	298 “self-referred” adults Women: 76% Avg. age: 39 Mean illness: 8.4	Oxford Followup: 83% Followup time: 1.5 yrs	57% employed	29% employed 3% recovered 17% improved 60% unchanged 20% worse Improvement was positively predicted by sense of control over symptoms, lower fatigue severity
Bombardier10391 1995 N = 445	226 adults from referral CFS clinic Women: 84% Avg. age: 38 Avg. illness 5	CDC 1988 with Schluederberg modification Followup: 89% Followup time: 1.5 yrs	37% unemployed 27% part time	61% some improvement 46% continued to meet symptom criteria for CFS 35% unemployed 19% part time Dysthymia reported at initial interview predicted overall less improvement
Wilson109 1994 Australia N = 139	103 adults previously enrolled in treatment trials (immunoglobulin, leukocyte extract, cognitive behavior) Women: 72% Avg. age: 41 Mean illness: 5.6 y	Australian Followup: 74% Mean followup time: 3.2 yrs		6% recovered 63% improved 30% unable to work 25% receiving disability 20% no physical activity 39% no social activity mean Karnofsky score 76.3 + 8.8
Tirelli107 1993 Italy N = 205	205 adults Setting: unclear Women: 61% Median age: 33 Median illness: 3	CDC 1988 Followup: unclear Followup time: 1 yr		1 yr followup 2% no symptoms 9% substantial resolution 89% persistent symptoms

Of 12 reports that used multivariate techniques to help distinguish groups of patients, 3 examined very focused or limited dimensions of CFS, ^52,54,57 3 characterized subgroups of patients with CFS, ^56,60,63 and 8 evaluated multiple dimensions of CFS (Table 5).^55,58-64

Studies that examined limited dimensions of CFS

Fischler compared sleep patterns between 49 patients who met the Oxford criteria for CFS with sleep patterns of 20 healthy participants.⁵² Sleep initiation and sleep maintenance disturbances were commonly observed among CFS patients and not among healthy participants. Using discriminant analysis, the authors showed that sleep disturbances correctly classified most of the patients with CFS.

Blakely⁵⁴ used discriminant analysis to investigate the psychological characteristics of patients with chronic fatigue syndrome. Forty-two patients with CFS as defined by the 1988 CDC case definition were compared to 39 patients with chronic pain and 72 healthy participants. There was considerable overlap in psychological symptoms between patients with CFS and patients with chronic pain; approximately one-fourth of the patients in these groups had similar psychological profiles. There was less overlap between patients with CFS and healthy participants; approximately 90% of the healthy participants had psychological profiles that were distinct from patients with CFS or chronic pain.

Vercoulen studied 51 patients who met Oxford criteria for CFS and 50 patients with multiple sclerosis.⁵⁷ He used structural equation modeling to explore a hypothesized model in which behavioral, cognitive, and affective factors might play a role in perpetuating fatigue. He found positive associations between cognitive and behavioral factors and the persistence of fatigue.

The three studies described above had multiple problems that limit their generalizability.^52,54,57 They involved relatively small selected study samples and focused on limited aspects of CFS. As such, their results must be considered exploratory.

Study that examined subgroups of patients with CFS

Hickie evaluated 565 patients who met Australian criteria for CFS.⁵⁶ Self-reported symptoms, depression, course of illness, impact of illness on social and occupational functioning, illness behavior and general health status were assessed. Using latent variable models, two groups of patients with CFS were identified. One group of patients (27% of the sample) frequently reported multiple severe symptoms, such as fatigue, headache, inability to concentrate, disturbed sleep and memory, joint and muscle pain, nausea, and palpitations. The other larger group of patients (73% of the sample) reported fewer and less severe symptoms. This latter, larger group was characterized by a greater proportion of men (33% v 18%), shorter duration of illness (5.9 v 8.3 years), less severe course of illness (61% v 80%), lower current psychiatric morbidity, fewer medical consultations (14.3 v 19.9 visits previous 12 months), less disability (49% v 60% currently unemployed; 23% v 38% receiving disability benefits; and 1.5% v 7.4% housebound), and less previous treatment with antidepressants (35% v 48%).

Studies that evaluated multiple dimensions of CFS

Vercoulen studied 298 self-referred patients with CFS who met Oxford criteria.⁵⁵ Participants completed a battery of functional status and general well-being instruments that included the following: the Checklist Individual Strength-CIS, self-reported complaints, the Beck Depression Inventory, the Sickness Impact Profile-SIP, the CIS Activity subscale, the Multidimensional Health Locus of Control, the SIP social life subscale, and the Symptom Checklist-SCL90 sleep problems subscale. Using principal component factor analysis, the authors identified nine dimensions of manifestations that characterized patients with CFS. The nine dimensions were psychological well-being, functional impairment in daily life, sleep disturbances, avoidance behavior, concentration problems, causal attributions related to the complaints, social functioning, and self-efficacy expectations, and subjective experience.

Friedberg⁶³ evaluated symptom patterns in CFS patients who had been ill for 10 or more years using principal component analysis. The 286 CFS patients had been diagnosed using the 1994 CDC case definition and were compared to 179 healthy controls (spouses/partners/significant others). The factor analysis resulted in three conceptually distinct dimensions with one factor, cognitive problems, accounting for 30% of the variance. The other two factors were flu-like symptoms (6.4% of the variance) and neurologic symptoms (4.1% of the variance).

There were a series of reports based on a single, large, unselected, random community sample. ^58-62,64 These studies examined multiple dimensions of CFS, rigorously assessed a comprehensive set of symptoms, and attempted to determine the diagnostic validity of the 1994 CDC criteria. Of 28,673 people who were called, 18,675 (65%) completed the study interview. Of these, 780 individuals reported chronic fatigue. They were classified into three subgroups: 408 CFS-like individuals (self-reported chronic fatigue with at least 4 of the CDC 1994 symptoms); 304 ICF-like individuals (idiopathic chronic fatigue); and 68 CF-explained-like individuals (chronic fatigue explained by medical or psychiatric conditions).

A principal component analysis resulted in a four-factor solution that explained 38% of the total variance. The four components were lack of energy, physical exertion, cognitive functioning, and fatigue and rest. Scores on these four components differed significantly among the three subgroups of patients (i.e., CFS-like, ICF-like, CF-explained) differed significantly. The CFS-like and the CF-explained-like groups had more severe lack of energy, exhibited more severe fatigue following physical exertion, and demonstrated more severe cognitive problems than the ICF-like group. The CFS-like group reported higher fatigue following rest or sleep compared to the ICF-like group that reported less severe fatigue.

In a followup analysis of 166 of the 408 CFS-like individuals who agreed to a structured psychiatric interview and a complete physical exam, individuals were reclassified as chronic fatigue syndrome (CFS), idiopathic chronic fatigue (ICF), chronic fatigue explained by a medical condition (CFM) or chronic fatigue explained by a psychiatric condition (CFP).⁶¹ The characteristics of the patients' fatigue fell into three clusters: 1) relatively low post-exertional fatigue; 2) most severe post-exertional fatigue and most improvement in fatigue following rest; and 3) high post-exertional fatigue and fatigue not alleviated by rest. The majority of patients with CFS (69%) fell into cluster 3.

Jason and colleagues conducted additional analyses to evaluate the diagnostic validity of somatic symptoms associated with five functional somatic syndromes: CFS, fibromyalgia, irritable bowel syndrome, depression, and anxiety.⁶⁴ Confirmatory factor analysis on 21 symptoms yielded a five-factor solution that specifically distinguished the five functional somatic symptoms from one another. The factor grouping for CFS included chronic fatigue, sore throat, lymph node pain, post-exertional malaise, memory/concentration problems, and unrefreshing sleep. The five-factor solution had reasonably good fit for the data. In contrast, a confirmatory factor analysis that tested the proposition that all five disorders represent a single underlying functional somatic distress syndrome did not fit the data well. Logistic regression showed that only the CFS symptom factor was significantly associated with actual CFS diagnosis as determined by two independent physicians.

Question 3. What are the Prevalence and Natural History of Chronic Fatigue Syndrome?

Prevalence

Several factors influence the reported prevalence of chronic fatigue syndrome including the following:

the definition or criteria that are used for diagnosis;

the relative intensity of methods used to exclude other medical and/or psychiatric conditions;

the method (e.g., self-report, physician report, chart review) of ascertainment;

the time course of ascertainment (e.g., retrospective, prospective);

the type of population that is studied (e.g., community, clinic);

the sampling method (e.g., random, consecutive, referred, volunteers); and

the response rate of the population that is studied.

In general, more restrictive definitions of chronic fatigue syndrome, such as those that exclude concomitant psychiatric diagnoses, require longer duration of fatigue, and/or demand that multiple other medical illnesses have been excluded, lead to lower estimates of prevalence compared to less restrictive definitions. In general, community populations have lower prevalence of illness and disease than clinic or hospital populations. However, conditions that are particularly mild, under-recognized, not accepted/addressed by the medical care system, or perceived as stigmatizing for social or financial reasons could conceivably be as or more common in community than clinical populations.

Table 2. Simple group comparison studies on CFS case-defining (more...)

Table 2. Simple group comparison studies on CFS case-defining symptoms

Case Definition	Author	FUNC	FATG	MLSE	THRO	LYMP	MUSC	JOIN	HEAD	SLEP	MEMR	CONC	FEVR	WEAK	DEPR
CFS vs. Healthy Controls
CDC 1988	Komaroff72 1996			x	x	x	x	x	x	x	x	x	x	x	x
	Krupp26 1993		x							x					x
	Sandman27 1993										x
	Schweitzer28 1995	x
	Buchwald29 1996	x
	Komaroff30 1996	x
	Tiersky31 1998											x
Australia 1990	Vollmer-Conna32 1997														x
Oxford 1991	Smith33 1993											x			x
	Swanink34 1995	x	x									x			x
	Smith35 1996	x	x							x
	Fischler52 1997									x
	Vercoulen36 1996	x	x	x		x	x	x	x	x	x	x		x
	Bazelmans37 1997	x	x												x
	Wearden38 1997											x
	Wessely41		x		x		x	x	x	x
	Vercoulen39 1998										x	x
	Smith40 1999		x		x	x	x	x	x	x	x	x	x		x
CDC 1994	Wessely41 1996			x	x	x	x	x	x	x	x		x	x
	Baraniuk42 1998		x
	Blenkiron43 1999		x
	Jason44 2000			x	x	x	x	x	x	x	x	x
	Jason45 2000	x
CFS vs. Depression
CDC 1998	Hickie46 1990
	Komaroff72 1996			x	x	x	x	x	x		x		x	x
	Komaroff30 1996	x
	Buchwald29 1996	x	x
Australia 1990	Vollmer-Conna32 1997
Oxford 1991	Morriss47 1999		x
CDC 1994	Johnson48 1996
CFS vs. Multiple Sclerosis
CDC 1998	Schweitzer28 1995	x
	Komaroff72 1996			x	x	x	x		x	x		x	x		x
	Krupp26 1993		x							x					x
CDC 1994	Johnson48 1996

Symbols x -chronic fatigue patients had greater difficulty than patients in the comparison group, and the difference was statistically significant

-chronic fatigue patients had less difficulty than patients in the comparison group, and the difference was statistically significant

-chronic fatigue patients did not differ significantly from patients in the comparison group

Abbreviations

FUNC-Functional Impairment

FATG-Fatigue

MLSE-Malaise

THRO-Sore throat

LYMP-Lymph nodes

MUSC-Muscle pain

JOIN-Joint pain

HEAD-Headache

SLEP-Sleep

MEMR-Memory

CONC-Concentration

FEVR-Fever/chills

WEAK-Muscle weakness

DEPR-Depression

We show several examples of prevalence estimates of chronic fatigue syndrome that have been published since 1990 in Table 2. Studies that addressed prevalence in special populations, such as nurses, Gulf War veterans, and patients with infections or known fatigue are not reviewed.^73-77 Studies in which physicians relied on their memories to estimate numbers of patients with CFS that they were caring for also were not reviewed.⁷⁸

The prevalence of CFS in community populations ranged from approximately 0.1 to 0.7%. Of note, lower prevalence bounds of less than 0.01% were reported in a large surveillance study with poor response rates and in a retrospective analysis of a community survey that did not ascertain all possible minor symptoms associated with current definitions of chronic fatigue syndrome.^79,80 The reported prevalence in primary care populations ranged from approximately 0.04% to 2.6%. Use of different case definitions and differences between primary care practices in different countries probably accounted for the varied estimates. In general, use of the 1988 CDC criteria, which are the most stringent, appeared associated with the lowest prevalence estimates.

