Contents
Contributors
Search

INTRODUCTION

Although the term fibromyalgia is relatively new, for centuries in the medical literature there have been a variety of semantic terms applied to persons who suffer from widespread musculoskeletal pain. The most recent term prior to fibromyalgia, fibrositis, implied that there was an inflammatory process within the connective tissues. This term was abandoned when research showed that this condition is not due to inflammation in the tissues. The currently preferred label, fibromyalgia, is a descriptive term that connotes that there is pain within the tissues, and does not attempt to define the pathogenesis of the pain.

To fulfill the criteria for fibromyalgia established by an American College of Rheumatology committee in 1990, an individual must have both a history of chronic widespread pain involving all four quadrants of the body (and the axial skeleton), and the presence of 11 of 18 "tender points" on physical examination (1).

One important type of study which has been performed using these criteria is population-based research to estimate the frequency of this illness. These studies have shown that in several different industrialized countries, the prevalence of fibromyalgia is fairly constant, ranging from 1 - 4% in the general population. This places fibromyalgia as the second most common rheumatologic disorder, behind osteoarthritis. An interesting finding in all of these studies is that pain and tenderness occur as a continuum in general population. Thus, some individuals rarely experience pain, others have intermittent pain, and yet others experience continuous widespread pain. Tenderness, or for that matter other somatic symptoms such as fatigue or memory difficulties, is also continuously distributed over a wide continuum in the general population. And symptoms such as pain, fatigue, and memory difficulties tend to aggregate in the population, with persons with any one of these symptoms likely to experience the other (2).

This clustering of somatic symptoms leads to several recognized systemic conditions defined on the basis of idiopathic chronic pain and fatigue. In addition to fibromyalgia, chronic fatigue syndrome (CFS), multiple chemical sensitivity, and somatoform disorders would fall within this spectrum. Both population- and clinic-based studies have demonstrated a tremendous clinical overlap between these disorders (3) (see Figure 33-1).

Figure 33-1.

There are also a variety of less frequent conditions that share considerable homology with these illnesses. Some of these conditions are referred to as "exposure syndromes," since the illness is defined on the basis of an exposure suspected to cause the symptom complex, rather than solely on the basis of symptoms (e.g., Gulf War Illness, sick building syndrome). Other illnesses within this spectrum affect only one organ system or portion of the body, with the seminal features being pain and/or dysfunction in this region (e.g., migraine headaches, irritable bowel syndrome, temporomandibular joint dysfunction).

In summary, the differences between the aforementioned syndromes characterized by chronic pain and fatigue are more semantic than real, with the demarcations that have been established to separate these conditions largely historical and artificial. The aggregate data on these syndromes suggest that at least 10 - 20% of the population suffer from otherwise unexplained chronic pain, chronic fatigue, or both. This symptom complex is more common in females than males, but not overwhelmingly so. And some of these individuals have an identifiable psychiatric condition which is responsible for these symptoms, but the majority do not.

To completely understand fibromyalgia, one must move well outside the field of rheumatology and examine the larger body of data that has been collected on these syndromes. For example, there is some evidence of familial aggregation for nearly all of the illnesses within this spectrum (3). Family members of patients with fibromyalgia display a higher than expected frequency of fibromyalgia, as well as conditions related to fibromyalgia, including irritable bowel syndrome, migraine headaches, and mood disorders (4). This group of inter-related systemic and regional syndromes that aggregate in families has been termed the "affective spectrum" by Hudson and colleagues, and "dysregulation spectrum" by Yunus.

ETIOLOGY

The "etiology" of fibromyalgia can be considered in many ways. Like many illnesses, the expression of fibromyalgia may occur when a person who is genetically predisposed comes in contact with certain environmental exposures that can trigger the development of symptoms. There are many environmental exposures that are generally accepted triggers of fibromyalgia, all of which fall into the general category of "stressors". Examples of stressors include physical trauma (especially to the axial skeleton), infections (e.g. parvovirus, Hepatitis C), emotional distress (acute or chronic), endocrine disorders (e.g., hypothyroidism), and immune stimulation, as may occur in a variety of autoimmune disorders (4;5). Although studies of groups of individuals suggest that there are many "stressors" that can trigger the development of this illness, because of the plethora of potential exposures an individual may be exposed to, it is sometimes difficult to assess the putative role of a single environmental stressor in a single individual.

