CENTER ON BEHAVIORAL MEDICINE

ADDITIONAL MATERIAL

RELATED PAPERS
Related Papers Menu

Fibromyalgia

Permission graciously given by the author to reproduce this paper: 
The following material is taken from the dissertation of Dr. Winter and is used with his permission.  Since fibromyalgia is such a common problem I decided to include major portions of the dissertation.  This paper begins with Chapter 1 which is a brief statement about fibromyalgia and neurofeedback.  Chapter 2, Review of the Literature, is reproduced in its completeness.  This chapter reviews what is know of fibromyalgia and the various uses of neurofeedback.  Chapter 3, Methodology, and Chapter 4, Results, have been edited but much of the original material remains.  Chapter 5 was omitted as were the appendixes.  The list of references was left complete.  A complete copy of Dr. Winter’s dissertation can be found at Northcentral University in Prescott, Arizona.
 
The Effect of EEG Neurofeedback Training in a Clinical Sample of Patients with Fibromyalgia

By

Earl F. Winter
2001

CHAPTER I

STATEMENT OF THE PROBLEM

INTRODUCTION

Electroencephalograph (EEG) neurofeedback is a training process whereby information about a person’s brainwaves is provided to the person in a manner that allows for operant conditioning (Thatcher,1999), or self-regulation, of these brainwaves. Using this technique, experimenters are able to train subjects to increase or decrease the amplitude of various brainwave frequency bands during a training session.  Once this training is complete, over several sessions, it is not necessary for the subject to attempt conscious brainwave control because this training appears to result in long-term improvements, as determined for example, by Lubar and Lubar (1999) for ADHD, Tansey (1993) for hyperactivity, Penniston and Kulkosky (1989) for alcoholics, and numerous others. 

This study evaluates EEG neurofeedback training efficacy in treating fibromyalgia patients.  Clinical patient records are used to evaluate efficacy by comparing pre- and post treatment results, and also by comparing this group to a previous group of fibromyalgia patients that had not received neurofeedback training.  The current sample includes all patients previously seen in a clinical setting (MD rheumatology office) who were treated with EEG neurofeedback for a minimum of 40 sessions.  The sample size is 15 patients.  The sample size for the previous fibromyalgia group not receiving neurofeedback training is 63 patients. 
 

CHAPTER II

REVIEW OF RELATED LITERATURE


Neurofeedback Training

The study of brainwaves and their control goes back to the early work of Kamiya (1968) and Kamiya et al. (1969) where the authors showed it was possible to train subjects to control their alpha rhythm using an audible tone as feedback.  The alpha rhythm is a brainwave rhythm between eight and 12 cycles per second, and having an amplitude of up to about 50 microvolts. 

Shortly thereafter, Sterman and Friar (1972) and Sterman (1973) demonstrated that neurofeedback training to increase the amplitude of the sensorimotor rhythm (SMR, 12-15 Hz) brainwave using visual and/or audible feedback was effective in suppressing seizures in epileptic patients.  Additional studies on seizure reduction were carried out by Seifert and Lubar (1975) who were able to obtain seizure reduction in five of six epileptics.  These patients represented a cross-section of epilepsies including grand mal, myoclonic, afocal and psychomotor types. 

Later, Lubar and Bahler (1976) were able to confirm this electroencephalograph (EEG) conditioning using a sample of 8 epileptic patients.  But also of importance in this study was the observation that one of the patients who had been markedly hyperactive prior to treatment (in addition to being epileptic) was significantly less hyperactive after SMR training (training the subject to increase the amplitude of the 12-15 Hz brainwave). 

Following this fortuitous discovery, Lubar and Shouse (1976, 1977) investigated the effect of SMR training on 24 hyperactive children without epilepsy and were able to show that training to increase SMR levels ameliorated their hyperactivity.  Since then, there have been numerous publications verifying these early findings. 

Follow-up research studies expand on these findings and include the work of Hardt and Kamiya (1978), who investigated the use of alpha feedback training as a treatment for anxiety.  They found that anxiety was reduced in proportion to alpha increases, but only for high anxiety subjects.  Low anxiety subjects were able to increase their alpha waves as well as high anxiety subjects, but their alpha changes were not related to anxiety changes.  The authors concluded that alpha feedback training of at least 5 hours duration (spread over several sessions) might be of value in treating high anxiety patients. 

Additional research was performed by Tansey and Brunner (1983) who utilized a combination of electromyograph (EMG) and EEG biofeedback training with a 10-year old boy who presented with an attention deficit disorder with hyperactivity (ADHD), developmental reading disorder, and ocular instability.  Symptoms were completely eliminated for all three of these disorders following EMG training initially to eliminate muscle tension over the central forehead area, and then followed with EEG training to increase the SMR response along the midline of the top of the skull.  The results were stable over a 24-month period following treatment.

Extensive research has been done on the use of neurofeedback in treating attention-deficit/hyperactivity disorder (ADHD).  A detailed description of this disorder, and its treatment with neurofeedback training, is given by Lubar (1995) who first began work in this area in the mid-1970s.  He reports that there are currently over 300 organizations using EEG neurofeedback in the treatment of ADHD.  This includes private clinics, university-based groups, and individuals working within school systems.

Lubar reports that anyone with a primary diagnosis of attention deficit disorder (ADD) or ADHD, between the ages of 7 and 45, without severely impaired intelligence, is a candidate for treatment with neurofeedback.  For these patients, symptoms that can be improved with neurofeedback include attention, focus, concentration, task completion and organizational skills, impulsiveness, and mild hyperactivity.  And the results of treatment include improved behavior and learning, improvement in school grades, increased self-esteem, better job performance, greater realization of innate potential, higher intelligence test scores, and improved scores on parent-teacher rating scales. 

Additionally, research has since been performed showing that neurofeedback therapy is effective in treating a variety of other disorders.  James and Folen (1996) successfully treated a case of chronic fatigue syndrome.  Tansey (1993) confirmed the long-term stability of results in a 10-year follow-up study of a boy who had originally failed fourth grade and had presented with symptomatology of a developmental reading disorder, hyperactivity, and an educational classification of perceptually impaired.  Ten years after his termination from successful neurofeedback treatment, he was classified as normal in both social and academic functioning.

In another interesting study, Tansey (1986) described the successful elimination of both a simple and a complex tic utilizing EEG SMR neurofeedback.  The subjects were a 32-year old male with a 17-year simple motor tic and a 14-year old boy with a 6-year complex tic.  The authors hypothesized that treatment was effective because exercising the sensorimotor cortex resulted in increased activation of the cerebrocortical subsystem with a resulting increased threshold for random motor discharge, thereby eliminating the tic responses.

Peniston and Kulkosky (1989) used alpha-theta brainwave training (reward alpha, while inhibiting theta) with alcoholics, and compared the results to a nonalcoholic control group and a traditionally treated alcoholic control group.  Results showed that alcoholics receiving neurofeedback training had a sharp reduction in self-assessed depression, and a 13-month follow-up showed a sustained prevention of relapse in those who completed the neurofeedback training.

Later, Saxby and Peniston (1995) studied alcoholics with depressive symptoms and found that alpha-theta brainwave training significantly reduced the following scores: schizoid, avoidant, dependant, histrionic, passive-aggression, schizotypal, borderline, anxiety, somatoform, hypomanic, dysthymic, alcohol abuse, drug abuse, psychotic thinking, and psychotic depression.

In the medical field, Nahmias, Tansey, and Karetsky (1994) were able to show that EEG neurofeedback training utilizing SMR enhancement resulted in reversal of variable extrathoracic upper airway obstruction in patients with laryngeal dyskinesis, an asthma-like disorder.  Then, Rozelle and Budzynski (1995) used SMR enhancement neurotherapy to treat a male subject who had experienced a stroke one-year prior to start of treatment.  Initially, the patient complained of hesitant speech with word finding difficulty and paraphasia, difficulty focusing his right eye, lack of balance and coordination, poor short-term memory, poor concentration, anxiety, depression, and tinnitus.  At the conclusion of neurotherapy there was noticeable improvement in speech fluency, word finding, balance and coordination, attention, and concentration.  Depression, anxiety, and tinnitus were greatly reduced.