Natural History

Table 3. Multivariate psychometric studies on CFS clinical (more...)

Table 3. Multivariate psychometric studies on CFS clinical manifestations

Study/Participants	Design/Methods	Findings/Conclusion
Blakely54 1991 Case-Definition: CDC 1988 Source: General practice clinics (CFS and healthy controls), Pain clinic (chronic pain patients) Location: New Zealand Mean Age: 39 years % Female: 61 Mean Duration of Illness: 5.2 years (CFS)	Objective: Investigate the psychological characteristics of chronic fatigue syndrome. Design: Battery of psychometric instruments (General Health Questionnaire, Beck Depression Inventory, Minnesota Multiphasic Personality Inventory, Lazarus Ways of Coping Inventory) N Surveyed: 243 (58 CFS, 81 CP or chronic pain, 104 healthy controls) N Analyzed: 153 (42 CFS, 39 CP or chronic pain, 72 healthy controls)	Discriminant analysis correctly classified 65.4% of cases. There was a marked overlap of classification for CFS and CP, and some overlap between CP and controls. 88% of healthy controls were correctly classified; only 55% of CFS and 44% of chronic pain patients were correctly classified. 26% of CFS patients were misclassified as chronic pain; 27% of chronic pain was misclassified as CFS.
Vercoulen55 1994Case-Definition: Oxford 1991 Source: Self-referred patients Location: Netherlands % Female: 75 Median Duration of Illness: 5 years	Objective: Identify dimensions of CFS using principal component analysis and to provide a description of patients presenting with unexplained fatigue using these dimensions. Design: Survey questionnaire (present illness, specialist visits, treatment) and fatigue, complaints, psychological well-being, daily functioning, avoidance behavior, cognitions and attributions, social interactions, and sleep disturbances questionnaires). N Surveyed: 395 N Analyzed: 298	Principal component analysis results in 8 easily interpreted and named factors (psychological well-being, functional impairment in daily life, sleep disturbances, avoidance behavior, concentration problems, causal attributions related to the complaints, social functioning, and self-efficacy expectations) and one difficult-to-interpret factor and named “subjective experience.” Fatigue severity predicted by dimensions psychological well-being, functional impairment, and self-efficacy expectations using stepwise multiple regression analysis (R2=.51)
Hickie56 1995Case-Definition: Australian 1990 Source: CFS clinical registry Location: Australia Mean Age: 38 years % Female: 71 Mean Duration of Illness: 6.5 years	Objective: To determine whether patients diagnosed as having CFS constitute a clinically homogeneous class using multivariate statistical analyses to derive symptom patterns and potential patient subclasses. Design: Patient survey using a self-report symptom instrument (40 symptoms), sociodemographic data, Zung depression scale, course of illness, impact of illness on social and occupational functioning, illness behavior, and the General Health Questionnaire. N Surveyed: 770 N Analyzed: 565	Latent class analysis resulted in two subgroups with distinguishable clinical features: 1) 27% of patients having characteristics suggestive of somatoform disorders; and 2) 73% presenting a more limited combination of fatigue and neuropsychological symptoms, and only moderate disability but remained clinically heterogeneous. Significant differences between the two groups suggested that Group 1 patients were almost twice as likely to be male, had fewer years duration of illness and less likely to have abilities limited by fatigue, muscle/joint pain, poor concentration, depression and amotivation. Group 2 patients also had lower scores on psychological symptoms, CFS symptom history and severity, fewer visits to the doctor, less disability, less likely to sleep > 4 hours in the daytime, and less likely to be treated with antidepressants.
Fischler52 1997Case Definition: Oxford 1991 Source: Fatigue clinic (CFS); students, administrative, nursing and medical personnel (controls) Location: Belgium Mean Age: 36 years % Female: 72 Mean Duration of Illness: 6.4 years	Objective: Compare polysomnographic data between CFS patients and healthy controls; assess whether there is an association between primary sleep disorders and disability in CFS. Design: Psychiatric interviews with SCID-P; other instruments included Hamilton Rating Scale for Depression, Sickness Impact Profile, multiple sleep variables with polysomnography. Discriminant analysis used with sleep variables. N Analyzed: 49 CFS, 20 healthy controls	Discriminant analysis showed a high level of correct classification of CFS patients and healthy controls. Sleep-onset latency and the number of stage shifts/hour contributed significantly to the discriminant function. The mean REM latency and the percentage of patients with a shortened REM latency were similar in CFS and healthy controls.
Vercoulen57 1998Case-Definition: Oxford 1991 Source: not reported Location: Netherlands Mean Age: 36 years % Female: 71 Median Duration of Illness: 5 years	Objective: Clarify ambiguities in previous studies on demographic and social factors and fatigue. Design: Structural equation modeling (causal modeling) using battery of instruments including Checklist Individual Strength, Beck Depression Inventory, Somatization Subscale of Symptom Checklist, Home Activities Subscale and Mobility Subscale of Sickness Impact Profile, Physical Activities Rating Scale, Causal Attributions List, Pain Cognition List. N Analyzed: 51 CFS, 50 multiple sclerosis (MS)	“Attributing complaints to a somatic cause produced low levels of physical activity, which in turn had a causal effect on fatigue severity. Depression had to be deleted from the model. Sense of control over symptoms and focusing on bodily symptoms each had a direct causal effect on fatigue. The model showed an excellent fit for CFS patients, but was rejected for MS patients. Therefore, a new model for MS patients had to be developed in which sense of control had a causal effect on fatigue. In the MS model, no causal relationship was found between the physical state as measured by the Expanded Disability Status Score and fatigue or functional impairment. The present study shows that cognitive and behavioral factors are involved in the persistence of fatigue. Treatment should be directed at these factors. The processes involved in the subjective experience of fatigue in CFS were different from the processes related to fatigue in MS.”
Jason58 1999 unpublishedJason59 1999Case-Definition: CDC 1994 Source: Random community sample Location: United States Mean Age: 44 (chronic fatigue) % Female: 67 (chronic fatigue) Mean Duration of Illness: 4.0 years	Objective: Address important issues involving the actual factor structure of symptoms of chronic fatigue within a community sample of individuals, unbiased by help-seeking behavior. Design: Community-based study of a random sample of 28,673 residential/working telephone numbers, using a CFS Screening Questionnaire containing a Fatigue Scale and other scales addressing symptoms of CFS and fatigue duration of fatigue. N Surveyed: 18,675 completing interview N Analyzed: 780 of 18,676 with chronic fatigue (408 CFS-like: chronic fatigue and 4+ CDC 94 symptoms; 304 ICF-like: idiopathic chronic fatigue; 68 with exclusionary medical or psychiatric conditions or CF-explained-like)	Principal component analysis resulted in a four-factor solution explaining 38% of total variance: lack of energy, physical exertion, cognitive functioning, and fatigue and rest. Compared to ICF-like, CFS-like had significantly more of the following symptoms: sore throat, painful glands, muscle aches or pain, feel worse for >24 hours after exercising, new headaches, joints pain, not rested after a night of sleep, concentration/memory interfere with work/study; and more difficulty with lack of energy, physical exertion, cognitive functioning, greater fatigue following rest or sleep. Compared to CF-explained, CFS-like had significantly more of the following symptoms: sore throat, painful glands, muscle aches or pain, not rested after a night of sleep.
Jason60 1999 unpublishedJason61 1999Case-Definition: CDC 1994 Source: Random community sample Location: United States Mean Age: ≈ 35 Female: 74 Mean Duration of Illness: not reported	Objective: Using cluster analysis, provide further empirical support for the observation of distinct aspects of chronic fatigue, and clarify differences between different subgroups of individuals with chronic fatigue in a smaller sample of CFS-like individuals undergoing full psychiatric and medical evaluation; and identify unique symptom and severity patterns. Design: Community-based study, n = 28,673. N Surveyed: 18,675 completing interview, 780 with chronic fatigue N Analyzed: 166 of 408 CFS-like individuals: chronic fatigue and 4+ CDC 94 symptoms	166 individuals classified by independent physician review panel into four groups: chronic fatigue syndrome (CFS), idiopathic chronic fatigue (ICF), chronic fatigue explained by medical condition (CFM), or chronic fatigue explained by psychiatric condition (CFP). CFS group had more severe post-exertional fatigue than ICF; CFM had more severe post-exertional fatigue than ICF and CFP. Cluster analysis resulted in three clusters: 1) relatively low post-exertional fatigue (n=1); 2) most severe post-exertional fatigue and most improvement in fatigue following rest (n=47); and 3) high post-exertional fatigue and fatigue not alleviated by rest (n=117)
Taylor62 2000 unpublishedCase-Definition: CDC 1994 Source: Random community sample Location: United States Mean Age: not reported % Female: not reported Mean Duration of Illness: not reported	Objective: Compare rates of psychiatric diagnosis among individuals with chronic fatigue and a healthy control group in a randomly selected, diverse community-based urban sample; and explore whether various sociodemographic indicators predicted higher or lower rates of Axis I psychiatric disorders. Design: Community-based study, n = 28,673 N Surveyed: 18, 675 completing interview; and completing the SCID (Structural Clinical Interview for DSM-IV): 408 identified as having chronic fatigue profiles, 199 controls N Analyzed: 301 (227 of 408 CF and 74 of 199 controls)	In a forward stepwise logistic regression analysis with current psychiatric diagnosis as the dependent variable, significant predictors were chronic fatigue status and socioeconomic status. In a forward stepwise logistic regression analysis with lifetime psychiatric diagnosis as the dependent variable, significant predictors were chronic fatigue status and work status
Taylor64 2000 unpublishedCase-Definition: CDC 1994 Source: Random community sample Location: United States Mean Age: not reported % Female: not reported Mean Duration of Illness: not reported	Objective: Evaluate the diagnostic validity of somatic symptoms associated with CFS, Fibromyalgia, Somatic Depression, Somatic Anxiety, and Irritable bowel syndrome. Design: Community-based study, n = 28,673 N Surveyed: 18, 675 completing interview; and completing the SCID (Structural Clinical Interview for DSM-IV): 408 identified as having chronic fatigue profiles, 199 controls N Analyzed: 213 individuals who underwent medical and psychological work-up. Of the 166 individuals reporting a CFS-like condition and 47 controls, 26 were diagnosed as CFS only, 18 were diagnosed with FMS only, and 6 were diagnosed with CFS & FMS.	Confirmatory factor analysis on 21 symptoms yielded a 5-factor solution that specifically distinguished the 5 functional somatic disorders from one another. The 5 distinct functional somatic disorder model had a reasonably good fit to the data. In contrast, a confirmatory analysis that tested the proposition that all 5 disorders represent a single underlying functional somatic distress syndrome did not fit the data well. Logistic regression predicting actual FMS diagnoses from the five-syndrome factor scores showed that only the FMS symptom factor score was significantly associated with actual FMS diagnosis. Logistic regression predicting actual CFS diagnoses from the five-syndrome factor scores showed that only the CFS symptom factor was significantly associated with actual CFS diagnosis.
Friedberg63 2000Case Definition: CDC 1994 Source: Advertisements in CFS journals/newsletters and physician referrals (CFS), and spouses/partners/significant-others of CFS participants (healthy controls) Mean Age: 49 years % Female: 67 Median Duration of Illness: 16 years (long duration CFS) and 3 years (short duration CFS)	Objective: Evaluate symptom patterns in patients with CFS who were ill for 10 or more years using Symptom Checklist, CES-D, Brief Symptoms Inventory, and Illness Management Questionnaire. Design: Simple group comparison, matched controls N Analyzed: 286 CFS and 179 healthy controls	Principal component analysis of illness-related symptoms in long-duration CFS yielded three conceptually distinct dimensions with the cognitive problems factor accounting for a substantial portion (30.5%) of the variance; of the remaining factors, flu-like symptoms accounted for 6.4% and neurologic symptoms 4.1% of the variance. Long-duration CFS patients had significantly greater percentage on the following symptoms compared to short-duration CFS: work block, bumping into things, putting the wrong word in, trouble with directions, and burning sensations. A number of viral and immune-related illnesses were reported at significantly higher frequencies in CFS patients than in healthy controls including chicken-pox/shingles, chronic sinus infections, infectious mononucleosis, herpes I/II, and hepatitis.