The epidemic of "Gulf War Illness" that occurred in troops deployed to the Persian Gulf in 1990 and 1991 affords an excellent example of how illnesses such as fibromyalgia and chronic fatigue syndrome may be "triggered". To review, in 1990 and 1991 the U.S. deployed approximately 700,000 troops to the Persian Gulf to liberate Kuwait from Iraqi occupation. Fortunately, there were relatively few combat-related injuries and diseases during this conflict, but up to 45% of deployed veterans (as compared to 15% of non-deployed veterans) developed a constellation of symptoms and syndromes including muscle and joint pain, fatigue, memory problems, headaches, and gastrointestinal complaints (6). This experience was not unique to U.S. troops, since veterans of this conflict from the United Kingdom experienced a similar increase in this spectrum of illness (7). Several expert panels have been convened to examine these illnesses. There is agreement that this is not a single illness, but rather a constellation of symptoms and syndromes very similar to that seen in fibromyalgia and CFS. And there is no single environmental exposure that is likely to have contributed to this illness. Instead, there are data from many sources that suggest that certain persons who are exposed to "stressors", including infections or other types of immune stimulation, drugs or chemicals, physical trauma, or emotional stress, will develop a chronic multi-system illness (8).

PATHOGENESIS

Most investigators in this field feel that the primary abnormality that leads to expression of symptoms in fibromyalgia and related conditions is aberrant central nervous system function (3;9-11). Furthermore, there is a general belief that the central components of the "stress response" are playing a major role in symptom expression, in that these systems are capable of being activated by a variety of stressors, and disturbances in this system can have effects on sensory processing, autonomic, and neuroendocrine function. The principal components of the human stress response are the corticotropin-releasing hormone (CRH) and locus ceruleus-norepinephrine/autonomic (sympathetic / LC-NE) nervous systems (12). The CRH system is primarily centered in the hypothalamus, and the sympathetic / LC-NE system in the brain stem. Activation of these systems by a variety of stimuli propagates a series of physiologic changes known as the "stress response."

There is a substantial body of literature regarding this system, which cannot be reviewed in detail. But there are several relatively recent advances in this area that may be germaine to illnesses such as fibromyalgia. For example, several authors have noted that although this system is adaptive in animals and early human species, it may be maladaptive in the twentieth century. In everyday life this "stress" system is much more likely to be activated by daily events which have no threat to survival (e.g., sitting in traffic) than for the intended purposes of this system, e.g., to protect against predators, starvation, etc.(13).

It is also important to recognize that different types of "stress" lead to markedly different biological responses, in both animals and humans. And the environment within which stress occurs may be the most important determinant of the physiologic consequences. Stressors perceived as inescapable or unavoidable, or which are accompanied by lack of predictability or support, evoke the strongest adverse biological consequences (12-14). This could conceivably explain why victims of trauma, such as motor vehicle accidents, appear to have a much higher rate of the development of fibromyalgia and myofascial pain than those that are responsible for the accident.

Finally, within any species, there are genetic differences in the activity of the biological stress response, as well as in systems that process sensory information such as pain. The aggregate data collected from such studies suggests that individuals may be born with a certain "set point" for the functioning of such systems, and that subsequent environmental exposures may change that set point over the life of that organism. For example, early life stressors can have a permanent and profound impact on the subsequent biological response to stress in animals, because of the plasticity of the nervous system. Studies in rodents have demonstrated that exposure to endotoxin, trauma, or separation in the neonatal period all lead to permanent changes in the subsequent biological response to stress, extending throughout the life of the animal (13-15). This plasticity may be due to changes in the numbers of neurons, number of circuits, and/or increases or decreases in gene expression - leading to permanent changes in molecules that define the function of the system. This permanent effect of early stressors could explain why individuals who develop fibromyalgia, CFS, somatoform disorders, IBS, and other disorders in this spectrum report a higher than expected incidence of childhood physical and sexual abuse.

The areas of nervous system function that may be playing some role in the pathogenesis of fibromyalgia include: sensory processing, neuroendocrine and autonomic, and psychobehavioral influences.

Abnormalities in sensory processing

The primary abnormality in fibromyalgia is that individuals experience a diffusely lowered pain threshold. The tenderness in fibromyalgia is not in any way confined to "tender points". Relatively speaking, individuals with fibromyalgia are just as tender in areas such as the forehead and thumbnail as they are in areas considered "tender points"; the only difference is that everyone is more tender in areas considered tender points. The diffuse nature of the tenderness, or allodynia, has been a primary reason that the search for the cause of fibromyalgia has moved into the central nervous system.