Also in the medical field, neurofeedback training is used in the diagnosis and treatment of head injury.  Hoffman, Stockdale, Hicks, and Schwaninger (1995) describe the advances in diagnosis of Mild Traumatic Brain Injury (MTBI) made possible by utilizing quantitative measures obtained from normative or reference databases developed by Matousek and Petersen (1973), Thatcher (1980), Epstein (1980), Hudspeth (1985), and Hudspeth and Pribram (1990) and (1991).  For the treatment of MTBI, they use an average of 40 neurofeedback training sessions.  Treatment protocols are aimed at normalizing the EEG, and include beta (15-18 Hz) enhancement with theta (4-7 Hz) suppression, SMR enhancement with theta suppression, alpha (10-14 Hz) enhancement with theta suppression, and/or coherence training.  Patients come in for this training at least twice weekly, and improvement is often noted in as few as five sessions.

Another interesting use of neurofeedback is in the treatment of Lyme disease, an inflammatory arthritic disorder.  It is the most common tick borne illness in the United States, with common symptoms of fatigue, chills and fever, headache, stiff neck, backache, and sore throat.  In one article, Brown (1995) describes the successful treatment of this disorder in a 39-year-old female patient who had been diagnosed with this disease five years prior.  After a five-phase treatment program of EEG neurotherapy, the patient was symptom-free, and remained so upon follow-up.  A single electrode sensor placement at Cz was used for the majority of treatments.

Similar results were obtained by Packard and Ham (1996) who treated a 44-year old male with advanced Lyme disease.  After 40 sessions of EEG neurofeedback, the patient reported a 50% improvement in functioning.  The protocol was designed to suppress theta and enhance beta.

Neurofeedback training is now used routinely for peak performance training of both athletes and business executives.  One of the first peer reviewed papers on this subject was published by Landers et al. (1991) who used EEG neurofeedback to improve archery performance.  They were able to show, using control groups, that EEG biofeedback resulted in significant improvements in this area of athletic performance. 

Norris and Currieri (1999) also give a detailed discussion of performance enhancement training (PET) through neurofeedback, and give a review of the research related to this subject.  For example, they report that anyone who works with their hands or their body, such as athletes, surgeons, or pianists should find neurofeedback training to decrease theta and increase their sensorimotor rhythm (SMR) extremely helpful in decreasing their reaction time.  They also report that training to increase alpha and decrease theta enhances relaxation, reduces stress, improves ideational fluency and complex problem solving, improves the ability to sustain accuracy, and results in more effective decision making and improved mental tracking performance.

The medical applications of neurofeedback are reviewed and summarized by Laibow (1999) who believes that any patient involved in neurofeedback therapy should also be involved in an integrated psychotherapeutic process to help integrate any relevant material that may arise from a neurotherapy session.  The author cites the strong relationship between emotion, perception, and neurological functioning to support this belief.

Laibow also states that the CNS is the integrative and regulatory organ of the patient.  There is a very strong mind-body connection, and all stress-related symptoms and diseases initially, and over the course of the disease, involve the ANS.  Thus dysregulation of the ANS, or dysautonomia, may be treatable by neurofeedback, but this should only be done after medical consultation and involvement.

In their review of neurofeedback, Othmer, Othmer, and Kaiser (1999) present a model of the brain that treats the brain as a self-regulatory control system.  Failure or dysregulation of this control system therefore manifests itself in pathology.  In this view, basic functioning of the system is manifested in EEG activity, and therefore some failure modes of the system may be restored via neurofeedback operant conditioning of the EEG.

This EEG conditioning is aimed at achieving: “(1) an improved ability to maintain homeostasis, and (2) an improvement in stability when responding to a sudden challenge or insult to the regulatory system.”  In control system language, these are called setpoint errors, and instabilities. 

In spite of the above-mentioned positive reports on the efficacy of neurofeedback training, there is still some controversy.  For example, Barkley (1992) wrote a very negative article about using neurofeedback to treat ADHD.  One of his primary criticisms was that there were no control groups for the studies and therefore placebo or maturation effects could not be ruled out.  A second criticism was that children in these ADHD studies were receiving treatments other than neurofeedback at the same time as their neurofeedback.  Such other treatments included intensive academic tutoring, and relaxation and self-control training.  A third criticism was that the researchers did not consider practice effects on the measures being utilized to evaluate the ADHD children.  A fourth criticism was that the population reported on in the studies was too small to be statistically relevant.  And a fifth criticism was that not enough details were provided to ascertain if the children in the study actually had ADHD, or were they learning disabled children, or children merely doing poorly in school, or any combination of these factors.

These are valid criticisms.  However, in response to Barkley, Othmer (1998a) emphasized that the studies on ADHD were outcome studies and, as such, were indeed legitimate and sufficient to show efficacy for this type of treatment.  Othmer countered that the effect of smoking on health was established by outcome studies.  So were the effects of aspirin on heart attack incidence, and the asbestos link to cancer. He also countered that controlled studies of EEG neurofeedback have been done in the field of epilepsy in at least a dozen papers going back at least twenty years, and these studies all confirm the efficacy of this form of treatment.

With regard to the concern about a possible placebo effect, Othmer argues that children and their parents are blinded, in a fashion, to many of the positive results of neurofeedback training which occur in addition to improvement of their ADHD.  One of these results, for example, is the cessation of bedwetting.  It is a fortuitous improvement that occurs in many children undergoing EEG neurofeedback training.  Because it is totally unexpected by both parent and child, and occurs so often, it cannot have been the result of the placebo effect.

With Barkley’s concern about the possibility of a maturation effect, Othmer replies that maturation is what EEG training for children is all about.  He points out that the EEG’s of children with ADHD look like those of a younger child.  The training produces a more-mature EEG in a matter of weeks in a population where such results are improbable under any other circumstances.

In summarizing, Othmer points out that Barkley has never performed EEG neurofeedback and therefore is not in as good a position to judge as the practicing clinicians.  In closing his response to Barkley, Othmer quotes from others as follows; “This technique yields the most dramatic, documented results I have ever seen in over twenty years of working with learning disabled children.” (Cliford Marks, Ph.D, clinical psychologist); and “I’ve seen some remarkable results with EEG biofeedback with a number of children I have referred.  It has seemed particularly helpful with children who have speech, memory, or attention problems.” (Norma Schlager, Ph.D., educational therapist).  His final quote was from a mother who had read Barkley’s critique: “Too expensive? I would have sold my house for what you have done for my child.”  Nevertheless, Barkley’s criticisms need to be addressed in future studies if the use of neurotherapy is to gain credibility and acceptance in the scientific community.

Another area of controversy arises when researchers look at brainwave levels in various frequency bands before and after treatment.  In some instances changes can be identified, but in many instances no changes can be identified.  Because they see no change, the concern of some researchers is that any improvement noted is a result of factors other than neurofeedback.  Othmer (1998b) comments that it is acceptable when no trend changes can be noticed in the frequency bands during training. 

One of the reasons no change may be observed in filtered or power spectral density data, according to Bendat and Piersol (1971) is that these techniques have a limited ability to detect the presence of correlation, or low-level periodic components in the overall brainwave signal.  These authors show that the presence of such periodicities may be detected by visual inspection of a power spectral density function only if the filter bandwidth is sufficiently narrow, and the amplitude of the periodicity is sufficiently large, which may not be the case for the EEG. 

Another method offering better identification of periodicities, or correlation, is by use of the autocorrelation function, which is a very sensitive method for identifying low-level periodicities in a signal.  This is accomplished by mathematically eliminating random noise from the signal, leaving only the correlated parts for analysis.  This type of autocorrelation analysis is not normally performed on EEG neurofeedback data, and is an area that should be researched in the future.   It is important because it may show changes that are not apparent in the filtered or power spectral density plots normally used in neurofeedback.  Such changes would strengthen the claims made by neurofeedback practitioners about the efficacy of their treatments. 