We found multiple studies relevant to the natural history of chronic fatigue syndrome.^{83,84,87-90,90-117} Of these, seven met all inclusion criteria (Table 3). Others were excluded for the following reasons: retrospective or cross-sectional data,^94,97,98 abstract only,¹¹⁸ mixed CFS with other patient data,^89,92,93 lack of case definition developed specifically for CFS,^102,105 short followup,¹¹⁷ and small sample sizes.^{83,84,87,88,95,96,99-101,104,112-116}

The seven prospective studies with more than 30 participants that had a followup duration of at least one year are outlined in Table 3. Two of the studies involved patients recruited from the same referral clinic in Seattle; some of the same patients may have participated in both studies.^103,108 In all, approximately 1,000 people with CFS, almost all of whom have been recruited from referral settings, were followed in these studies. More than two-thirds of the participants have been middle-aged women. Rates of self-reported global or overall improvement at 12 to 18 months have ranged from 11% to 64%.^{90,103,107,110} Rates of self-reported worsening of symptoms at 12 to 18 months in two studies were 15% and 20%.^90,110 One study from the CDC surveillance system estimated that the cumulative probability of recovery at five years was 31%.¹⁰⁶ Six months after the first reported recovery, 28% reported return of symptoms of fatigue while 15% were not classified as recovered or relapsed because of incomplete data. ¹⁰⁶

Question 4. Do Controlled Studies Show that Particular Therapies Improve Clinical Symptoms of CFS, When Compared to Placebo, No Therapy, or Each Other?>

Overview of the Evidence

There were 38 controlled trials examining various therapies for the treatment of chronic fatigue syndrome among adults. One trial involving adolescents and treatment with immunoglobulin was found but for the purpose of this report, not reviewed further.¹²⁰ One trial examining the effect of exposure to polio vaccine was found, but the vaccine was not administered as a therapeutic agent and therefore, excluded from further review.¹²¹ One trial comparing an antiviral to amino acid therapy was found, but the only data reported were pre- and posttest results for each separate arm and therefore, not reviewed further.¹²² Four trials were found but excluded from this report because the authors used a case-definition other than one of the four allowed in the selection criteria, or they did not report the case-definition that was used. Excluded trials examined a multidisciplinary therapeutic approach, essential fatty acid therapy, vitamin/mineral therapy, or sulbutiamine.^123-126

Table 6. Controlled trials: country of origin by publication (more...)

Table 6. Controlled trials: country of origin by publication year and case definition

Year	CDC 1988	Australia 1990	Oxford 1991	CDC 1994
1988	U.S.127
1989	U.S.128
1990	Australia129
	U.S.130
1991	U.K.131	U.K.131
1992
1993	U.K.133	Australia132
1994	U.S.134,135
1995
1996	U.S.136,138,140		Iceland141	U.S.136,142,143
			Netherlands139
			U.K.137,144
1997	U.S.148	Australia147	U.K.145,146	U.K.145
1998	U.K.154		U.K.151	Belgium153
	U.S.149,150,152			Canada155
				Sweden156
				U.S.149,150,152
1999			U.K.157,158	U.K.158
				U.S.159
2000			U.K.162	Australia161
				Netherlands163
				U.K.164
				U.S.160

Note: References appearing more than once reflects studies where authors used more than one case definition.

Table 7. Controlled trials: types of therapies studied (more...)

Table 7. Controlled trials: types of therapies studied

# of Trials	Intervention Category	Treatment		Comparison Group
1127	Immunological	Antiviral		Placebo
1138		Antihistamine
1156		Vaccine
1135		Immunomodulator/Antiviral
3129,130,147		Immunomodulator	Immunoglobulin
1143			Transfer Factor
1140			Interferon
1133				No Treatment
1132	Immunological and Behavioral	Immunomodulator (Transfer Factor) and Cognitive Behavioral Therapy		Placebo and/or Attention (Attend Clinic) Control
2149,160	Pharmacological (other than Immunological)	Steroid	Mineralcorticoid	Placebo
2150,158			Glucocorticoid
1141		Anticholinergic
1153		Hormone
1159		NADH
4136,139,152,161		Antidepressant
1151	Pharmacological and Behavioral	Antidepressant and Graded Exercise		Placebo and/or Non-Specific Advice Control
5134,137,145,163,164	Behavioral	Cognitive Behavioral Therapy		Medical Care; Relaxation; Counseling; Natural Course; or No Treatment
1146		Graded Exercise Therapy		Flexibility and Relaxation
1162		Educational Intervention		Standard Medical Care
1144	Complementary and Alternative Therapy	Homeopathy		Placebo
1148		Massage Therapy		Sham TENS
1154		Osteopathy		Normal Care
1131	Nutritional Supplements	Magnesium		Placebo
1157		Essential Fatty Acids
1128		Liver Extract
1142	Other Interventions	Social Support		Waiting List
1155		Comprehensive Care		NoTreatment (Assessed Only)

Table 8. Evidence tables of randomized controlled trials (more...)