Further studies have demonstrated that fibromyalgia patients can not detect electrical, pressure, or thermal stimuli at lower levels than normals, but the point at which these stimuli cause pain or unpleasantless is lower (16;17). Other studies have examined regional differences in pain sensitivity, and have demonstrated that although tender points are anatomic locations that are more sensitive to pressure, these regions are actually less responsive to both electrical and thermal stimuli than "control points" (16). And although nearly all of the research on sensory processing in fibromyalgia has focused on the processing of pain, there are some data suggesting a more generalized disturbance in sensory processing. For example, many patients experience sensitivity to loud noises, bright lights, odors, drugs, and chemicals. These symptoms of generalized sensitivity to multiple stimuli account for the significant number of persons with fibromyalgia that also could be classified with "multiple chemical sensitivity" (an acknowledged misnomer).

Hypothalamic-pituitary dysfunction

There are substantial data indicating that the hypothalamic-pituitary axes function abnormally in subsets of persons with fibromyalgia and related disorders (18). Beginning with the early work of Demitrack and colleagues at the NIH, there has been a recognition that individuals with chronic fatigue syndrome (CFS) display a number of abnormalities in neuroendocrine function, especially involving the hypothalamic pituitary adrenal (HPA) axis (11). Most early studies of HPA function in CFS demonstrated a low 24 hour free cortisol excretion, increased adrenocortical sensitivity to ACTH, and attenuated ACTH responses to CRH. (19;20). Early neuroendocrine studies in fibromyalgia yielded similar results, with a relative hyporesponsiveness of the adrenal glands (decreased production of cortisol in response to CRH or ACTH), low 24 hour urine free cortisol, and an exaggerated pituitary response to CRH (21;22). Some of the data appear to indicate a primary adrenal insufficiency (supported by a single study reporting diminished size of adrenal glands in CFS (23)) whereas others suggest a hypothalamic defect.

However, more recent studies have yielded somewhat different results. When fibromyalgia subjects are studied in a more natural environment (exposed to daily or experimental stressors), cortisol levels at several points throughout the day (in one study salivary and the other plasma) have been shown to be consistently elevated, and urinary free cortisol is normal (24). In these latter studies, subjects were compared either while performing normal daily activities, or while in a strictly monitored CRC setting, being exposed to standardized physiologic stressors. Finally, many recent studies have noted no differences between CFS or fibromyalgia subjects and controls with respect to HPA function, with these studies generally having been more vigilant to studying only medication-free subjects with no psychiatric co-morbidities (25;26).

A more consistent finding in both CFS and fibromyalgia is hypo-responsiveness of effector arms of the stress response to standardized acute stressors. In one study in fibromyalgia, cortisol levels paradoxically fall rather than rise in response to physical exertion (27). Torpy and colleagues demonstrated a related hyporesponsiveness to the stress of an IL-6 infusion in fibromyalgia subjects, noting exaggerated ACTH but normal cortisol levels (28). Perhaps the most consistent finding regarding autonomic function is that fibromyalgia patients have an attenuated HPA and/or catecholamine response to several different stressors (e.g., exercise, hypoglycemia, muscle contraction, and noise) (29-32).

The attractiveness of the "blunted stress response" theory is that similar abnormalities suggesting a blunting of the HPA axis have also been noted in many of the less common chronic fatigue states, such as dysthymia or seasonal affective disorder (12;33-36). These changes are opposite to those seen in melancholic depression, which is characterized by chronically increased stress system activity (12;35;36). In the only study to directly compare CFS subjects to those with depression, the CFS subjects were found to have low 24 hour urinary free cortisol (regardless of whether they had concurrent depression or not), and the depressed subjects exhibited the expected high urinary free cortisol values (37).

The reason for these opposite changes in the function of the stress response in clinically similar disorders is not clear. Some have hypothesized that any disturbances in stress system activity, be it increased or decreased, can upset homeostasis, and, thus, impair performance (12;38). Alternatively, subtypes of major depression may have biologically disparate causes, as has been suggested by genetic studies. A closer examination may reveal that only certain subtypes of depression are associated with fibromyalgia and CFS. Finally, it is possible that the HPA abnormalities are surrogates for other changes in the central nervous system which then directly causes mood disturbances, or have effects on areas such as the dopaminergic mesocorticolimbic system or the reticular activating system. Regardless, these disparate findings between chronic pain and fatigue states and melancholic depression further supports the contention that these are biologically different illnesses.