To help define and clarify how power spectral density functions and autocorrelation functions are calculated, the following information was extracted from Bendat and Piersol (1971):

The power spectral density function is estimated by the following operations:
1.    Frequency filtering of the signal by a narrow band-pass filter having a bandwidth
       of Be Hz.
2.    Square of the instantaneous value over the sampling time.
3.    Averaging of the squared instantaneous value over the sampling time.
4.    Division of the mean square output by the bandwidth Be. As the center frequency
       of the narrow band-pass filter is moved, a plot of the power spectral density
       function versus frequency (power spectrum) is obtained.

The autocorrelation function is estimated by the following operations:
1.    Delaying the signal by a time displacement equal to ? seconds, called the lag time.
2.    Multiplying the signal value at any instant by the value that had occurred ? seconds before.
3.    Averaging the instantaneous product values over the sampling time.  (This, in effect, eliminates the random portions of the signal and leaves only the correlated part.)

As the lag time is moved, a plot of the autocorrelation function versus lag time (autocorrelogram) is obtained.

Othmer, Othmer, and Kaiser (1999) present another argument why frequency band levels may not show a permanent increasing or decreasing trend over time.  The reason is that the levels in any given frequency band may be entirely adequate under some conditions, but the ability of the band to change under other conditions may be limited.  It is a control system problem, and the exercise during a session enables the control system to maintain homeostasis more appropriately, upon demand.  Thus we may see increases during a session, but not long-term trend increases, with a net result however, of an improved management of the autonomic nervous system.

In summary, neurofeedback is not without controversy.  However, it has been shown by many to be successful in ameliorating a variety of disorders, both medical and psychological.  Many of the above referenced researchers have hypothesized that the success of this type of treatment may be the result of the neurofeedback training restoring or improving control to some CNS autonomic related functions that are not controlling properly.

Fibromyalgia

Fibromyalgia syndrome is the term used to describe a syndrome that includes widespread aching pain in muscles and connective tissue (ligaments and tendons), axial skeletal pain, and marked sensitivity to pressure at numerous “tender points.”  Although children occasionally get fibromyalgia, it is more common in young females, 45-55 years of age.  Adult women have this syndrome nine times more often than men, with approximately 2% of the population afflicted.  It is estimated that 5,000,000 Americans have fibromyalgia, and the cost to the U.S. economy in service utilization costs alone is $2274 per person per year in 1996 dollars according to Wolfe, et al. (1997a).  This amounts to more than 11 billion dollars per year, exclusive of lost work and other costs.

Diagnostic criteria for fibromyalgia, accepted by the American College of Rheumatology in 1990 (Wolfe et al. 1990), requires the presence of widespread pain in soft tissues both above and below the waist, and on both sides of the body.  In addition, axial skeletal pain (cervical spine or anterior chest or thoracic spine or low back) must be present.  A patient’s widespread pain must have lasted for longer than three months, and there should be pain in at least eleven of eighteen “tender points.”  Note that experts are not in agreement about the minimum number of points that must be tender for the diagnosis, and at any given time not all points may be particularly painful according to Wolfe, Hawley, Cathey, Caro, and Russell, (1985), and Bennett (1995).

The eighteen tender points are as follows:
•    Occiput: bilateral, at the suboccipital muscle insertions.
•    Low cervical: bilateral, at the anterior aspects of the intertransverse spaces at C5-C7.
•    Trapezius: bilateral, at the midpoint of the upper border.
•    Supraspinatus: bilateral, at origins, above the scapula spine near the medial border.
•    Second rib: bilateral, at the second costochondral junctions, just lateral to the junctions 
      on the upper surfaces.
•    Lateral epicondyle: bilateral, 2 cm distal to the epicondyles.
•    Gluteal: bilateral, in upper outer quadrants of buttocks in anterior fold of muscle.
•    Greater trochanter: bilateral, posterior to the trochanteric prominance.
•    Knee: bilateral, at the medial fat pad proximal to the joint line.

In non-technical terms these locations may be summarized as:
2 - neck, front
2 – neck, rear
2 – chest, upper
4 – shoulder, upper back
4 – hips, side, rear
2 – knees, inner
2 – elbow, side

The most common symptom of fibromyalgia is widespread total body pain in association with tender points.  It may be accompanied by chronic fatigue, disturbed nonrestorative sleep, significant neurocognitive problems such as confusion and memory lapses, morning stiffness, severe myogenic headaches, migraines, mood swings and other psychiatric disorders, hearing or vestibular abnormalities, sensations of pins and needles, burning and stabbing pains, carpal tunnel pains, irregular bowel habits (irritable bowel syndrome), frequent painful urination, allergic symptoms, alopecia, sun sensitivity, sicca syndrome symptoms, Raynauds phenomenon, and livedo reticularis.  Painful symptoms may be exacerbated by weather or temperature changes, fatigue or muscle overexertion, or emotional stress (Caro 1989).

The diagnosis and treatment of fibromyalgia has been hampered by the lack of a simple discriminatory objective test which may be applied to all patients with the diagnosis.  In search of such discriminants, Weigent, Bradley, Blalock and Alarcón (1998) reviewed most of the well-controlled studies looking for abnormalities in muscle tissue.  They concluded that there is no strong evidence that defects in muscle tissue are a primary cause of fibromyalgia. 

In the same article, Weigent et al. (1998) reviewed most of the studies looking for neuroendocrine abnormalities in fibromyalgia patients.  They reported that investigations of the hypothalamic-pituitary-adrenal (HPA) axis in fibromyalgia patients have produced evidence of adrenal insufficiency, suggesting that the resulting abnormalities in hormone levels may contribute to the many symptoms seen in these patients.  They also note however, that the exact interaction between the HPA axis and fibromyalgia is unknown.  Therefore, it is unknown whether the pain, fatigue, and other symptoms associated with fibromyalgia were a cause of, or a result of the adrenal insufficiency. 

In a related article, Demitrack and Crofford (1998) studied the HPA axis in fibromyalgia patients and suggest that this reduced HPA axis activity is due, in part, to impaired central nervous system drive.  Consistent with this, Griep et al. (1998), in a study comparing fibromyalgia patients and low-back-pain patients to controls, concluded that fibromyalgia patients had a dysregulated HPA axis, and this is due, in part, to a reduced containment of the stress-response system.  What is not clear from these studies is whether the observed dysregulation is a cause of, or the result of fibromyalgia.  It should be noted however that the hypothalamus is one area of the CNS involved in control of the ANS

Many fibromyalgia patients report symptoms that are indicative of primary sleep disorders.  These symptoms include insomnia, frequent awakenings during sleep, awakening feeling tired and not refreshed, and mood and cognitive disturbances.  In a review of the literature on sleep patterns of fibromyalgia patients, Harding (1998) found no consistent sleep abnormality that identified the fibromyalgia population.  Instead, numerous abnormal variables were identified that occurred in fibromyalgic patients and in some normals.  These included lower amounts of slow wave sleep, a higher number of arousals and awakenings, long awakenings (?10 minutes) and abnormal intrusion of alpha waves during delta wave sleep.  In general, these variables were found to respond to pharmacologic agents such as tricyclic antidepressants, selective serotonin reuptake inhibitors (SSRI’s), other antidepressants, chlorpromazine, and zopiclone, but the pain and chronic fatigue symptoms of fibromyalgia were not permanently eliminated with these medications.

In a review of the role of neurochemicals in fibromyalgia, Russell (1998) discussed the influence of serotonin, substance P (a chemical of eleven amino acids which has a potent effect on smooth muscle, and causes pain), nerve growth factor, and other neurochemicals in their possible role as causative agents for fibromyalgia.  Low serotonin levels have been found in many patients with fibromyalgia.  However comparable low levels have also been found in persons who do not exhibit the symptoms of fibromyalgia.  Likewise, high and normal levels of substance P have been found in the cerebral spinal fluid (CSF) of Fibromyalgic patients.  Thus the role of serotonin and substance P is uncertain in the pathogenesis of fibromyalgia.  Likewise, with nerve growth factor and other neurochemicals, the results are unclear.