Table 8. Evidence tables of randomized controlled trials

Author/Year Study Methods	Intervention Characteristics	Patient Characteristics	Outcomes Reported (How Measured)*
Straus127 1988Randomized: yes Double-blinded: yes Crossover Trial: yes # Enrolled: 27 % Dropout: 11% # Analyzed: 24	Intervention: Antiviral Treatment: Acyclovir (n=27) Comparison: Placebo (n=27) Protocol: Intravenous placebo or acyclovir (500 mg/m2) every 8 hours for 7 days of hospitalization, then discharged to take placebo or 800 mg acyclovir tablets 4x daily for 30 days; 6-week washout period. Duration of Therapy: 5 weeks	Case Definition: CDC 1988 Psychiatric Exclusions: not stated Location: United States Patient Source: NIH referrals Mean Illness Duration: 6.8 years Mean Age: 34.1 years % Female: 70% Baseline to Followup: 11 weeks	General: Wellness Somatic: Fatigue, Vigor (POMS) Functional: Rest (hours per day) Mood: Anxiety, Depression, Anger (POMS) Cognitive: Confusion (POMS)
Steinberg138 1996Randomized: yes Double-blinded: yes Crossover Trial: no # Enrolled: 30 % Dropout: 7% # Analyzed: 28	Intervention: Antihistamine Treatment: Terfenadine (n=15) Comparison: Placebo (n=15) Protocol: Terfenadine or placebo, 60 mg twice daily Duration of Therapy: 2 months	Case Definition: CDC 1988 Psychiatric Exclusions: not stated Location: United States Patient Source: Patient registry Illness Duration: not reported Mean Age: 36.2 years % Female: 77% Baseline to Followup: 2 months	General: Health perceptions (MOS) Somatic: Symptoms: fatigue, post-exertional fatigue, muscle weakness, myalgia, sleep disturbance, headache, arthralgia (percent) Functional: Physical, Social functioning (MOS) Mood: Mental Health (MOS)
Strayer135 1994Randomized: yes Double-blinded: yes Crossover Trial: no # Enrolled: 92 % Dropout: 9% # Analyzed: 84	Intervention: Immunomodulator/Antiviral Treatment: Ampligen (n=45) Comparison: Placebo (n=47) Protocol: Ampligen intravenous infusion, four 200-mg doses initially, followed by 400 mg 2x-weekly, or saline placebo of equal volume Duration of Therapy: 6 months	Case Definition: CDC 1988 Additional entry criteria: Severely debilitated (KPS score 20 to 60) Psychiatric Exclusions: yes Location: United States Patient Source: not stated Mean Illness Duration: 5.2 years Mean Age: 40.7 years % Female: 75% Baseline to Followup: 6 months	Functional: Physical performance (KPS), Activity (ADL) Mood: Anxiety, Depression (SCL) Cognitive: Cognitive deficit (SCL)
Lloyd129 1990 Randomized: yes Double-blinded: yes Crossover Trial: no # Enrolled: 49 % Dropout: 4% # Analyzed: 49	Intervention: Immunomodulator Treatment: IgG (n=23) Comparison: Placebo (n=26) Protocol: 3 intravenous influsions of immunoglobulin or placebo solution (maltose) at a dose of 2 g/kg/month Duration of Therapy: 3 months	Case Definition: CDC 1988 Psychiatric Exclusions: not stated Location: Australia Patient Source: not stated Median Illness Duration: 3.9 years Mean Age: 36.0 years % Female: 49% Baseline to Followup: 6 months	General: Improvement (reduction in symptoms, improvement in functional capacity) Quality of Life (QAL) Functional: Change in work, leisure, social activities Mood: Depression (HAMD, Zung)
Peterson130 1990Randomized: yes Double-blinded: yes Crossover Trial: no # Enrolled: 30 % Dropout: 7% # Analyzed: 28	Intervention: Immunomodulator Treatment: IgG (n=15) Comparison: Placebo (n=15) Protocol: 6 intravenous infusions of immunoglobulin or placebo solution (albumin) at a dose of 1 g/kg/month Duration of Therapy: 5 months	Case Definition: CDC 1988 Psychiatric Exclusions: not stated Location: United States Patient Source: Research program Mean Illness Duration: 3.8 years Mean Age: 40.8 years % Female: 73% Baseline to Followup: 5 months	General: Health perception (MOS) Somatic: Symptom severity (fatigue, post-exertional fatigue, muscle weakness, myalgia, arthralgia, sleep disturbance, headache) Physical and social functioning Functional: (MOS) Mental health (MOS) Mood:
Vollmer-Conna147 1997Randomized: yes Double-blinded: yes Crossover Trial: no # Enrolled: 99 % Dropout: 4% # Analyzed: 99	Intervention: Immunomodulator Treatment 1: 0.5 g/kg IgG (n=22) Treatment 2: 1.0 g/kg IgG (n=28) Treatment 3: 2.0 g/kg IgG (n=23) Comparison: Placebo (n=26) Protocol: 3 intravenous infusions administered at monthly intervals or placebo (albumin/maltose) Duration of Therapy: 3 months	Case Definition: Australia 1990 Psychiatric Exclusions: no Location: Australia Patient Source: not stated Mean Illness Duration: 6.3 years Mean Age: 39.8 years % Female: 76% Baseline to Followup: 6 months	General: Quality of Life (QAL) Somatic: Energy, Fatigue (POMS) Functional: Daily activity (KPS; hours per day) Mood: Depression (POMS) Cognitive: Confusion (POMS)
De Vinci143 1996 Randomized: yes Double-blinded: yes Crossover Trial: no # Enrolled: 20 % Dropout: 10% # Analyzed: 20	Intervention: Transfer Factor Treatment: Dialyzable extract from immune lymphocytes (n=14) Comparison: Placebo (n=6) Protocol: Capsules of ‘specific’ TF (known activity against Epstein-Barr virus, human herpes virus 6 and cytomegalovirus), 90-970 days, or placebo (unspecific TF or lactose placebo) 140-360 days Duration of Therapy: 90-970 days (mean 328 days)	Case Definition: CDC 1994 Psychiatric Exclusions: not stated Location: United States Patient Source: not reported Illness Duration: not reported Mean Age: 35 % Female: 60% Baseline to Followup: End of treatment	General: Clinical Improvement (diminished sore throat, muscular and/or joints pain, headaches, increased ability of mental concentration)
Lloyd132 1993 Randomized: yes Double-blinded: yes Crossover Trial: no # Enrolled: 90 % Dropout: 2% # Analyzed: 88	Intervention: Transfer Factor (DLE - dialyzable leukocyte extract) and/or Cognitive Behavioral Therapy (CBT) Treatment 1: DLE & CBT (n=20) Treatment 2: DLE & “attend clinic” (n=26) Treatment 3: CBT and saline placebo (n=21) Comparison: Saline placebo & “attend clinic” (n=23) Protocol: DLE (5x108 leukocytes/dose), 8 biweekly intramuscular injections or normal saline placebo; CBT (delivered by 3 psychiatrists who focused on behavioral and psychological issues, difficulties that warranted attitude shifts, attitude changes, and concepts of being active including a home-based graded exercise schedule), 6 biweekly sessions, or “attend clinic” placebo Duration of Therapy: 4 months	Case Definition: Australia 1990 Psychiatric Exclusions: no Location: Australia Patient Source: not stated Median Illness Duration: 5.5 years Mean Age: 39.6 years % Female: 76% Baseline to Followup: 7 months	General: Quality of Life (QAL) Somatic: Fatigue, Vigor (POMS) Functional: Physical capacity (diaries), Daily activity (KPS), Non-sedentary hours Mood: Anger, Anxiety, Depression (POMS) Cognitive: Confusion (POMS)
See140 1996Randomized: yes Double-blinded: yes Crossover Trial: yes # Enrolled: 30 % Dropout: 13% # Analyzed: 26	Intervention: Immunomodulator Treatment: Interferon-(α2a (n=15) Comparison: Placebo (n=15) Protocol: 3 million units of interferon-(α2b administered subcutaneously thrice weekly; comparison group treated after 3 month followup; washout period not stated, or placebo (NaCl) Duration of Therapy: 3 months	Case Definition: CDC 1988 Psychiatric Exclusions: no Location: United States Patient Source: Referrals by local internists and university faculty Mean Illness Duration: 4.6 years Mean Age: 37.2 years % Female: 80% Baseline to Followup: 4 months	General: Quality of Life
Brook133 1993Randomized: yes Double-blinded: no Crossover Trial: no # Enrolled: 20 % Dropout: 10% # Analyzed: 20	Intervention: Immunomodulator Treatment: Interferon-α2b (n=11) Comparison: No treatment (n=9) Protocol: 3 million units of interferon-α2b administered subcutaneously thrice weekly; comparison group treated after 3 month followup Duration of Therapy: 3 months	Case Definition: CDC 1988 Psychiatric Exclusions: not stated Location: United Kingdom Patient Source: not reported Illness Duration: 1-11 years Mean Age: not reported % Female: 70% Baseline to Followup: 6 months	Functional: Activity (ECOG)
Peterson149 1998Randomized: yes Double-blinded: yes Crossover Trial: yes # Enrolled: 25 % Dropout: 20% # Analyzed: 20	Intervention: Mineralcorticoid Treatment: Fludrocortisone (n=22) Gp 2: Placebo (n=24) Protocol: Fludrocortisone acetate tablet (0.1 initially; dosed doubled to 0.2 mg if patient reported no improvement in fatigue after first two weeks of treatment) or lactose placebo; 6-week washout period. Duration of Therapy: 6 weeks	Case Definition: CDC 1988/CDC 1994 Psychiatric Exclusions: no Location: United States Patient Source: Patient registries, regional CFS research program and local CFS clinic Mean Illness Duration: 7.0 years Mean Age: 39.7 years % Female: 76% Baseline to Followup: 6 weeks	General: General well-being (MOS) Somatic: Symptom severity: fatigue, unrefreshing sleep, sore throat, joint pains, muscle pains, headaches, painful lymph nodes, lightheadedness (VAS) Functional status: physical, Functional: social, role limitations (MOS) Emotional well-being (MOS), Depression (VAS), Positive Mood: affect (PANAS) Concentration, Forgetfulness (VAS); Reaction time (HICK) Cognitive:
Rowe160 2000Randomized: yes Double-blinded: yes Crossover Trial: no # Enrolled: 100 % Dropout: 17% # Analyzed: 83	Intervention: Mineralcorticoid Treatment: Fludrocortisone (n=50) Comparison: Placebo (n=50) Protocol: Fludrocortisone (one capsule of 0.025 mg capsule/day for 1 week, then increased to 2 capsules per day for 1 week, then 4 per day for remaining 7 weeks) or methylcellulose placebo. Duration of Therapy: 9 weeks (started 2 weeks after table-tilt test screen)	Case definition: CDC 1994; patients with neurally-mediated hypotension Psychiatric Exclusions: yes Location: United States Patient Source: CFS patient registries; ads with patient advocacy programs, local newspapers and internet postings Mean Illness Duration: 6.4 years Mean Age: 36.8 years % Female: 66% Baseline to Followup: 11 weeks	General: Wellness Somatic: Fatigue, Vigor (POMS) Functional: Physical function (MOS), Activity (DASI) Mood: Depression (BDI), Mental health (MOS) Cognitive: Mental fatigue (WMFI)
McKenzie150,167 1998, 2000Randomized: yes Double-blinded: yes Crossover Trial: no # Enrolled: 70 % Dropout: 10% # Analyzed: 65-68	Intervention: Glucocorticoid Treatment: Hydrocortisone (n=35) Comparison: Placebo (n=35) Protocol: Oral hydrocortisone, 13 mg/m2 of body surface area every morning and 3 mg/m2 every afternoon, or placebo, for 12 weeks Duration of Therapy: 3 months	Case definition: CDC 1988/94 Psychiatric Exclusions: yes Location: United States Patient Source: not reported Mean Illness Duration: 4.4 years Mean Age: 37.5 years % Female: 80% Baseline to Followup: 3 months	General: Wellness Somatic: Symptom severity (SCL), Fatigue, Vigor (POMS) Functional: Sickness impact (SIP), Activity Depression (BDI, HAMD), Anger, Mood: Anxiety (POMS) Confusion (POMS) Cognitive:
Cleare158 1999Randomized: yes Double-blinded: yes Crossover Trial: yes # Enrolled: 35 % Dropout: 9% # Analyzed: 32	Intervention: Glucocorticoid Treatment: Hydrocortisone (n=32) Comparison: Placebo (n=32) Protocol: Hydrocortisone (5 mg first 16 patients, 10 mg for remainder) or placebo first for 28 days before crossover to other treatment; washout not reported Duration of Therapy: 1 month	Case Definition: CDC 1994/Oxford 1991 Psychiatric Exclusions: yes Location: United Kingdom Patient Source: CFS clinics Meant Illness Duration: 3.0 years Mean Age: 35.3 years % Female: 57% Baseline to Followup: 1 month	General: Improvement Somatic: Fatigue (CFQ) Functional: Physical function, Role limitation (MOS), Disability (WSAS) Psychological symptoms (GHQ) Mood:
Snorrason141 1996 Randomized: yes Double-blinded: yes Crossover Trial: yes # Enrolled: 49 % Dropout: 20% # Analyzed: 30-40	Intervention: Anticholinergic Treatment: Galanthamine (n=25) Comparison: Placebo (n=24) Protocol: Galanthamine hydrobromide (10 mg t.i.d., reached by escalating dosage) or matched treatment with placebo tablets; 70% of patients followed this schedule and remaining 30% dosage reduced due to adverse effects. Originally designed as a parallel trial, changed to optional crossover after 2 weeks; washout not reported. Duration of Therapy: 2-8 weeks	Case Definition: Oxford 1991 Psychiatric Exclusions: yes Location: Iceland Patient Source: General and rheumatological outpatient clinics Mean Illness Duration: 12.8 years Mean Age: 43.9 years % Female: 86% Baseline to Followup: 1-8 weeks	Somatic: Fatigue, Myalgia, Sleep, Dizziness (VAS) Functional: Work capacity (VAS) Mood: Anxiety (STAIS, DSM), Depression, Emotional instability (DSM) Cognitive: Concentration (WMS), Memory (WMS, VAS)
Moorkens153 1998 Randomized: yes Double-Blinded: yes Crossover Trial: no # Enrolled: 20 % Dropout: 15% # Analyzed: 17	Intervention: Hormone Treatment: Growth hormone (n=10) Comparison: Placebo (n=10) Protocol: Growth hormone therapy, 6.7 μg/kg/day, or placebo for 12 weeks, then all patients enrolled in open trial with GH Duration of Therapy: 3 months RCT, 9 months open trial	Case Definition: CDC 1994 Psychiatric Exclusions: no Location: Belgium Patient Source: CFS clinic Illness Duration: not reported Mean Age: not reported % Female: 65% Baseline to Followup: 12 months	General: Quality of Life (NHP, QOL)
Forsyth159 1999 Randomized: yes Double-blinded: yes Crossover Trial: yes # Enrolled: 35 % Dropout: 26% # Analyzed: 26	Intervention: Nicotinamide Adenine Dinucleotide (NADH) Treatment: NADH (n=26) Comparison: Placebo (n=26) Protocol: NADH, two 5mg tablets daily, for a 4-week period; 4-week washout period. Duration of Therapy: 4 weeks	Case Definition: CDC 1994 Psychiatric Exclusions: yes Location: United States Patient Source: Referrals and recruited from medical center Mean Illness Duration: 7.2 years Mean Age: 39.6 years % Female: 65% Baseline to Followup: 4 weeks	General: Symptom Improvement
Natelson136 1996Randomized: yes Double-blinded: yes Crossover Trial: yes # Enrolled: 24 % Dropout: 25% # Analyzed: 18	Intervention: Antidepressant Treatment: Phenelzine (n=15) Comparison: Placebo (n=9) Protocol: First 2 weeks placebo phase-in all patients, 1 placebo pill per day; then phenelzine or placebo during 2nd 2 weeks with patients in active treatment group receiving one 15-mg tablet phenelzine per day alternating with placebo; during final two weeks, patients in active treatment group received one 15-mg tablet per day; placebo group took one placebo pill per day. Duration of Therapy: 4 weeks	Case Definition: CDC 1988/94 Psychiatric Exclusions: yes Location: United States Patient Source: CFS center Illness Duration: not reported Mean (SD) Age: 34.6 yrs (2.8) % Female: 83% Baseline to Followup: 6 weeks (includes 2 week placebo phase-in)	Somatic: Symptom severity (Checklist), Fatigue (POMS, FSS), Vigor (POMS) Functional status (FSQ), Activity Functional: (ADL), Disability (ISS) Depression (POMS, CESD) Confusion (POMS) Mood:   Cognitive:
Hickie161 2000Randomized: yes Double-Blinded: yes Crossover Trial: no # Enrolled: 90 % Dropout: 14% # Analyzed: 90	Intervention: Antidepressant Treatment: Moclobemide (n=47) Comparison: Placebo (n=43) Protocol: Moclobemide or placebo at one 150-mg tablet 2x daily for 1 week, then increased to 450 mg/day, then increased to 600 mg/day if tolerated Duration of Therapy: 6 weeks	Case Definition: CDC 1994 Psychiatric Exclusions: yes Location: Australia Patient Source: Infectious disease and immunology outpatient clinics Mean Illness Duration: 1.7 years Mean Age: 43.5 years % Female: 54% Baseline to Followup: 6 weeks	General: Improvement Somatic: Fatigue, Vigor (POMS) Functional: Disability (KPS) Mood: Depression (POMS)
Vercoulen139 1996 Randomized: yes Double-blinded: yes Crossover Trial: no # Enrolled: 107 % Dropout: 10% # Analyzed: 96	Intervention: Antidepressant Treatment: Fluoxetine (n=54) Comparison: Placebo (n=53) Protocol: Fluoxetine or placebo 20 mg per day; groups sub-divided into depressed and non-depressed patients. Duration of Therapy: 2 months	Case Definition: Oxford 1991 Psychiatric Exclusions: yes Location: Netherlands Patient Source: Referrals to general outpatient clinic Mean Illness Duration: 5.5 years Mean Age: 38.9 years % Female: 76% Baseline to Followup: 4 months	General: Recovery Somatic: Fatigue (CIS, SIP) Mood: Depression (BDI)
Wearden151 1998Randomized: yes Double-blinded: yes Crossover Trial: no # Enrolled: 136 % Dropout: 29% # Analyzed: 136	Intervention: Antidepressant and Graded Exercise Therapy (GET) Treatment 1: Fluoxetine & GET (n=33) Treatment 2: GET & drug placebo (n=34) Treatment 3: Fluoxetine and exercise placebo (n=35) Comparison: Placebo only (n=34) Protocol: All patients attended hospital on 8 sessions for treatment by physiotherapist. GET included instructions to carry out preferred aerobic activity for 20 minutes, 3x per week; GET placebo patients not offered any specific advice but told to do what they felt capable and to rest when felt needed. Fluoxetine or placebo, 20 mg daily. Duration of Therapy: 6 months	Case Definition: Oxford 1991 Psychiatric Exclusions: yes Location: United Kingdom Patient Source: Referrals to outpatient clinic Median Illness Duration: 2.3 years Mean Age 38.7 years % Female: 71% Baseline to Followup: 3 and 6 months	General: Health perception (MOS) Somatic: Fatigue (CFQ), Pain (MOS) Functional: Physical function, Role or occupational function, Social function (MOS; Functional work capacity (BERG) Depression (HADS) Mood:
Kaslow128 1989Randomized: no Double-blinded: yes Crossover Trial: yes # Enrolled: 15 % Dropout: 7% # Analyzed: 14	Intervention: Supplement Treatment: LEFAC (n=15) Comparison: Placebo (n=15) Protocol: LEFAC (extract of bovine liver, 10 μg/mL, with folic acid, .4 mg/mL, and cyanocobalamin, 100 μg/mL) or placebo. Self-administration of 2-mL IM injections daily, 1-week supply; returned after one week for alternate therapy; then all patients entered in open trial for 2 additional weeks. Duration of Therapy: 1 week	Case Definition: CDC 1988 Psychiatric Exclusions: not stated Location: United States Patient Source: Community physician or self-referral Illness Duration: not reported Mean Age 40.4 years % Female: 78% Baseline to Followup: 1 week	Somatic: Energy, Overall symptoms Functional: Daily activities (FSQ) Mood: Mental health (FSQ)
Cox131 1991Randomized: yes Double-blinded: yes Crossover Trial: no # Enrolled: 32 % Dropout: 3% # Analyzed: 31	Intervention: Supplement Treatment: Magnesium (n=15) Comparison: Placebo (n=17) Protocol: Magnesium sulfate (1 g in 2 ml) or placebo (2 ml water), intramuscular, weekly Duration of Therapy: 6 weeks	Case Definition: CDC 1988/Australia 1990 Psychiatric Exclusions: no Location: United Kingdom Patient source: Complementary medicine research center and general practitioners Illness Duration: not reported Mean Age: 36.4 years % Female: 69% Baseline to Followup: 6 weeks	General: Health profile (NHP) Somatic: Energy, Pain, Sleep (NHP) Functional: Social isolation, Physical mobility (NHP) Emotional reactions (NHP) Mood:
Warren157 1999Randomized: yes Double-blinded: yes Crossover Trial: no # Enrolled: 50 % Dropout: 24% # Analyzed: 38	Intervention: Essential Fatty Acids Treatment: EFA (Efamol Marine) (n=24) Comparison: Placebo (n=26) Protocol: EFA (evening primrose oil and concentrated fish oil), two 500-mg capsules, 4x day or sunflower oil placebo (same number of capsules) Duration: 3 months	Case Definition: Oxford 1991 Psychiatric Exclusions: no Location: United Kingdom Patient Source: Referrals to regional infectious disease unit Mean Illness Duration: 4.0 years Mean Age: 37.1 years % Female: 58% Baseline to Followup: 3 months	General: Improvement (Self-report) Somatic: Symptom severity (PSC) Mood: Depression (BDI)
Sharpe137,168 1996, 1998Randomized: yes Double-blinded: no Crossover Trial: no # Enrolled: 60 % Dropout: 0% # Analyzed: 60	Intervention: Cognitive Behavioral Therapy Treatment: CBT and medical care (n=30) Comparison: Medical care only (n=30) Protocol: CBT - consisting of 16 one-hour individual treatment sessions delivered by three experienced cognitive therapists that focused on explanation of illness, strategies to increase gradual and consistent increases in activity, active problem solving for interpersonal and occupational difficulties, and avoidance of excessive perfectionism and self-criticisms; Medical care - general practitioner followup Duration of Therapy: 4 months	Case Definition: Oxford 1991 Psychiatric Exclusions: yes Location: United Kingdom Patient Source: Outpatient clinic referrals Mean Illness Duration: 2.6 years Mean Age: 36.0 years % Female: 68% Baseline to Followup: 5, 8, 12 months	General: Overall change Somatic: Fatigue (severity) Functional: Activity (KPS, Interference; # days in bed/week) Mood: Anxiety, Depression (HADS)
Deale145,169170 1997, 1998, 2000 Randomized: yes Double-blinded: no Crossover Trial: no # Enrolled: 60 % Dropout: 12% # Analyzed: 53	Intervention: Cognitive Behavioral Therapy Treatment: CBT (n=30) Comparison: Relaxation (n=30) Protocol: CBT consisting of 13 weekly sessions aimed at establishing a program of consistent planned activity and rest; a sleep routine; graded increases in activity and cognitive restructuring of unhelpful illness beliefs and assumptions. Relaxation included progressive muscle relaxation, visualization and rapid relaxation. Duration of Therapy: 4 to 6 months	Case Definition: Oxford 1991/CDC 1994 Psychiatric Exclusions: yes Location: United Kingdom Patient Source: Referrals to CFS clinic Mean Illness Duration: 4.0 years Mean Age: 34.5 years % Female: 68% Followup: 1-, 3-, and 6-month post-treatment; 5 years	General: Improvement (global) Somatic: Fatigue (severity, degree, improvement) Functional: Functional Impairment (MOS, WSAS, Long-term goals) Mood: Anxiety (GHQ), Depression (GHQ, BDI)
Ridsdale164 2000 Randomized: yes Double-Blinded: no Crossover Trial: no # Enrolled: 160 (45 with CFS) % Dropout: 19.4% (18% CFS) # Analyzed: 129 (37 CFS)	Intervention: Cognitive Behavioral Therapy Treatment: CBT (n=80, 25 CFS) Comparison: Counseling (n=80, 20 CFS) Protocol: CBT was delivered by qualified therapists focused on activity planning, homework, establishing a sleep routine and other cognitive interventions. Counseling was delivered by qualified counselors and focused on a psychodynamic model of client-centered therapy. 6 therapy sessions (content video-taped and validated by independent raters). Duration of Therapy: 3 months	Case-Definition: Other/CDC 1994 Psychiatric Exclusions: yes Location: United Kingdom Patient Source: General practice patients Mean Illness Duration: 3.2 years Mean Age: 39.4 years % Female: 73% Baseline to Followup: 6 months	General: General Health (MOS) with care Fatigue (CFQ) Somatic: Social Adjustment (MOS) Functional: Anxiety, Depression (HADS) Mood:
Prins163 2000 Randomized: yes Double-Blinded: no Crossover Trial: no # Enrolled: 278 % Dropout: 3% # Analyzed: 270	Intervention: Cognitive Behavioral Therapy Treatment 1: CBT (n=93) Treatment 2: Guided support (n=94) Comparison: Natural course (n=91) Protocol: CBT consisting of 16 sessions delivered by 13 behavioral therapists focused on self-control, restructuring fatigue-related cognitions, attaining base level of daily activity, gradual increases of physical activity, and returning to work or personal activities; Guided Support consisting of 11 meetings delivered by a social worker aimed at exchanging experiences and obtaining mutual understandings; Natural course was defined as no intervention. Duration of Therapy: 8 months	Case-Definition: CDC 1994 Psychiatric Exclusions: no Location: Netherlands Patient Source: Internal medicine department outpatient clinic Mean Illness Duration: 5.6 years Mean Age: 36.7 years % Female: 78% Baseline to Followup: 8, 14 months	General: Quality of Life (EQ), Improvement (self-rated) Somatic: Fatigue Severity (CIS) Functional: Functional Impairment (SIP), Functional Status (KPS), Hours working Mood: Psychological Well-Being (SCL)
Fulcher146 1997 Randomized: yes Double-blinded: no Crossover Trial: yes # Enrolled: 66 % Dropout: 29% # Analyzed: 66	Intervention: Graded Exercise Therapy (GET) Treatment: GET (n=33) Comparison: Flexibility/relaxation (n=33) Protocol: GET included 12 weeks of supervised treatment and next week's exercise prescription. Lab sessions supervised by exercise physiologist. Flexibility/relaxation included 12 weekly sessions and taught stretching routine and relaxation techniques; specifically told to avoid extra physical activities. Duration of Therapy: 12 weeks	Case Definition: Oxford 1991 Psychiatric Exclusions: yes Location: United Kingdom Patient Source: Clinic referrals Median Illness Duration: 2.7 years Mean Age: 37.2 years % Female: 74% Baseline to Followup: 3 and 12 months after supervised treatment	General: Overall change Somatic: Fatigue; Sleep (PSQI) Functional: Physical function (MOS) Mood: Anxiety, Depression (HADS)
Powell162 2000 Randomized: yes Double-Blinded: no Crossover Trial: no # Enrolled: 148 % Dropout: 14% # Analyzed: 148	Intervention: Educational Intervention Treatment 1: Maximum educational intervention (MAX) (n=38) Treatment 2: Telephone intervention (TEL) (n=39) Treatment 3: Minimal educational intervention (MIN) (n=37) Comparison: Standardized medical care (n=34) Protocol: One of three educational intervention groups about evidence on physical and psychological effects of physical deconditioning and circadian dysrhythmia, with the intention of encouraging a self-managed graded exercise program (2 individual treatment sessions, 2 telephone follow-up calls, comprehensive educational pack. Minimal intervention (MIN) included no further treatment; Telephone intervention (TEL) included 7 additional calls; Maximum intervention (MAX) included 7 additional face-to-face sessions. Comparison group included standardized medical care (advice and return to primary care). Duration of Therapy: 3-4 months	Case-Definition: Oxford 1991 Psychiatric Exclusions: yes Location: United Kingdom Patient Source: Patient referrals to chronic fatigue clinic and infectious disease outpatient clinic. Mean Illness Duration: 4.3 years Mean Age: 33.4 years % Female: 78% Baseline to Followup: 3 months, 6 months, 1 year	General: Global change Somatic: Fatigue (CFQ), Sleep Functional: Physical function (MOS) Mood: Anxiety, Depression (HADS)
Awdry144,166 1996Randomized: yes Double-blinded: yes Crossover Trial: no # Enrolled: 64 % Dropout: 5% # Analyzed: 61	Intervention: Alternative Therapy Treatment: Homoeopathy (n=32) Comparison: Placebo (n=32) Protocol: Single selected homoeopathic remedies (with emphasis on symptoms and person with the symptoms) or placebo (inert powder). Duration of Therapy: 12 months	Case Definition: Oxford 1991 Psychiatric Exclusions: no Location: United Kingdom Patient source: Volunteer patients Mean Illness Duration: 5.1 years Mean Age: 40.2 years % Female: 70% Baseline to Followup: 12 months	General: Improvement Somatic: Fatigue, Sleep Functional: Disability, Employment Mood: Mood disturbance
Kaslow128 1989Randomized: no Double-blinded: yes Crossover Trial: yes # Enrolled: 15 % Dropout: 7% # Analyzed: 14	Intervention: Supplement Treatment: LEFAC (n=15) Comparison: Placebo (n=15) Protocol: LEFAC (extract of bovine liver, 10 μg/mL, with folic acid, .4 mg/mL, and cyanocobalamin, 100 μg/mL) or placebo. Self-administration of 2-mL IM injections daily, 1-week supply; returned after one week for alternate therapy; then all patients entered in open trial for 2 additional weeks. Duration of Therapy: 1 week	Case Definition: CDC 1988 Psychiatric Exclusions: not stated Location: United States Patient Source: Community physician or self-referral Illness Duration: not reported Mean Age 40.4 years % Female: 78% Baseline to Followup: 1 week	Somatic: Energy, Overall symptoms Functional: Daily activities (FSQ) Mood: Mental health (FSQ)
Field148 1997 Randomized: yes Double-Blinded: no Crossover Trial: no # Enrolled: 20 % Dropout: 0% # Analyzed: 20	Intervention: Massage Therapy Treatment: Massage (n=10) Comparison: SHAM TENS (n=10) Protocol: Massage, twice each week by trained massage therapist consisting of gentle pressure to arms, torso, legs and head. SLAM TENS (attention and tactile stimulation control) also provided by massage therapist; TENS device used but never turned on (no electrical current) Duration of Therapy: 5 weeks	Case Definition: CDC 1988 Psychiatric Exclusions: not stated Location: United States Patient Source: Physician referral Illness Duration: not reported Mean Age: 47 years % Female: 80% Baseline to Followup: 5 weeks	Somatic: Pain (MPQ, PSQ), Sleep (PSQ), Fatigue, Somatic symptoms (PFRS) Mood: Emotional distress (PRFA), Depression (POMS, CESD), Anxiety (STAIS) Cognitive distress (PRFA) Cognitive:

*Abbreviations ADL: Activities of Daily Living BDI: Beck Depression Inventory BERG: Bosch ERG 551 Ergometer BSI: Brief Symptom Inventory CESD: Center for Epidemiology Studies Depression Scale CFQ: Chalder Fatigue Questionnaire CIS: Checklist Individual Strength CGI: Clinical Global Impressions COPE: Coping Strategies Scales CPRS: Comprehensive Psychopathological Rating Scale DAIS: Duke Activity Status Index DSM: DSM-III-R ECOG: Eastern Cooperative Oncology Group Activity Scale EQ: EuroQol Quality of Life visual analogue scale FRCS: Fatigue-Related Cognitions Scale FSQ: Functional Status Questionnaire FSRS: Fatigue Self-Rating Scale FSS: Fatigue Severity Scale GHQ: General Health Questionnaire HADS: Hospital Anxiety and Depression Scale HAMD: Hamilton Depression Rating Scale HICK: Hick Paradigm Reaction Time ISS: Illness Severity Scale KPS: Karnofsky Performance Scale MOS/SF36: Medical Outcomes Study Short Form MPQ: McGill Pain Questionnaire NHP: Nottingham Health Profile PANAS: Positive and Negative Affect Scale POMS: Profile of Mood States PFRS: Profile of Fatigue-Related Symptoms PSC: Physical Symptom Checklist PSQ: Pain and Sleep Questionnaire PSQI: Pittsburgh Sleep Quality Index QAL: Quality of Life visual analogue scale QOL: Quality of Life assessment in GH-deficient adults questionnaire SCL: Symptom Checklist SCL-90 SIP: Sickness Impact Profile STAIS: State and Trait Anxiety Index of Spielberger SX: Symptom Severity Checklist VAS: Symptom severity visual analogue scale WMFI: Wood Mental Fatigue Inventory WMS: Wechsler Memory Scale WSAS: Work and Social Adjustment Scale Zung: Zung Depression Scale