There are a number of other neuroendocrine changes identified in patients with CFS and fibromyalgia which suggest a blunted stress response. These include:

• Growth hormone and Insulin-like growth factor-1 [IGF-1] (somatomedin-C). IGF-1 is produced in the liver, primarily in response to Growth Hormone (GH), and is responsible for many of the biological activities of GH. The first work done of this axes in these illnesses was by Bennett, who found that IGF-1 was low in fibromyalgia patients, and in individuals who have fibromyalgia and a concurrent rheumatic disorder such as osteoarthritis or rheumatoid arthritis (39;40). Only very low levels of IGF-1 are specific for fibromyalgia, and occur in approximately one third of patients (41). The defect in GH secretion that leads to the low IGF-1 appears to be hypothalamic in origin, since these individuals fail to secrete GH in response to a variety of types of stimulations (41). 
  More recent studies in CFS have confirmed and extended these findings. Moorkens et. al. studied 73 non-obese CFS subjects and compared these results to 21 controls, and found low nocturnal GH secretion in CFS, and blunted GH responses to hypoglycemia, although in this cohort IGF-1 levels were not found to be lower (26). Of note, in this study HPA function as measured by plasma cortisol and ACTH was normal. Conflicting results, though, were noted by Cleare et. al. who found that basal and stimulated (with insulin and GHRH) levels of GH and IGF-1 were normal in 37 medication-free CFS with no concurrent psychiatric diagnoses, compared with matched controls (42).
  
• Reproductive hormones. There have been few systematic studies of the levels of sex hormones, or of the integrity of the hypothalamic-pituitary-gonadal axis, in these illnesses. Low DHEA and DHEA-S have been noted in several abstracts in both CFS and fibromyalgia, perhaps the most consistent finding 
  
• Thyroid hormones. Several studies have addressed this issue, with most suggesting that individuals with either CFS or fibromyalgia display blunted secretion of thyrotropin and thyroid hormones in response to thyroid releasing hormone, again suggesting a blunted hypothalamic response (43).

Autonomic nervous system

There are also identifiable abnormalities in autonomic nervous system function in many of the disorders in this spectrum. Just as with studies of neuroendocrine function, though, only a subset of fibromyalgia patients will have "abnormal" autonomic function, depending on how this is defined. In summary, various studies have demonstrated that subsets of persons with fibromyalgia, as well as other similar disorders such as CFS, display low baseline sympathetic tone, and an inability to respond to stressors (27;29;32). The clinical manifestations that are related to autonomic dysfunction are not entirely clear, but may include orthostatic intolerance (e.g., as in neurally mediated hypotension), vasomotor instability, and visceral dysfunction.

Psychiatric, psychological, and behavioral factors

There has been a longstanding debate over the role of psychiatric, psychological, and behavioral factors in fibromyalgia. Some contend that all of these symptoms are "supra-tentorial" in origin, or that fibromyalgia represents a state of distress or vulnerability, whereas others counter that the rate of psychiatric co-morbidities in these conditions is similar to any chronic disease (44;45). A review of the accumulated data in these conditions supports a few consistent observations.

Approximately 20 - 40% of individuals with fibromyalgia seen in tertiary care centers have an identifiable current mood disorder such as depression or anxiety disorder (46). The lifetime incidence of psychiatric co-morbidities in tertiary care patients may be as high as 40 - 70% over several studies (46;47). These data are among those used by Hudson and colleagues to posit that there is a spectrum of disorders including fibromyalgia, migraines, irritable bowel, and affective disorders which may share a common genetic predisposition, and underlying pathogenic mechanisms.

However, some of these differences in the current and lifetime history of mood disorders are likely due to health care seeking behaviors, since lower lifetime incidences of affective disorders are typically noted in individuals with fibromyalgia who are identified in the general population (48). This same relationship between the setting of care, and the rate of co-morbid psychiatric conditions, has been consistently noted in irritable bowel syndrome.