Slotkoff and Clauw (1996) reviewed the research on cognitive dysfunctioning, which affects approximately two-thirds of fibromyalgia patients.  Their review covered the spectrum of memory loss, poor concentration, forgetfulness, and confusion.  They concluded that little is known about the cause of these dysfunctions, but possible contributing factors include fatigue, medications, mood disturbances, pain, and sleep disturbances.

Because fibromyalgia is a syndrome, which is the term applied to a collection of symptoms, it is not necessarily a specific disease entity.  Therefore, there may be multiple causes for this disorder, and as a result, the cause of pain in one patient may be different than the cause of pain in another.  These different causes therefore do not necessarily contradict each other.

Another proposed cause of fibromyalgia symptoms is dysregulation of the ANS.  This dysregulation theory arises, in part from studies such as those by Elam, Johansson, and Wallin (1992), Qiao, Vaerøy, and Mørkrid (1991), and van Denderan, Boersma, Zeinstra, Hollander, and van Neerbos (1992) who showed, via experiments utilizing the stressors of exercise, muscle contraction, and acoustic noise, that fibromyalgia patients have an impaired ability to respond to stressors compared to normals.  Clauw (1995) concluded that this is indicative of ANS malfunctioning.

Supporting this ANS malfunctioning theory, Clauw (1995), in reviewing the pathogenesis of fibromyalgia, covers the peripheral hypothesis (abnormalities in the areas of increased tenderness), the central hypothesis (abnormalities in the CNS appear likely to be responsible for the majority of fibromyalgia findings), sleep disturbances (disruptive sleep and alpha intrusions on the EEG during sleep), neurotransmitter abnormalities (such as inadequate amounts of serotonin, or excessive amounts of Substance P), autonomic nervous system abnormalities (“There is clearly dysregulation of the autonomic nervous system, although the precise nature of the defect is unclear.”), neuroendocrine abnormalities (hypothalamus-pituitary-adrenal (HPA)) axis hyporesponsiveness, and immune abnormalities (decreased natural killer cell (NK) function, altered IL-2 secretion, and increased deposition of immunoglobulin at the dermal-epidermal junction).

In another review of fibromyalgia, Wallace (1997) reports that although the existence of fibromyalgia has been controversial, nearly all Rheumatologists now accept is as a distinct diagnostic entity.  He reviews the literature on pain pathways in fibromyalgia and identifies elevated levels of Substance P in spinal fluid as being implicated in some patients, but not all.  Likewise, serotonin deficiency has been implicated in some fibromyalgia patients, but not all.  Additionally, his review concludes that muscle abnormalities play no role in fibromyalgia.

In reviewing the studies on ANS functioning in patients, he states “There is overwhelming evidence that dysautonomia is a prominent feature of fibromyalgia.”  (The ANS consists of the sympathetic and the parasympathetic nervous systems.)
In another review of fibromyalgia, Goldenberg (1996a) mentions the prevalence of fibromyalgia in the general population as being 2%.  He also suggests that fibromyalgia is not a specific disease, but best thought of as a “continuum of pain and tenderness in the population.”  However, he states that in spite of this, the diagnostic criteria of fibromyalgia is now accepted by the rheumatology community for classification purposed, even though they continue to be challenged by others as to whether they define a discrete clinical disorder.  This is an interesting and long-standing debate, as evidenced by the editorial by Bennett (1987) in the Journal of the American Medical Association, who argued that the disorder of fibromyalgia truly existed.

In his review article, Goldenberg (1996a) reviews studies of abnormalities in the CNS.  Of particular interest was a relatively small study that found that cortical regional cerebral blood flow was lower in fibromyalgia patients than in controls.  Also in his review of controlled therapeutic trials for fibromyalgia patients, he noted the lack of major improvements resulting from medicinal and non-medicinal intervention.

When assessing treatment efficacy, or severity of fibromyalgia, the measurement of morning stiffness is an important marker, even though it is a symptom of other disease processes, such as rheumatoid arthritis, as well.  Different researchers have used different criteria for measuring stiffness. Rhind, Unsworth, and Haslock (1987) performed a study of 100 patients with rheumatoid arthritis, and found that duration of morning stiffness was found to correlate only moderately well with severity of morning stiffness, and poorly with severity of stiffness at the time of interview.  In their study, they used the standard question: “How long did it take for your stiffness to begin to ease after you got out of bed this morning?”  In tabulating their results, they used an upper limit of 360 minutes for the duration of morning stiffness.  In discussing their results, they note that some who claim to be stiff are equally likely to be referring to pain, limited movement, or a combination of the two.  Thus the patient’s report of stiffness may not truly reflect the subjective stiffness of which the patient complains.

In a related article, Vlieland, Zwinderman, Breedveld, and Hazes (1997) described their studies of morning stiffness in rheumatoid arthritis patients.  In assessing stiffness duration, they used the standard question: “How long does your morning stiffness last from waking until maximum improvement occurs?”  They used an upper limit of 240 minutes in their study.  Severity of morning stiffness was determined via a visual analog scale, a 10-cm horizontal line with 0=none on the left, and 10=very severe on the right.  Their results showed that the measurement of morning stiffness by means of severity is more sensitive to changes over time than the measurement of its duration.

In another article on morning stiffness, Wolfe and Pincus (1995) use a standard question “How long is it until you are as limber as you will be for that day?”  The response is scored in minutes.  If the stiffness lasts for greater than 5 hours, the stiffness is scored as 301 minutes.  Thus different researchers use different upper limits in scoring stiffness duration.

The prognosis of fibromyalgia depends on patient selection and practitioner skill, with a general trend showing more success in a general practitioners office, and less in a specialized rheumatology office.  This appears to be the result of patients with a less severe form of the disease being successfully treated in the general practitioners office, with no need for a referral to a specialist.  Usually, only those cases that are refractory to treatment are referred to a specialist.  Thus the fibromyalgia patient population seen in a specialty rheumatology office usually has a more severe form of the disease.

MacFarlane et al. (1996) studied the patient population of 2 general practitioners in the UK.  After 2 years, only 35% of the fibromyalgia patients were symptomatic.  In another study in Canada, Crook, Weir, and Tunks (1989) showed that the percentage of fibromyalgia patients whose pain had resolved at 2 years was greater in family practices (36%) than in specialty pain clinics (13%).  And in children with fibromyalgia, the prognosis is favorable.  In a study by Buskila et al. (1995), it was shown that 73% of children resolve their symptoms within 2 years of diagnosis, and no longer meet the criteria for fibromyalgia. In another interesting study, Goldenberg (1996b) reports on a review of longitudinal studies of fibromyalgia.  His summary is as follows:

Author        Year    N    Duration(yr)    % improved    % worse
Felson        1986    39    3        ?        60
Ohghl        1990    58    3.3        31        68
Mallison    1992    28    2.2        ?        61
Nørregaard    1993    83    4        10        71
Ledingham    1993    72    4        26        74
Granges    1994    44    2        47 (not FM)    ?
Kennedy    1994    29    10        65        25

In a study performed by Wolfe et al. (1997b), the authors looked at the longitudinal outcome of fibromyalgia in six rheumatology clinics having a special interest in fibromyalgia. The study looked at measures of pain, global severity, fatigue, sleep disturbance, anxiety, depression, and health status over a seven-year period.  They concluded that for these patients, the measures were essentially unchanged over the study period.  In spite of these somewhat pessimistic results for the more seriously afflicted patient, it is important to note, as Goldenberg (1989) points out, that fibromyalgia is not life threatening and it does not, in general, degenerate over time into other life threatening diseases.  Also, cosmetic or structural abnormalities will not develop over time. 