Table 9. Evidence tables of non-randomized controlled (more...)

Table 9. Evidence tables of non-randomized controlled trials

Author/Year Study Methods	Intervention Characteristics	Patient Characteristics	Outcomes Reported (How Measured)*
Andersson156 1998 Randomized: unclear Double-Blinded: yes Crossover Trial: no # Enrolled: 28 % Dropout: 14% # Analyzed: 24	Intervention: Vaccine Treatment: Staphylococcus toxoid vaccine (n = 14) Comparison: Placebo (n = 14) Protocol: Vaccine given at increasing doses containing .01, .05, .1, .2, .5 and 1.0 ml of fully potent vaccine, or sterile water placebo; 1 injection weekly. Duration of Therapy: 12 weeks	Case-Definition: CDC 1994 and fibromyalgia Psychiatric Exclusions: no Location: Sweden Patient Source: Referral from care centers, hospitals, psychiatric units Mean Illness Duration: 12.9 years Mean Age: 47.0 years % Female: 100% Baseline to Followup: 12 weeks	General: Improvement (CGI) Somatic: Severity of Illness (CGI), Pain (VAS), Sleep (CPRS) Mood: Depression (Zung), Psychiatric symptoms (CPRS) Concentration, Memory (CPRS) Cognitive:
Natelson 152 1998 Randomized: no Double-Blinded: unclear Crossover Trial: yes # Enrolled: 25 % Dropout: 24% # Analyzed: 19	Intervention: Antidepressant Treatment: Selegiline (n=25) Comparison: Placebo (n=25) Protocol: 2 weeks placebo phase-in, 1 placebo pill 2x daily; next 2 weeks, 1 5-mg tablet of selegiline replaced 1 of the placebo pills; final 2 weeks, 1 5-mg selegiline tablet 2x daily; washout period not stated. Duration of Therapy: 6 weeks	Case Definition: CDC 1988/94 Psychiatric Exclusions: yes Location: United States Patient Source: CFS Center Illness Duration: not reported Mean Age: not reported % Female: not reported Baseline to Followup: 12, 26 weeks	Somatic: Fatigue (FSS), Vigor (POMS, ISS), Symptom severity (SX) Disability (FSQ) Functional: Tension, Anxiety (POMS), Mood: Depression (POMS, CESD)
Friedberg134 1994Randomized: no Double-blinded: yes Crossover Trial: no # Enrolled: 64 % Dropout: 0% # Analyzed: 64	Intervention: Cognitive Behavioral Therapy Treatment 1: CBT (n=22 CFS) Treatment 2: CBT(n=22 depressed) Comparison: No treatment (n=22) Protocol: 6 weekly sessions focusing on coping; relaxation, tolerance, coping skills and cognitive restructuring. Emphasis on lifestyle changes compatible with activity limitations imposed by CFS. Duration of Therapy: 6-9 weeks	Case Definition: CDC 1988 Psychiatric Exclusions: not stated Location: United States Patient Source: Outpatients with CFS, neurology clinic, and local CFS support group; depressed patients of author Mean Illness Duration: 4.1 years Mean Age: 36.9 years % Female: 75 Baseline to Followup: 6-9 weeks	Somatic: Fatigue (FSS) Mood: Depression (CESD), Psychological distress (BSI) Fatigue-related thinking (FRCS) Cognitive:
Perrin154 1998 Randomized: no Double-Blinded: no Crossover Trial: no # Enrolled: 80 % Dropout: 28% # Analyzed: 58	Intervention: Osteopathic Therapy Treatment: Osteopathic therapy (n=40) Comparison: Normal care (n=40) Protocol: Twenty sessions (minimum) including multiple osteopathic treatments. Duration of Therapy: 12 months	Case Definition: CDC 1988 and London criteria for ME Psychiatric Exclusions: yes Location: United Kingdom Patient Source: Patient referrals; control volunteers Illness Duration: not reported Age Range: 18 to 55 years % Female: 68% Baseline to Followup: 12 months	Somatic: Symptom severity (BDI, NHP, PFRS, general health, back pain, anxiety, sleep, cognitive function)
Shlaes142 1996 Randomized: no Double-Blinded: no Crossover Trial: no # Enrolled: 12 % Dropout: 17% # Analyzed: 10	Intervention: Social Support Treatment: Buddy and Mentor (n=6) Comparison: Waiting list control (n=6) Protocol: Buddy 1 hour per week (emotional support, social companionship and instrumental support), Mentor 2 hours per month (information and emotional support regarding living with CFS). Duration of Therapy: 4 months	Case Definition: CDC 1994 Psychiatric Exclusions: not stated Location: United States Patient Source: CFS specialists, support groups, CFS Association Illness Duration: not reported Age Range: 36 to 57 years % Female: 75% Baseline to Followup: 4 months	Somatic: Physical fatigue (FSRS) Functional: Physical functioning (MOS), Coping (COPE) Mood: Depression (CESD), Psychological functioning (MOS), Psychological distress (BSI) Cognition Mental fatigue (FSRS)
Marlin155 1998 Randomized: no Double-Blinded: no Crossover Trial: no # Enrolled: 71 % Dropout: 69% # Analyzed: 22	Intervention: Comprehensive Multidisciplinary Intervention (CMI) Treatment: CMI (n=51) Comparison: Assessed only (n=20) Protocol: Bringing patient under optimal medical management, treating ongoing affective or anxiety disorder pharmacologically, and implementing a comprehensive CBT program by therapist in patient's own environment (individually tailored but included physical exercise, sleep management, activity management, regulation of stimulant intake and reduction in use of symptomatic medications, cognitive intervention designed to deal with patient's beliefs concerning nature of disorder, participation of patient's family, and efforts to establish specific vocational and avocational goals. Duration of Therapy: 3-8 months	Case-Definition: CDC 1994 Psychiatric Exclusions: not stated Location: Canada Patient Source: Multidisciplinary clinic Mean Illness Duration: 4.1 years Mean Age: 41.6 years % Female: 77% Mean Baseline to Followup: 32 months	Functional: Return to work

*Abbreviations ADL: Activities of Daily Living BDI: Beck Depression Inventory BERG: Bosch ERG 551 Ergometer BSI: Brief Symptom Inventory CESD: Center for Epidemiology Studies Depression Scale CFQ: Chalder Fatigue Questionnaire CIS: Checklist Individual Strength CGI: Clinical Global Impressions COPE: Coping Strategies Scales CPRS: Comprehensive Psychopathological Rating Scale DAIS: Duke Activity Status Index DSM: DSM-III-R ECOG: Eastern Cooperative Oncology Group Activity Scale EQ: EuroQol Quality of Life visual analogue scale FRCS: Fatigue-Related Cognitions Scale FSQ: Functional Status Questionnaire FSRS: Fatigue Self-Rating Scale FSS: Fatigue Severity Scale GHQ: General Health Questionnaire HADS: Hospital Anxiety and Depression Scale HAMD: Hamilton Depression Rating Scale HICK: Hick Paradigm Reaction Time ISS: Illness Severity Scale KPS: Karnofsky Performance Scale MOS/SF36: Medical Outcomes Study Short Form MPQ: McGill Pain Questionnaire NHP: Nottingham Health Profile PANAS: Positive and Negative Affect Scale POMS: Profile of Mood States PFRS: Profile of Fatigue-Related Symptoms PSC: Physical Symptom Checklist PSQ: Pain and Sleep Questionnaire PSQI: Pittsburgh Sleep Quality Index QAL: Quality of Life visual analogue scale QOL: Quality of Life assessment in GH-deficient adults questionnaire SCL: Symptom Checklist SCL-90 SIP: Sickness Impact Profile STAIS: State and Trait Anxiety Index of Spielberger SX: Symptom Severity Checklist VAS: Symptom severity visual analogue scale WMFI: Wood Mental Fatigue Inventory WMS: Wechsler Memory Scale WSAS: Work and Social Adjustment Scale Zung: Zung Depression Scale

Table 10. Immunologic and supplemental therapies (vs. (more...)