There are a myriad of complex psychosocial factors which play a significant role in some individuals with fibromyalgia, as with nearly any chronic medical illness. These include behavioral pathways, such as sick role behavior and maladaptive coping mechanisms, cognitive pathways such as victimization and loss of control, and social pathways, such as interference with role functioning and deterioration of social or other support networks. Psychosocial factors are known to play a particularly prominent role in the transition form acute pain to chronic pain and disability. As pain progresses from the acute phase into chronicity, problems emerge for the individual such as job loss, financial constraints, distancing of friends, etc. If patients' responses to these problems are maladaptive such as avoidance of work, friends, financial responsibilities and physical activity, the patient may become distressed and overwhelmed by the pain and its negative impact on life. Increased stress, learned helplessness, depression, increased anxiety, anger, distrust, entitlement, and somatization can all emerge and worsen symptoms, probably by inter-related physiologic and psychologic mechanisms. All of these factors can be important in dictating how individuals report symptoms, how and when they seek health care, and how they respond to therapy. This may also explain why cognitive-behavioral therapy, which addresses many of these issues, has generally been effective in the treatment of individuals with fibromyalgia, as well as nearly any other chronic medical condition (49).

CLINICAL FEATURES

As noted previously, the American College of Rheumatology criteria for fibromyalgia require both a history of chronic widespread pain involving all four quadrants of the body (and the axial skeleton), and the presence of 11 of 18 "tender points" on physical examination (1) (Figure 33-2).

Figure 33-2. Fibromyalgia  Tender Points

These criteria were never intended to be strictly applied to individual patients as diagnostic criteria, and at least half of the individuals who have the clinical diagnosis of fibromyalgia will not fulfill this definition.

There are problems with the ACR criteria, and especially with the requirement that an individual needs a certain number of tender points to fulfill these criteria. Tender points are pre-defined anatomic points that are present in various areas of the body, and are considered to be "positive" when an individual complains of pain when four kilograms (approximately nine pounds) of pressure is applied (approximately the amount of pressure required to blanch the examiner's nail). Although early studies suggested that fibromyalgia patients experienced tenderness only in these discrete regions, recent data show that individuals with fibromyalgia display increased sensitivity to pain throughout the body (50). Tender points (e.g., the mid-trapezius region, epicondyles, etc.) appear to merely represent regions of the body where everyone is more tender, thus individuals who are more diffusely tender will generally have a greater number of tender points. Also, tender points measure not only how tender an individual is, but also how "distressed" they are. Finally, tenderness is influenced by many factors. Female gender, increasing age, poor aerobic fitness, and mood disorders all tend to increase cutaneous pressure sensitivity. Therefore, rigidly adhering to the ACR criteria in clinical practice will skew the diagnosis of fibromyalgia towards older females with poor aerobic fitness, and a high level of distress.

Although both pain and tenderness are defining features of fibromyalgia, the latter is rarely a presenting complaint. The pain of fibromyalgia frequently waxes and wanes, may be quite migratory, and may be accompanied by dysesthesias or paresthesias following a non-dermatomal distribution. In some instances patients will present with "aching all over", whereas in other instances patients experience several areas of chronic regional pain. In this setting, regional musculoskeletal pain typically involves the axial skeleton, or areas of "tender points", and may originally be diagnosed as a local problem (e.g., low back pain, lateral epicondylitis). Regional pain involving non-musculoskeletal regions is also common, including a higher than expected prevalence of both tension and migraine headaches, TMD (or TMJ) syndrome, noncardiac chest pain, irritable bowel syndrome, a number of entities characterized by chronic pelvic pain, and plantar or heel pain.

In addition to pain and tenderness, most individuals with this illnesses also experience a high lifetime and current prevalence of non-defining symptoms (51) (see Figure 33-2). For example, most patients with fibromyalgia experience fatigue, and at least half of individuals who meet ACR criteria for fibromyalgia will also meet criteria for chronic fatigue syndrome (CFS). The fatigue is commonly worse after activities, and may be accompanied by memory difficulties.

Memory difficulties, especially with attention and short-term memory, may be the most debilitating aspect of their illness. Other constitutional symptoms include fluctuations in weight, heat and cold intolerance, and the subjective sensation of weakness.

Patients with fibromyalgia and related illnesses also display a wide array of "allergic" symptoms, ranging from adverse reactions to drugs and environmental stimuli (as seen in multiple chemical sensitivity), to higher than expected incidences of rhinitis, nasal congestion, and lower respiratory symptoms. Although some of these individuals may truly be atopic, many of these symptoms are due to neural (e.g., hypersensitivities, vasomotor rhinitis) mechanisms.