Like other disorders, fibromyalgia is not without controversy.  St. Amand and Marek (1999), for example, claim success in treating fibromyalgia with guaifenesen, a widely available medication used in cold preparations, which is well tolerated and relatively inexpensive.  St. Amand has treated thousands of fibromyalgia patients over the years with guaifenesen, and claims to have seen fibromyalgia symptoms eliminated in 90 % of them.  Unfortunately, he has not published his results in a peer reviewed medical journal but has chosen, instead, to publish only in book form.  He is an endocrinologist and an assistant clinical professor of medicine at the UCLA School of Medicine, and has endorsements of his treatment approach, and his book, by some university associated physicians.  If his treatment method works as well as he claims, it should be published eventually in a peer reviewed journal.  In the meantime, the controversy continues. 

In summary, fibromyalgia is a chronic syndrome with many abnormalities having been found.  These abnormalities have pointed in part to prominent central nervous system (CNS) components to this disorder. Thus neurofeedback training is a reasonable treatment approach to try because it appears to work by restoring proper control to portions of the CNS that are not controlling properly.

TOVA – Measures of Attention
The TOVA (Test of Variables of Attention) test will be utilized in this study to aid in the measurement of CNS attention related difficulties that may be experienced by fibromyalgia patients.  This is a non-language-based test that requires a subject to pay continuous attention to two non-verbal stimuli for 21.6 minutes without interruption or rest.  During this time the subject responds either to visual target stimuli on a computer screen or to auditory target stimuli played through computer controlled loudspeakers.  The method of target and non-target presentation, coupled with the responses of the subject, provides measures of inattention, impulsivity, response time, and response time variability.

TOVA is a continuous performance test (CPT).  Various other CPT attention-testing systems have been used.  For example, O’Dougherty, Neuchterlein, and Drew (1984) use signal detection methods to estimate attention sensitivity and bias.  Halperin, Wolf, Greenblatt, and Young (1991) estimate attention related impulsivity by means of error analysis.  Cornblatt, Risch, Faris, Friedman, and Erlenmeyer-Kimling (1988) use identical pairs identification within verbal and non-verbal stimuli.  TOVA measures attentional and impulse control processes with either auditory of visual stimuli.

The TOVA tests consist of two versions, the TOVA-auditory, and the TOVA-visual.  During the visual version, two easily discriminated geometric pictures are shown on a computer screen.  During the auditory TOVA, the stimuli are two easily discriminated audible tones played through computer controlled loudspeakers (Middle G and Middle C.)

During the test, one of two stimuli, either the designated target or the nondesignated target, is presented for 100 milliseconds every 2 seconds.  The test subject is instructed to respond to the designated target only, by pressing a switch as quickly as possible each time the designated target is presented.  The designated target is presented 22.5% during the first half of the test and 77.5% during the second half.

The test software records the subject’s responses with one-millisecond accuracy.  From this, four variables are determined: (a) errors of omission or number of target stimuli missed (inattention); (b) errors of commission or number of responses following incorrect stimuli (impulsivity); (c) mean correct response time (response speed); and (d) response time variability (consistency of attention).  In addition, a D Prime score is presented which is a measure of performance deterioration over time.  It helps distinguish impaired from non-impaired individuals. 

The TOVA test was originally developed as an aid in the diagnosis of ADHD.  For that purpose, an ADHD score was developed which utilized the response measures described above.  Additionally, validity and reliability measures were determined regarding its’ use as a diagnostic aid for ADHD.  A detailed description for that application is presented in Leark, Dupuy, Greenberg, Corman, and Kindschi (1996).
Note that the validity and reliability measures described in that document are not applicable to this case because the TOVA test is not being used to diagnose ADHD in this application.  Instead, the TOVA auditory and visual tests are being used simply to provide the direct response measures of errors of omission, errors of commission, response time, and response time variability. 

Measurements of these variables were made with the TOVA tests prior to start of neurofeedback training, and after completion of every ten neurofeedback training sessions.  The purpose was to see if these measures changed over time.  It is believed that any changes that would be observed would be the result of neurofeedback training because there are no appreciable test-retest practice effects, according to Greenberg and Kindschi (1996).  They reported that there was a nonsignificant tendency for commission errors to decrease (improve) during the first half of the test from the first test to the second test, but not for subsequent tests.  There were no other significant differences on the other variables.

Summary of Literature Review

The review on fibromyalgia covered several studies implicating autonomic nervous system dysregulation.  These include studies by Elam Johansson, and Wallin (1992), Qiao, Vaerøy, and Mørkind (1991), and van Denderan, Boersma, Zeinstra, Hollander, and van Neerbos (1992).  Also, Clauw (1995) and Wallace (1997) state that the evidence for dysregulation of the ANS is overwhelming. 

The ANS is part of the peripheral nervous system, but is under the direct control of the cerebral cortex, the hypothalamus, and the medulla, all within the central nervous system.  Thus attempts to restore proper regulation to the ANS might reasonably focus on the CNS.

The review of the literature on EEG neurofeedback gave several illustrations of how this type of therapy appears to work by impacting the central nervous system.  Specifically, the works of Sterman and Friar (1972), Sterman (1973), Seifert and Lubar (11975), and Lubar and Bahler (1976) found that training to enhance the amplitude of the SMR over the sensorimotor cortex was effective in seizure reduction for epileptics.

Tansey and Brunner (1983) were among the first to effectively treat ADHD and related disorders by training to enhance SMR.  This protocol, with modifications to accommodate individual differences, is now used in over 300 organizations to treat ADHD, according to Lubar (1995), another pioneer in the field.

In another study, Tansey (1986) reported the successful elimination of tics using the SMR enhancement protocol.  Rozelle and Budzynski (1995) also were successful with SMR enhancement in treating a stroke victim.  Nahmias, Tansey, and Karetsky (1994) likewise reversed symptoms of extrathoracic airway obstruction with SMR enhancement.  This same protocol is also used routinely in performance enhancement training according to Norris and Currieri (1999).

Thus training to enhance SMR over the sensorimotor cortex appears to affect the CNS and result in positive outcomes in numerous disorders.  Therefore this neurofeedback training protocol might reasonably be expected to result in positive outcomes for some patients with fibromyalgia.

CHAPTER III
METHODOLOGY

Research Design

This study was performed in a clinical setting.  In such settings, controlling for all possible variables is impossible.  Furthermore, chronic diseases such as fibromyalgia present with a complex symptomatology and are almost always accompanied by a host of other disorders.  Notwithstanding these limitations, if the results of this preliminary study appear promising, then a more controlled follow-on study would be justified.

This study involved an evaluation of: (1) subjective data and comments from the patient and the examining physician; and (2) objective TOVA? test data taken periodically as neurotherapy progressed.  The patient population consisted of all fibromyalgia patients seen between September, 1996 and August, 1999 in a specialty rheumatology practice who completed 40 or more sessions of neurofeedback training for their fibromyalgia.  Neurofeedback training was given only after other medical interventions had failed.

The initial sample size of patients that underwent more than 40 sessions of neurotherapy was 21.  From this sample, two were excluded because they did not meet the clinical diagnosis criteria of fibromyalgia, but instead were being treated for attention deficit disorder.  Two were excluded because of evidence of dementia or prior stroke, one was excluded because other psychological testing indicated the strong possibility of malingering, and one was excluded because treatment was not continuous over time.  This left a sample size of 15.  Of this sample, 14 were women. 

All these remaining patients had a diagnosis of fibromyalgia and an average of 3-4 other diagnoses, which were confounds in this study.  These other diagnoses were standard co-morbid symptoms of many fibromyalgia patients seen in a Rheumatologist’s office, and included Sicca Syndrome, Sjogrens syndrome, Irritable Bowel Syndrome, Osteo Arthritis, Rheumatoid Arthritis, Chronic Fatigue Syndrome, Neurocognitive Syndrome, Sleep Disorders, Osteoporosis, Crest Syndrome, Low-grade Polyarthritis, Raynauds Phenomenon, Myofascial Pain Syndrome, and Rotator Cuff Tears.  In addition to the use of neurofeedback training, all diagnoses were treated with appropriate medical care.