Table 10. Immunologic and supplemental therapies (vs. placebo): General health and somatic outcomes

Intervention Type	Case Definition	Study	Therapy	Quality				General			Somatic
				A	B	N	D%	I M P R	Q O L	W E L L	F A T G	V I G R	S L E P	S E V R	P A I N	M Y L G	H E A D	A R T H	P F A T	W E A K	D I Z Z
Immunologic	CDC 1988	Strauss127	Antiviral	R	D	27	11%
		Steinberg138	Antihistamine	R	D	30	7%
		Lloyd129	Immunomodulator	R	D	49	4%	↑
		Peterson130		R	D	30	7%
		See140		R	D	30	13%
		Strayer135	Immunomodulator/Antiviral	R	D	92	9%
	CDC 1994	De Vinci143	Transfer Factor	R	D	20	10%
		Andersson156	Vaccine	U	D	28	14%	↑
	Australia 1990	Vollmer-Conna147	Immunomodulator	R	D	99	4%
		Lloyd132	Transfer Factor	R	D	90	2%
Supplement	CDC 1988	Cox131	Magnesium	R	D	32	3%			↑		↑			↑
		Kaslow128	Liver Extract	R	D	15	7%
	Oxford 1991	Warren157	Essential Fatty Acid	R	D	50	24%
Symbols and Abbreviations:Results: ↑ Favors therapy group No difference between groups ↓ Favors placebo group Quality A: Assignment MethodR - Randomized; N - Not Randomized; U - Unclear B: BlindingD - Double-blinded; N - Not Blinded; U - Unclear N: # Enrolled     D%: % Dropout Outcome Variables:IMPR-Improvement QOL-Quality of Life WELL-Wellness FATG-Fatigue VIGR-Vigor/Energy SLEP-Sleep Disturbance SEVR-Symptom Severity PAIN-Pain MYLG-Myalgia HEAD-Headache ARTH-Arthralgia PFAT-Postexertional Fatigue WEAK-Muscle Weakness DIZZ-Dizziness/Lightheadedness

Table 11. Pharmacologic therapies (vs. placebo): General (more...)

Table 11. Pharmacologic therapies (vs. placebo): General health and somatic outcomes

Intervention Type	Case Definition	Study	Therapy	Quality				General			Somatic
				A	B	N	D%	I M P R	Q O L	W E L L	F A T G	V I G R	S L E P	S E V R	P A I N	M Y L G	H E A D	A R T H	P F A T	W E A K	D I Z Z
Pharmacologic	CDC 1988	Natelson136	Antidepressant	R	D	24	25%	↑
		Natelson152		N	U	25	24%	↑				↑
		McKenzie150	Glucocorticoid	R	D	70	10%
		Peterson149	Mineralcorticoid	R	D	25	20%
	CDC 1994	Hickie161	Antidepressant	R	D	90	14%					↑
		Cleare158	Glucocorticoid	R	D	35	9%				↑
		Rowe160	Mineralcorticoid	R	D	100	17%
		Forsyth159	NADH	R	D	35	26%	↑
	Oxford 1991	Vercoulen139	Antidepressant	R	D	107	10%
		Wearden151		R	D	136	29%
Symbols and Abbreviations:Results: ↑ Favors therapy group No difference between groups ↓ Favors placebo group Quality A: Assignment MethodR - Randomized; N - Not Randomized; U - Unclear B: BlindingD - Double-blinded; N - Not Blinded; U - Unclear N: # Enrolled     D%: % Dropout Outcome Variables:IMPR-Improvement QOL-Quality of Life WELL-Wellness FATG-Fatigue VIGR-Vigor/Energy SLEP-Sleep Disturbance SEVR-Symptom Severity PAIN-Pain MYLG-Myalgia HEAD-Headache ARTH-Arthralgia PFAT-Postexertional Fatigue WEAK-Muscle Weakness DIZZ-Dizziness/Lightheadedness

Table 12. Behavioral therapies with/without immunologic (more...)

Table 12. Behavioral therapies with/without immunologic therapy contrasts: General health and somatic outcomes

Intervention Type	Case Definition	Study: Therapy Contrast	Quality				General			Somatic
			A	B	N	D%	I M P R	Q O L	W E L L	F A T G	V I G R	S L E P	S E V R	P A I N	M Y L G	H E A D	A R T H	P F A T	W E A K	D I Z Z
Immunologic and Behavioral	Australia 90	Lloyd:132 Transfer factor & CBT (reestablish activity) vs. drug placebo & “CBT placebo” (attend clinic)	R	D	90	2%		↑
Behavioral		Lloyd:132 CBT (reestablish activity) & drug placebo vs. drug placebo & “CBT placebo” (attend clinic)
	Oxford 91	Sharpe:137 CBT (increase activity) & medical care vs. medical care only	R	N	60	0%				↑
	CDC 94	Deale:145 CBT(increase activity) vs. relaxation	R	N	60	12%	↑			↑
		Ridsdale:164 CBT (activity planning) vs. counseling; (CFS subgroup)	R	N	45	18%
		Prins:163 CBT (increase activity) vs. guided support group	R	N	278	3%		↑		↑
		Prins:163 CBT (increase activity) vs. no treatment								↑
	CDC 88	Friedberg:134 CBT: (activity limitations) vs. no treatment	N	D	64	0%
Symbols and Abbreviations:Results: ↑ Favors therapy group No difference between groups ↓ Favors placebo group Quality A: Assignment MethodR - Randomized; N - Not Randomized; U - Unclear B: BlindingD - Double-blinded; N - Not Blinded; U - Unclear N: # Enrolled     D%: % Dropout Outcome Variables:IMPR-Improvement QOL-Quality of Life WELL-Wellness FATG-Fatigue VIGR-Vigor/Energy SLEP-Sleep Disturbance SEVR-Symptom Severity PAIN-Pain MYLG-Myalgia HEAD-Headache ARTH-Arthralgia PFAT-Postexertional Fatigue WEAK-Muscle Weakness DIZZ-Dizziness/Lightheadedness

Table 13. Behavioral therapies with/without pharmacologic (more...)

Table 13. Behavioral therapies with/without pharmacologic therapy, complementary/alternative medicine therapies, and other therapy contrasts: General health and somatic outcomes

Intervention Type	Case Definition	Study: Therapy Contrast	Quality				General			Somatic
			A	B	N	D%	I M P R	Q O L	W E L L	F A T G	V I G R	S L E P	S E V R	P A I N	M Y L G	H E A D	A R T H	P F A T	W E A K	D I Z Z
Pharmacologic and Behavioral	Oxford 91	Wearden:151 Fluoxetine with graded exercise vs. placebo only	R	D	136	29%
Behavioral		Wearden:151 Graded exercise with drug placebo vs. placebo only151
		Fulcher:146 Graded exercise vs. flexibility/relaxation	R	N	66	29%	↑		↑	↑
		Powell:162 Exercise education, maximum intervention vs. standard care	R	N	148l	14%	↑			↑
		Powell:162 Exercise education, telephone intervention vs. standard care					↑			↑
		Powell:162 Exercise education, minimum intervention vs. standard care					↑			↑
Complementary/Alternative Medicine	CDC 88	Field:148 Massage therapy vs. sham TENS	R	N	20	0%				↑		↑			↑
		Perrin:154 Osteopathy vs. normal care	N	N	80	28%	↑
Other	CDC 94	Shlaes:142 Social support (buddy/mentor) vs. waiting list (no treatment)	N	N	12	17%				↑
Symbols and Abbreviations:Results: ↑ Favors therapy group No difference between groups ↓ Favors placebo group Quality A: Assignment MethodR - Randomized; N - Not Randomized; U - Unclear B: BlindingD - Double-blinded; N - Not Blinded; U - Unclear N: # Enrolled     D%: % Dropout Outcome Variables:IMPR-Improvement QOL-Quality of Life WELL-Wellness FATG-Fatigue VIGR-Vigor/Energy SLEP-Sleep Disturbance SEVR-Symptom Severity PAIN-Pain MYLG-Myalgia HEAD-Headache ARTH-Arthralgia PFAT-Postexertional Fatigue WEAK-Muscle Weakness DIZZ-Dizziness/Lightheadedness

Table 14. Immunologic and supplemental therapies (vs. (more...)

Table 14. Immunologic and supplemental therapies (vs. placebo): Functional, mood and cognitive outcomes

Intervention Type	Case Definition	Study	Therapy	Quality				Functional					Mood						Cognitive
				A	B	N	D%	P H Y S	A C T V	I M P R	S O C L	W O R K	D E P R	A N X Y	M N T L	D S T R	C O N F	M F T G	M E M R	C O G N	C O N C
Immunologic	CDC 1988	Strauss127	Antiviral	R	D	27	11%						↓	↓			↓
		Steinberg138	Antihistamine	R	D	30	7%
		Lloyd129	Immunomodulator	R	D	49	4%
		Peterson130		R	D	30	7%				↓
		See140		R	D	30	13%
		Strayer135	Immunomodulator/Antiviral	R	D	92	9%	↑	↑											↑
	CDC 1994	De Vinci143	Transfer Factor	R	D	20	10%
		Andersson156	Vaccine	U	D	28	14%
	Australia 1990	Vollmer-Conna147	Immunomodulator	R	D	99	4%
		Lloyd132	Transfer Factor	R	D	90	2%
Supplement	CDC 1988	Cox131	Magnesium	R	D	32	3%									↑
		Kaslow128	Liver Extract	R	D	15	7%
	Oxford 1991	Warren157	Essential Fatty Acid	R	D	50	24%
Symbols and Abbreviations:Results: ↑ Favors therapy group No difference between groups ↓ Favors placebo group Quality A: Assignment MethodR - Randomized; N - Not Randomized; U - Unclear B: BlindingD - Double-blinded; N - Not Blinded; U - Unclear N: # Enrolled     D%: % Dropout Outcome Variables:PHYS-Physical functioning ACTV-Activity level IMPR-Impairment/impact SOCL-Social functioning WORK-Occupational functioning DEPR-Depression ANXY-Anxiety MNTL-Mental health DSTR-Distress CONF-Confusion MFTG-Mental fatigue MEMR-Memory/forgetfulness COGN-Cognitive function CONC-Concentration

Table 15. Pharmacologic therapies (vs. placebo): Functional, (more...)