Hearing, ocular, and vestibular abnormalities have also been noted, including a high incidence of sicca symptoms, a decreased painful sound threshold, exaggerated nystagmus and ocular dysmotility, and asymptomatic low frequency sensorineural hearing loss.

Individuals with fibromyalgia likewise suffer from a number of symptoms of "functional" disorders of visceral organs, including a high incidence of recurrent non-cardiac chest pain, heartburn, palpitations, and irritable bowel symptoms. However, prospective studies of randomly selected individuals with fibromyalgia have detected a high frequency of objective evidence of dysfunction of several visceral organs, including echocardiographic evidence of mitral valve prolapse, esophageal dysmotility, and diminished static inspiratory and expiratory pressures on pulmonary function testing. Neurally mediated hypotension and syncope also occur more frequently in these individuals. Similar syndromes characterized by visceral pain and/or smooth muscle dysmotility are also seen in the pelvis, including dysmenorrhea, urinary frequency, and urinary urgency, interstitial cystitis, endometriosis, and vulvar vestibulitis or vulvodynia.

The physical examination is generally unremarkable in fibromyalgia, other than finding tenderness. The tenderness may be virtually anywhere, and is not just confined to tender points. The former concept of "control points", previously described as areas of the body that should not be tender, has been abandoned.

Laboratory testing should be used judiciously. Even if the individual has acute or subacute onset of symptoms, ordering serologic assays such as ANA and rheumatoid factor should generally be avoided unless there is strong evidence for an autoimmune disorder. There are several reasons for this, including the fact that these tests have a low predictive value in the setting of non-specific symptoms, and that the rate of ANA positivity may be higher in persons with illnesses within this spectrum (52;53).

TREATMENT

General approach.
At present, the treatment of fibromyalgia is as much art as science. Once an individual with this diagnosis is identified, the practitioner first has to consider whether to "label" the individual. For the majority this label will help them understand their symptoms and the most appropriate treatment, but there may be some individuals for whom this is harmful (45).

The practitioner should schedule a prolonged visit, or series of visits, when this diagnosis is considered. Although there are no data to support this, it is likely that this "up-front" time is extremely useful for both patients and providers, as it helps the physician understand precisely what is bothering the patient, and assists the patient in understanding the goals and rationale of treatment. The physician should explore the symptoms that are most bothersome, the impact these symptoms are having on various aspects of the patient's life, their perception about what is causing these symptoms, and the stressors that may be exacerbating the problem.

Education.
Some patients who present with symptoms of fibromyalgia want only to be told that this is a benign, non-progressive condition. These patients generally have milder symptoms that have been present for some time, and they possess adequate strategies for improving symptoms and maintaining function.

But for all patients with fibromyalgia, education about the nature of this disorder is critical. The physician should describe this condition in terms they feel most comfortable with, and then refer the patient to reputable sources of information such as the Arthritis Foundation, several national patient support organizations (e.g., American Fibromyalgia Syndrome of America, National Fibromyalgia Research Association, and Fibromyalgia Alliance of America), or up-to-date websites (www.fms-cfs.org).

Pharmacologic therapies.
Tricyclic compounds, notably amitriptyline and cyclobenzaprine, have been the most extensively studied agents for the treatment of fibromyalgia. Low doses of tricyclic drugs have also been demonstrated to be of short-term benefit in treating a number of other conditions and symptoms, including both migraine and tension headaches, noncardiac chest pain, irritable bowel syndrome, rheumatoid arthritis, insomnia, and a variety of chronic pain conditions (54;55).

A problem with all controlled trials of tricyclic drugs in fibromyalgia is that there is tremendous variability in the optimal dose of these compounds across groups of patients, and blinded trials do not allow for highly individualized dosing regimens that can be used in clinical practice. To increase the tolerance of cyclobenzaprine and amitriptyline, these compounds should be administered several hours before bedtime, begun at low doses (10 mg or less), and increased slowly (10 mg every 1 - 2 weeks) until the patient reaches the maximally beneficial dose (up to 40 mg of cyclobenzaprine, or 70 - 80 mg of amitriptyline).