Results from this patient population were compared to another group of 63 patients with fibromyalgia that did not receive neurotherapy.  Patients in this other group were examined and treated in the same Los Angeles office by the same examining physician before neurotherapy became available in that office and were part of a larger study monitored by the Wichita Arthritis Center (Wolfe, et al. 1997b).  Thus both groups received the same basic treatment, with the exception of neurotherapy.  However, because the two groups received basic medical support therapy during different time periods, this is a limitation of this study.

Fibromyalgia was defined using the guidelines established by The American College of Rheumatology (Wolfe, et al. 1990).  This required the presence of widespread pain in soft tissue both above and below the waist, and on both sides of the body.  In addition, axial skeletal pain must be present.  Pain must have lasted for longer than three months, and there should be pain in at least eleven of eighteen “tender points.” 
Over a 4.6-year period, the Los Angeles office non-neurotherapy treatment group experienced a 6% reduction in pain, fatigue was reduced 5%, anxiety increased 4%, and depression increased 3%.  These results are the therapeutic baselines used for this study.

Note that the Wichita Arthritis Center studied a total of 538 fibromyalgia patients.  Combined results for all 538 patients showed no change in pain (0%), no change in fatigue (0%), a decrease of 2% in anxiety, and an increase of 3% in depression.  These published results were slightly worse than in the Los Angeles office. 

Neurofeedback Training

The neurofeedback training on these patients was performed with the EEG Spectrum Neurocybernetics neurofeedback system manufactured by EEG Spectrum, Inc., Encino, California.  The training protocol consisted of the following:
1.    A sensing electrode was placed at the Cz position (American Electroencephalographic Society, 1991).  This is the topmost position on the midline of the head, halfway between the inion (just underneath the occipital condyle, the bump at the back of the head) and the nasion (the top of the bridge of the nose where the forehead is indented).
2.    Ground and reference electrodes were placed on the lobe of opposite ears.
3.    Signals from the electrodes were amplified and fed to a computer program that displayed the brainwave in amplitude time-history format on a computer screen viewed only by the therapist.  The unfiltered time-history signal was displayed, as well as three filtered time-history segments.
4.    The three filtered time-history segments were chosen to show the Theta (4-7 Hz) band levels, sensorimotor rhythm (SMR, 12-15 Hz) band levels, and high Beta (22-30 Hz) band levels.
5.    The segments were chosen so that the patient would try to augment (reward) SMR, with concurrent inhibition of Theta and high Beta via a separate computer screen viewed by both patient and therapist.
6.    A criterion was established by the therapist for each brainwave frequency band of concern.  For the reward band, the patient would attempt to keep his/her brainwave level above the established criterion.  For the two inhibit bands, the patient would attempt to keep his/her brainwave level below the established criteria.
7.    Augmentation and inhibition of these frequency bands was facilitated via a computerized game seen on the patient’s computer screen.  Control of the game was via brainwaves.  One of the games was a maze similar to PACMAN, where movement and brightness of the maze object was determined by the brainwaves.  The patient was rewarded by motion of the maze object, by the playing of a tone, and by the accumulation of game points whenever the three chosen frequency band levels were within the criteria specified by the therapist.  These criteria levels were changed periodically by the therapist to accommodate changes in the brainwave levels during a session, or from session to session.

The Cz location, directly over the sensorimotor (Rolandic) cortex, was chosen because: (1) many studies have shown this location to result in optimum improvement in other disorders; (2) training at this location might result in changes in areas of the CNS which might be involved in regulation of the ANS, and dysregulation of the ANS has been proposed as a causal agent in fibromyalgia; and  (3) the previous undocumented experience byDr.Caro while investigating the efficacy of other electrode placement locations showed that reward of the SMR frequency band at the Cz location, with concurrent inhibition of the Theta and high Beta frequency bands, usually resulted in optimum improvement in his fibromyalgia patients.
High amplitude Theta levels are also indicative of insufficient cortical control, and can be the result of fatigue, brain injury, emotional memories, or other biochemical or physiological disorders.  This can result in poor vigilance and attentiveness, general low arousal states, seizures, mood swings, or dissociative states.  Training attempts to decrease these levels if they are excessively high.

High amplitude high Beta levels are correlated with muscle tension or over-arousal, and may also reflect brain irritability or chemical sensitivity.  Training attempts to decrease these levels if they are excessively high.

Low amplitude SMR waves are correlated with poor CNS control, fatigue, inattention, or over-efforting.  As the SMR amplitude is increased, patients report an increase in feeling calm and focused, and improvement in mental and physical functioning.  Training attempts to increase these levels for all patients.

Keeping the above in mind, one might expect that changes in levels of the various frequency bands during neurofeedback training might result in permanent changes to these levels away from training.  This, however, was not the case.  Sometimes there are permanent changes, and sometimes not.  What is important is that the training is more properly seen as an exercise in strengthening the mechanisms in the brain which regulate and accommodate change, homeostasis, arousal, attention, and vigilance.

The amplitude of a given frequency band may be entirely adequate and appropriate for an individual under some conditions.  However, the ability of that particular band to change amplitude in response to changing external demands may be limited.  The exercise during a session will enable that particular band to change levels more appropriately at other times, upon demand.  Thus one does not necessarily expect to see permanent increases in amplitude when one trains to increase SMR during a session.

At each neurotherapy session, a log was made of patient self-reports of their subjective evaluations of pain, and anxiety/mood/depression. The verbal questions to the patients were phrased as follows: “For today, how would you rate your pain, on a scale of zero to ten, where zero is no pain, and 10 is maximum pain?”  The responses to each question were noted by the neurofeedback therapist in the patient’s chart on a form specifically designed to record pertinent data for each neurotherapy session.  Included on this chart was information about electrode placement, training protocol, changes during the session, duration of session, number of session, date, time of day, points earned by patient during training, patient comments, therapist comments, and anything else that might relate to that particular neurotherapy session.  No other data collection instruments were utilized at a session.  If the patient reported any unusual symptoms, the physician was notified before proceeding with the neurotherapy.

Physician Observations

Patients saw the physician prior to the start of neurofeedback training and periodically thereafter to evaluate treatment progress.  During these examinations, which covered all medical problems, the physician made an estimate of fibromyalgia pain levels experienced by the patient in both the upper and lower torso, averaged over the fibromyalgia tender points, and questioned the patient regarding the duration of their stiffness.  This information, as well as the physician’s general impression of patient well-being and treatment progress, was noted in the patient’s chart.  There were no rigorous criteria used in quantifying patient improvement and well-being, but the same physician made all estimates on all patients for this group. 
This physician has approximately 20 years experience in rheumatology and has a special interest in fibromyalgia.  In addition, he has authored or co-authored several publications in the medical literature on fibromyalgia.  Thus his estimates of the fibromyalgia related pain, well-being, and treatment progress in his patients are assumed to be reliable, accurate, and consistent.  No other instruments were used to acquire this data. 

TOVA TESTING – Measures of Attention

The TOVA (Test of Variables of Attention) test was utilized in this study to aid in the measurement of CNS attention related difficulties that may be experienced by fibromyalgia patients.  TOVA is a non-language-based test that requires a subject to pay continuous attention to two non-verbal stimuli for 21.6 minutes without interruption or rest.  During this time, the subject responds either to visual target stimuli on a computer screen or to auditory target stimuli played through computer-controlled loudspeakers.  The method of target and non-target presentation, coupled with the responses of the subject, provides measures of inattention, impulsivity, response time, and response time variability.
The TOVA tests consist of two versions, the TOVA auditory, and the TOVA visual.  During the visual version, two easily discriminated geometric pictures are shown on a computer screen.  During the auditory TOVA, the stimuli are two easily discriminated audible tones played through computer-controlled loudspeakers (Middle G and Middle C.)