Table 15. Pharmacologic therapies (vs. placebo): Functional, mood and cognitive outcomes

Intervention Type	Case Definition	Study	Therapy	Quality				Functional					Mood						Cognitive
				A	B	N	D%	P H Y S	A C T V	I M P R	S O C L	W O R K	D E P R	A N X Y	M N T L	D S T R	C O N F	M F T G	M E M R	C O G N	C O N C
Pharmacologic	CDC 1988	Natelson136	Antidepressant	R	D	24	25%
		Natelson152		N	U	25	24%							↑
		McKenzie150	Glucocorticoid	R	D	70	10%
		Peterson149	Mineralcorticoid	R	D	25	20%
	CDC 1994	Hickie161	Antidepressant	R	D	90	14%
		Cleare158	Glucocorticoid	R	D	35	9%			↑	↑	↑
		Rowe160	Mineralcorticoid	R	D	100	17%
		Forsyth159	NADH	R	D	35	26%
	Oxford 1991	Vercoulen139	Antidepressant	R	D	107	10%
		Wearden151		R	D	136	29%
Symbols and Abbreviations:Results: ↑ Favors therapy group No difference between groups ↓ Favors placebo group Quality A: Assignment MethodR - Randomized; N - Not Randomized; U - Unclear B: BlindingD - Double-blinded; N - Not Blinded; U - Unclear N: # Enrolled     D%: % Dropout Outcome Variables:PHYS-Physical functioning ACTV-Activity level IMPR-Impairment/impact SOCL-Social functioning WORK-Occupational functioning DEPR-Depression ANXY-Anxiety MNTL-Mental health DSTR-Distress CONF-Confusion MFTG-Mental fatigue MEMR-Memory/forgetfulness COGN-Cognitive function CONC-Concentration

Table 16. Behavioral therapies with/without immunologic (more...)

Table 16. Behavioral therapies with/without immunologic therapy: Functional, mood and cognitive outcomes

Intervention Type	Case Definition	Study: Therapy Contrast	Quality				Functional					Mood						Cognitive
			A	B	N	D%	P H Y S	A C T V	I M P R	S O C L	W O R K	D E P R	A N X Y	M N T L	D S T R	C O N F	M F T G	M E M R	C O G N	C O N C
Immunologic and Behavioral	Australia 90	Lloyd:132 Transfer factor & CBT (reestablish activity) vs. drug placebo & “CBT placebo” (attend clinic)	R	D	90	2%
Behavioral		Lloyd:132 CBT (reestablish activity) & drug placebo vs. drug placebo & “CBT placebo” (attend clinic)
	Oxford 91	Sharpe:137 CBT (increase activity) & medical care vs. medical care only	R	N	60	0%		↑	↑			↑
	CDC 94	Deale:145 CBT(increase activity) vs. relaxation	R	N	60	12%	↑		↑	↑	↑
		Ridsdale:164 CBT (activity planning) vs. counseling; (CFS subgroup)	R	N	45	18%
		Prins:163 CBT (increase activity) vs. guided support group	R	N	278	3%	↑		↑					↑
		Prins:163 CBT (increase activity) vs. no treatment					↑		↑					↑
	CDC 88	Friedberg:134 CBT: (activity limitations) vs. no treatment	N	D	64	0%
Symbols and Abbreviations:Results: ↑ Favors therapy group No difference between groups ↓ Favors placebo group Quality A: Assignment MethodR - Randomized; N - Not Randomized; U - Unclear B: BlindingD - Double-blinded; N - Not Blinded; U - Unclear N: # Enrolled     D%: % Dropout Outcome Variables:PHYS-Physical functioning ACTV-Activity level IMPR-Impairment/impact SOCL-Social functioning WORK-Occupational functioning DEPR-Depression ANXY-Anxiety MNTL-Mental health DSTR-Distress CONF-Confusion MFTG-Mental fatigue MEMR-Memory/forgetfulness COGN-Cognitive function CONC-Concentration

Table 17. Behavioral therapies with/without pharmacologic (more...)

Table 17. Behavioral therapies with/without pharmacologic therapy, complementary/alternative medicine therapies, and other therapy contrasts: Functional, mood and cognitive outcomes

Intervention Type	Case Definition	Study: Therapy Contrast	Quality				Functional					Mood						Cognitive
			A	B	N	D%	P H Y S	A C T V	I M P R	S O C L	W O R K	D E P R	A N X Y	M N T L	D S T R	C O N F	M F T G	M E M R	C O G N	C O N C
Pharmacologic and Behavioral	Oxford 91	Wearden:151 Fluoxetine with graded exercise vs. placebo only	R	D	136	29%
Behavioral		Wearden:151 Graded exercise with drug placebo vs. placebo only151
		Fulcher:146 Graded exercise vs. flexibility/relaxation	R	N	66	29%	↑
		Powell:162 Exercise education, maximum intervention vs. standard care	R	N	148l	14%	↑
		Powell:162 Exercise education, telephone intervention vs. standard care					↑
		Powell:162 Exercise education, minimum intervention vs. standard care					↑
Complementary/Alternative Medicine	CDC 88	Field:148 Massage therapy vs. sham TENS	R	N	20	0%						↑	↑
		Perrin:154 Osteopathy vs. normal care	N	N	80	28%
Other	CDC 94	Shlaes:142 Social support (buddy/mentor) vs. waiting list (no treatment)	N	N	12	17%
Symbols and Abbreviations:Results: ↑ Favors therapy group No difference between groups ↓ Favors placebo group Quality A: Assignment MethodR - Randomized; N - Not Randomized; U - Unclear B: BlindingD - Double-blinded; N - Not Blinded; U - Unclear N: # Enrolled     D%: % Dropout Outcome Variables:PHYS-Physical functioning ACTV-Activity level IMPR-Impairment/impact SOCL-Social functioning WORK-Occupational functioning DEPR-Depression ANXY-Anxiety MNTL-Mental health DSTR-Distress CONF-Confusion MFTG-Mental fatigue MEMR-Memory/forgetfulness COGN-Cognitive function CONC-Concentration

In this section, we present eight tables that summarize the controlled trials by particular characteristics. Table 6 classifies the controlled trials according to the CFS definition they used, their country of origin and year the study was published. Table 7 categorizes the studies by types of therapy and Tables 8 and 9 (at the end of this section) present relevant information regarding study methods, intervention protocol, participants and outcomes for each trial. Tables 10 through 13 (at the end of this section) present a method for qualitatively summarizing the literature.

Chronology of controlled trials

Table 6 classifies the controlled trials according to the CFS definition they used, their country of origin and year the study was published. The earliest of these trials was reported in 1988.¹²⁷ The frequency of subsequent trials when initially reported in the medical literature was as follows: one in 1989;¹²⁸ two in 1990;^129,130one in 1991;¹³¹ two in 1993;^132,133 two in 1994;^134,135 nine in 1996;^136-144four in 1997;^145-148 eight in 1998;^149-156 three in 1999^157-159 and five in 2000.^160-164

Case definitions

The trials used different CFS case definitions for enrolling patients and the selection appears to have been based primarily on when the study was conducted and which case definitions were available. Seven studies used more than one case definition to enroll patients or stated that enrolled patients met the criteria of multiple case definitions.^{131,136,145,149,150,152,158} Of the 38 controlled trials included in this evidence report, 26% used the CDC 1988 case definition to enroll patients (3 of these 10 studies used this earlier definition only even though these studies were published between 1996 and 1999); 8% used the Australia 1990 case definition, 24% used the Oxford 1991 case definition, and the remaining 42% used the CDC 1994 case definition (trials using more than one case definition were counted, for this purpose, as using whichever case definition was the latest).

Study origin

Approximately three-fourths of the controlled trials that met the selection criteria for this report were conducted in the United States (42%) or the United Kingdom (34%). Sixteen percent were conducted in Australia (11%) or the Netherlands (5%). The remaining studies (11%) were equally divided between Belgium, Canada, Iceland and Sweden.

Types of therapies studied

Thirty-six of the 38 controlled trials compared a single therapy to a placebo or other type of control group. The two remaining trials were multi-arm trials comparing an immunological therapy (an immunomodulatory, transfer factor/dialyzable leukocyte extract) with or without cognitive behavioral therapy (CBT),¹³² and a pharmacological therapy (an antidepressant, fluoxetine) with or without graded exercise.¹⁵¹ The trial comparing transfer factor with or without CBT used a drug placebo and an “attention” placebo defined as “attending clinic;” the antidepressant with or without exercise study used a drug placebo and as a placebo for excise, gave non-specific advice regarding activity level. Of the remaining 36 single-therapy trials, 23 compared the therapy to a placebo. Nine of these placebo-controlled single-therapy trials compared an immunological therapy to placebo: antiviral,¹²⁷ antihistamine,¹³⁸ immunomodulator (immunoglobulin,^129,130,147 interferon,¹⁴⁰ and a transfer factor using dialyzable extract from immune lymphocytes¹⁴³) an immunomodulator/antiviral (Ampligen),¹³⁵ and a vaccine (staphylococcus toxoid).¹⁵⁶ The remaining single-therapy immunological trial compared an immunomodulator (interferon) to a no treatment comparison group.¹³³ Eleven of the single-therapy placebo-controlled trials evaluated a pharmacological therapy: steroids (fludrocortisone^149,160 and hydrocortisone^150,158), anticholinergic (galanthamine),¹⁴¹ growth hormone,¹⁵³ NADH (nicotinamide adenine dinucleotide),¹⁵⁹ or antidepressants (phenelzine, fluoxetine, moclobemide and selegiline).^{136,139,152,161} Three single-therapy, placebo controlled trials evaluated a supplement: magnesium,¹³¹ essential fatty acid,¹⁵⁷ and liver extract.¹²⁸ The remaining single-therapy, placebo-controlled trial intervened with homeopathic therapies.¹⁴⁴ Five of the remaining single-therapy trials evaluated complex multidimensional therapies including CBT, social support and therapy described as “comprehensive.” The five CBT trials compared CBT to: medical care,¹³⁷ relaxation,¹⁴⁵ counseling,¹⁶⁴ natural course (no treatment),¹⁶³ or no treatment.¹³⁴ The social support trial compared a “buddy/mentor” program to a waiting-list control.¹⁴² The “comprehensive” therapy intervention compared a multitude of services (e.g., optimal medical management, treatment of affective or anxiety disorders, CBT, etc.) to a no treatment control.¹⁵⁵ Four trials compared a physical intervention: graded exercise vs. flexibility and relaxation,¹⁴⁶ an educational intervention about the importance of exercise vs. standard medical care,¹⁶² massage therapy vs. sham TENS,¹⁴⁸ and osteopathy vs. normal care.¹⁵⁴

Quality of the trials

Six trials were clearly not randomized^{128,134,142,152,154,155} and randomization was unclear in one trial.¹⁵⁶ Among trials that evaluated immunologic, nutritional supplements, or pharmacologic agents, all but two were reported as double-blinded.^133,152 Among non-pharmacologic/non-immunologic trials, double-blinding was not reported in eleven studies;^{134,137,142,145,146,148,154,155,162-164}Numbers of CFS participants in the studies were fewer than 90 except for nine studies that had sample sizes ranging from 90 to 278.^{132,135,139,147,151,160-163}Dropout rates ranged from 0 to 69%; three-fourths of the studies had dropout rates less than 20%. Approximately 20% of the studies used an intention to treat analyses. Follow-up periods, from baseline, ranged from 1 week to 32 months. Seventy-five percent of the studies had follow-up periods less than 6 months, eight percent had follow-up periods of 7-8 months, and thirteen percent had follow-up periods of 11-12 months.