Because of the side effects of tricyclics, recent studies have examined the efficacy of better tolerated compounds in these conditions. Studies show conflicting results regarding the efficacy of fluoxetine in fibromyalgia; one trial showed a beneficial effect (either alone or with amitriptyline), and the other demonstrated no advantage over placebo. Single trials suggest that sertraline, and venlafaxine, may be of some benefit. There is some evidence that medications that augment central adrenergic tone may be more effective central analgesics that drugs that work exclusively via serotonergic mechanisms. Anecdotally, other compounds with more prominent noradrenergic and/or dopaminergic mechanisms, such as buproprion, nefazadone, and pemoline, may have some clinical utility, especially if given during the day to patients with prominent fatigue or cognitive complaints.

Other drugs may be useful for treating certain symptoms of fibromyalgia, without necessarily leading to a globally beneficial effect. For example, tramadol has been demonstrated to be an effective analgesic in this disorder.

Gabapentin may also have some efficacy as a central analgesic when used at high doses (e.g. 1000 - 2000 mg/day) but has not been tested in fibromyalgia patients. For treating insomnia in persons intolerant to tricyclic compounds, bedtime doses of trazadone and zolpidem may be of benefit. And in persons with symptoms suggestive of autonomic dysfunction, such as orthostatic intolerance, vasomotor instability, or palpitations, increased fluid and sodium/potassium intake, and/or low doses of beta blockers, might be of benefit.

Non-pharmacologic therapies.
Many feel that "multi-modal" treatment programs that combine symptom-based pharmacologic therapy with extensive use of non-pharmacologic therapies (e.g. cognitive behavioral therapy [CBT], aerobic exercise) are the most effective in this illness.

Cognitive behavioral therapy (CBT).
Cognitive behavioral therapy refers to a structured education program that focuses on teaching individuals skills that they can utilize to improve their illness. CBT has been shown to be effective in improving patient outcomes in nearly every chronic medical illness including fibromyalgia, and needs to be tailored to the specific condition being treated (49). The skills most commonly associated with CBT for pain include relaxation training, activity pacing, pleasant activity scheduling, visual imagery techniques, distraction strategies, focal point and visual distraction, cognitive restructuring, problem solving, and goal setting. A goal of CBT is to allow the patient to gain more control of their illness, and to give them the tools to accomplish this.

Aerobic exercise. Aerobic exercise has likewise been demonstrated to be effective at improving outcomes for a wide range of conditions including fibromyalgia (56). The reason for the benefit is likely multifacorial. Since aerobic exercise has an analgesic as well as anti-depressant effect, and can enhance the sense of well-being and control.

In designing an aerobic exercise program, it appears as though careful planning is required to enhance tolerability, and to ensure long-term compliance. Especially in illnesses such as fibromyalgia, patients may experience a worsening of symptoms immediately after exercise, and thus fear that any form of exercise will exacerbate their condition. To reduce the pain associated with exercise, low-impact exercises such as aquatic exercise, walking, swimming or stationary cycling are recommended. Just as with medication, a "start low, go slow" approach appears to be most effective, with a gradual progression in exercise intensity and a focus on adherence to a lifelong program being of paramount importance.

Complementary therapies. There are several different types of complementary therapies that are used by physicians and patients to treat fibromyalgia. Some of these are physical modalities such as trigger point injections, myofascial release therapy (or other "hands on" techniques), acupuncture, and chiropractic manipulation, each of which has some data supporting efficacy. Others fall under the general category of "cure du jour": nutritional supplements, diets, devices, etc. that are frequently advertised over the Internet, usually accompanied by testimonials to their efficacy.

Since there are very few controlled trials can guide the practitioner in how to grapple with these treatment modalities, a general approach is suggested. The practitioner should first evaluate the safety of the proposed treatment, and point out to the patient any potential harmful effects. The physician should then consider whether this treatment is reinforcing a maladaptive belief, that in the long run will be harmful to the patient (e.g., a treatment program of prolonged bed rest, or of isolation). If the treatment is neither harmful nor maladaptive, then the practitioner may suggest that the patient conduct the equivalent of a clinical trial on themselves (as is done in "n of 1" trials). In this setting, the patient begins a single treatment (keeping all other variables constant) and determines if the treatment is beneficial. If the patient judges the treatment to be helpful, then the treatment should be discontinued to determine if the symptoms worsen. If the treatment withstands this test of efficacy, a placebo effect cannot be excluded, but in clinical practice, especially in an enigmatic condition such as fibromyalgia, it is difficult to argue with success.