During the test, one of two stimuli, either the designated target or the nondesignated target, is presented for 100 milliseconds every 2 seconds.  The test subject is instructed to respond to the designated target only, by pressing a switch as quickly as possible each time the designated target is presented.  The designated target is presented 22.5% during the first half of the test and 77.5% during the second half.
The test software records the subject’s responses with one-millisecond accuracy.  From this, four variables are determined: (a) errors of omission or number of target stimuli missed (inattention); (b) errors of commission or number of responses following incorrect stimuli (impulsivity); (c) mean correct response time (response speed); and (d) response time variability (consistency of attention).  In addition, a D prime score and an ADHD score is presented.  These scores help distinguish impaired from non-impaired individuals.


CHAPTER IV

FINDINGS

Overview

This chapter presents the data obtained, and analyses performed to determine if neurotherapy is of value in treating patients with fibromyalgia, and if the TOVA? test can be of use in monitoring treatment progress.  To determine if neurotherapy is of value, the results of this patient group are compared to another therapeutic baseline group that did not receive neurotherapy.  The data and analyses show that neurotherapy provides significant symptom reduction for global pain, fibromyalgia pain, and fatigue compared to the non-neurotherapy group.  It also shows that the TOVA? visual and auditory tests, as currently administered and interpreted, do not provide a reliable monitor of neurotherapy treatment progress for fibromyalgia.
Reasons why the neurotherapy group had a better therapeutic outcome than the non-neurotherapy group are not known.  The data collected do not reveal changes that can be ascribed to causal agents, and no consistent brainwave changes or trends were observed.  Possible primary contributors to this lack of observed change are:

1.    The choice of measurement location for the sensing electrodes might have been incorrect for detecting brainwave changes that occurred. 
2.    Methods of data analysis might have been inappropriate to detect the actual brainwave changes that occurred.  For example, changes might have been in signal phase properties as opposed to signal amplitude properties.  If that were the case, auto-correlation function analyses would show phase changes and therefore be more appropriate than the filtering analyses that were used.  Correlation function analyses are not normally performed on this type of data.
3.     Methods of data analysis might have been too crude or incorrect to show changes that might have actually occurred.  Filter bandwidths and sampling times are important factors that affect the ability to detect the presence of small signal properties that may be buried within data.  The software used in this study was that supplied by the manufacturer of the neurotherapy equipment and the data analyses were performed using factory settings.
4.    Brainwave levels may be entirely adequate under some circumstances and not under others.  Their ability to change amplitude according to need may have been restricted.  The neurotherapy might have enabled the homeostasis system to respond more appropriately upon demand, while exhibiting no change during a neurotherapy session, compared to other neurotherapy sessions.  Thus no change would have been observed from session to session even though appropriate changes might have occurred upon demand at other times.  This would, in effect, be a re-normalization of appropriate neural control systems, and one might not expect such re-normaliztions to show up in neurotherapy session results, only upon other special “demand” situations external to the neurotherapy session.

Findings

Patient Comments Regarding Improvement

The neurofeedback technician solicited comments from patients during neurofeedback therapy sessions, and the examining physician solicited comments from the patients during periodic office visits.  No quantifiable criteria were established for measuring these patient comments regarding their feelings of improvement.  Thus they have limited objective scientific basis. The comments are summarized below:
Patient Number    Patient comments
01    Improved.  Residual pain primarily in the hips, legs, and thumbs after 25 sessions.  Increased energy, doing better overall, better sleep, and thinking was clearer.
02    Improved.  Pain diminished in the back and neck.  Patient credits neurofeedback.  Hand pain was the primary remaining problem, but it was also diminished.
03    Improved, but only after completion of approximately 60 neurofeedback sessions.  Switched to alpha training at 83 sessions, and commented at completion of 90 sessions that she hadn’t felt this good in years.  Felt as if she had gone into remission.
04    Improved.  Felt better than ever, but then developed an unrelated medical problem after 25 sessions of neurofeedback.  This required hospitalization and resulted in a return of her previous symptoms.  After release from the hospital, patient resumed neurofeedback.  At completion, patient felt good enough to return to work after being on disability for 6 months.  Patient credited neurofeedback for the improvement.
05    Improved.  Initially there were fluctuations in how the patient felt, and the patient reported no specific changes in pain, mood, or fatigue, but rather an overall increased feeling of well being.  Later, with more neurofeedback sessions, the patient reported significant improvement for the first time in many years.  Pain was approximately 30% improved, much less morning stiffness, better sleep quality, and decreased fatigue.  Also patient reported a steady improvement in neurocognitive function, with better and clearer thinking.
06    Improved.  Patient reported that her pain and fatigue fluctuated.  Patient later estimated that she was at least 90% improved in her musculoskeletal pain, and had an increased freedom of movement.
07    Improved.  Patient reported an improvement in overall sense of well-being.  Also noticed an improvement in memory, and feeling more alert.  Pain was diminished slightly, and also slightly less fatigue.
08    Improved.  Patient reported that pain and fatigue were less, and she was sleeping much better.
09    Improved.  Patient noticed a constant improvement after starting neurofeedback, and had virtually no pain after the 21st session.  Residual stiffness was only in the hands.  Later, at one session, the patient reported pain from a smashed finger, which affected his neurofeedback, but only for that session.
10    Little To No Improvement.  Patient reported minimal change in fibromyalgia symptoms.
11    Improved.  After completion of neurotherapy, patient was able to walk without her cane.  She stated she was moving about better than before, and had more energy.
12    Improved.  Patient felt more calm and said her mind was not racing as before.  Pain level was significantly lower, and she was experiencing an overall feeling of well-being.  Sleep was a continuing problem.
13    Improved.  Patient’s mood improved, and she reported she no longer felt stressed at home.  Also, her pain and fatigue were less.
14    Improved.  Greatly improved energy and reduced pain.  Felt much better.  Recalled how she had to “crawl to the office” when she first started neurotherapy.
15    Improved.  Patient started to sleep better and have less pain after approximately 10 sessions.  Noticed less fatigue and more energy after approximately 20 sessions.  Patient still had periods of severe localized pain.  At 40 sessions, patient was still having localized pain, but fibromyalgia pain was much less intense.  Sleeping better, feeling better, relaxed and more alert.  Still had flare-ups, but they were primarily localized.  At 60 sessions, patient stated that her overall pain was much less than when she started neurofeedback.

In summary, 14 of the 15 patients (93%) in this study reported feelings of improvement after their neurofeedback training sessions.  The one patient that did not report feelings of improvement was a female, age 66, with fibromyalgia and accompanying disorders similar to the other participants in this study. 

Symptom Scores for Patients Receiving Neurotherapy
All data presented in this section are based upon the complete set of patient data obtained during this study, and presented in the appendix as follows:

•    Appendix A: Neurofeedback Session Patient Data
•    Appendix B: Physician Observations
•    Appendix C: TOVA Raw Scores
•    Appendix D: TOVA Standard Scores

Table 1 presents the global pain scores (range 0-10, 0=no pain, 10=maximum pain) reported to the neurotherapy technician at 0, 10, 20, 30, and 40 sessions of neurotherapy, and at completion.  Also shown is the number of neurotherapy sessions to completion for each patient.  This number is the same for all symptoms, so will not be repeated on the other tables.

Tables omitted 

Table 2 presents the fibromyalgia pain scores (range 0-3) as determined by the examining physician at 0, 10, 20, 30, and 40 sessions of neurotherapy, and at completion.  These scores are the average between the upper and lower torso scores.  These scores were not always obtained at the exact 10 session intervals.  Thus scores at the 10 session intervals were sometimes estimated from data trends and scores immediately before and after the 10 session intervals.  No scores are listed when a score was obtained at a session number significantly different from the 10-session intervals.

Tables 3, 4, and 5 below present the fatigue scores (range 0-10), mood/depression scores (range 0-10), and stiffness durations (range 0-360 minutes) reported to the neurotherapy technician at 0, 10, 20, 30, and 40 sessions of neurotherapy, and at completion

Table 3  Fatigue Scores With Neurotherapy

Table 4  Mood/Depression Scores With Neurotherapy

Table 5  Stiffness Duration With Neurotherapy - Minutes

Symptom Mean Score Changes for Patients Receiving Neurotherapy

The number of neurotherapy sessions required to complete therapy for this group of patients ranged from 40-98, with an average of 58 sessions.  The symptom mean scores and standard deviations at the start of neurotherapy, every ten sessions up to 40, and at completion are shown in Table 6.  Below each symptom mean score is the Wilcoxon calculated Z score of statistical significance.  Calculations of Z scores are made every 10 sessions, up to 40 sessions, and at completion so that data trends, if they exist, can be detected. 

The Wilcoxon test is a non-parametric test for determining the significance of the difference between the distributions of two correlated small samples involving repeated measures. 

Also in Table 6, an asterisk is shown whenever the Z score shows statistical significance at the one-tailed 0.05 level (p=0.05), or greater.  This corresponds to a requirement for the Z score to be +/- 1.645 or greater in amplitude.  As the magnitude of the Z score increases, statistical significance also increases.  Thus a Z score of 1.960 corresponds to a one-tailed level of significance of 0.025, a Z score of 2.326 corresponds to a one-tailed level of significance  of 0.01, and a Z score of 2.576 corresponds to a one-tailed level of significance of 0.005. 

Table 6  Symptom Mean Scores and Z Scores For Patients Receiving Neurotherapy

Because of the small sample size for this study (n=15), a power analysis was conducted to determine the probability of correctly rejecting a false hypothesis.  For this study, the power was calculated comparing baseline to completion.  Results are given in Table 7 below. 

Table 7  Power Analysis Results: Baseline Compared to Completion

As shown, fibromyalgia symptom scores have the greatest power, global pain and fatigue have less power, and mood/depression and stiffness have very little power.  These scores are consistent with the Z scores for these symptoms, shown in Table 6 above.

Table 8 below presents the overall percentage reductions in symptom mean scale values, and the results of the Wilcoxon test of the statistical significance between symptom scores obtained prior to starting the neurotherapy program (baseline), and at completion Table 8  Mean Symptom Score Reduction With Neurotherapy

Symptom Mean Score Changes For Patients Not Receiving Neurotherapy

The therapeutic baseline group consists of a group of fibromyalgia patients who were studied by Wolfe, et al (1997) at the Wichita Arthritis Center.  All of these patients received standard medical care for their disorder, and none received neurotherapy.  There were a total of 538 patients from 6 rheumatology centers in the United States who participated in that study.  One of those centers (Los Angeles) was that ofDr.Caro, who provided 63 fibromyalgia patients to that study.  His office also provided all the patients for this fibromyalgia/neurofeedback study.  Table 9 below shows the mean pre- and post-scores and mean of all scores for applicable symptoms at each participating center, including the Los Angeles center.  Table 10 shows the percent score changes 

For all patients in that study, the median duration of their fibromyalgia at first assessment was 7.8-years, and the final assessment took place after 7 years.  For the LosAngeles office, the median follow-up time in the study was 4.6 years. 

Table 9  Symptom Mean Score Changes For Patients Not Receiving Neurotherapy

Table 10  Symptom Percentage Score Changes For Patients Not Receiving Neurotherapy

TOVA Test Results and Mean Score Changes With Neurotherapy

TOVA visual (V) and auditory (A) tests were administered prior to starting the  neurofeedback therapy program and then re-administered after completion of every ten neurofeedback sessions.  They were performed to determine if these tests could be used as a reliable monitor of treatment progress.  Raw data from these tests, and standard scores for all patients, are presented in the appendix. 

Table 11 shows the overall mean percent reductions of the TOVA scores after completion of all neurotherapy sessions.  These reductions ranged from 0 to 64 percent.  The TOVA auditory test does not currently calculate an ADHD score, so it is not included in the table.  Also, ADHD scores are calculated from other test data, so percent changes in this score are not included here because they are not a meaningful indicator of percentage change. 

The number of sessions to completion for each patient varied from 40-98.  Improvement in mean score changes were calculated based upon individual raw score changes, whereas the Wilcoxon signed rank test was calculated using standard scores, which eliminated age and gender effects in the data. 
Table 11  TOVA Score Changes With Neurofeedback

Correlation of TOVA scores to Fibromyalgia Symptom Scores

To determine if the TOVA visual test could be used as a monitor of neurotherapy treatment of fibromyalgia, correlations were calculated between the three most statistically significant patient symptom mean score changes and the three most statistically significant mean TOVA visual scores.  Results are given in Table 12. 
Table 12  Correlation Between TOVA Sub-score Means and Patient Symptom Means

The correlation between mean fibromyalgia pain scores and mean TOVA Commission errors standard scores was –0.85, which indicated that this TOVA sub-score might be of value in tracking neurotherapy treatment progress for fibromyalgia patients.  To investigate further, individual correlations between this sub-score and fibromyalgia average pain were calculated for each patient.  Results are given in Table 13 below.
Table 13 Correlation Between Individual TOVA Visual Test Commission Errors Standard Score and Fibromyalgia Pain Scores

Power, and other statistical analyses were not performed on TOVA scores because they were subsequently determined to have minimal value in tracking neurofeedback therapy progress, even though they provide good measures of attention.

ANALYSIS AND EVALUATION OF FINDINGS

Patient Comments Regarding Improvement

Patient comments regarding their improvement were obtained as part of the unstructured, open-ended interviews held with each participant during their neurotherapy treatment program.  Purpose of these interviews was to provide an opportunity for each participant to describe, explain, and question factors regarding their neurotherapy treatment and progress, and to provide information to the neurofeedback therapist and physician regarding treatment progress and problems.

Their comments are of great interest and importance during the course of the study because they help validate, or invalidate, the process as treatment progresses.  However, such comments are not amenable to unequivocal interpretation because of the lack of scientific rigor in acquiring such information, and there were no rigorous quantifiable criteria established for measuring patient comments regarding their feelings of improvement..  Also, there are several conflicting reasons why participants may report feelings of improvement.  These include:
1.    They are truly feeling improved.
2.    They do not wish to confront their provider with negative results.
3.    They may falsely claim improvement to more easily justify termination of a relatively expensive treatment program.
4.    Some patients report feeling improved while still maintaining that their pain has not diminished.  What has apparently changed is their reaction to, or perception of their pain.  Thus improvement can refer not just to pain, but also to other subtle factors.

It is reasonable to suspect that all 4 reasons cited above, plus others, played some role in influencing patient comments.  Nevertheless, reports from 14 of the 15 patients indicating that they felt improvement (93%) was sufficient motivation to keep this study progressing.  The one patient that did not report improvement was a 66-year-old female with fibromyalgia and other accompanying disorders similar to the other patients in this study.

Global Pain

At the start of each neurotherapy session the patient was asked to rate their overall pain on a scale from 0 to 10, with 0 representing no pain, and 10 representing maximum pain.  For these ratings, no discrimination was made between localized joint pain and diffuse pain more representative of fibromyalgia.  For this study, their overall pain rating is called “global pain.”

There is scatter in the data, most likely caused by normal fluctuations in pain symptoms that occur naturally from the other disorders the patients have.  Nevertheless, the data in Tables 1 and 6 show a trend of decreasing global pain score as the patients complete neurotherapy sessions. 

At the beginning of therapy, the patients had global pain levels ranging from 0- 10 on the ten-point scale, with an average of 4.6.  At the completion of therapy, the range was 0-7, with an average global pain level of 2.8.    Thus the range decreased from 0-10 to 0-7, and the average global pain decreased from 4.6 to 2.8.  This is a 39% reduction, considering all patients.  The number of neurofeedback sessions required to complete therapy for this group ranged from 40 –90 sessions, with an average of 58 sessions. 

Of these 15 patients, three reported an increase in their global pain level by one point each at the completion of therapy (5?6, 0?1, and 1?2), two reported the same global pain level at completion (2?2, and 3?3), and the remaining ten patients all reported a decrease in global pain an average of 3 points each (range of 1-7) for an average reduction of 52%.  In general, the greatest decreases in global pain intensity were reported by those patients starting neurotherapy with the most severe pain.  The global pain scores at the start of neurotherapy are different from the completion scores at a one-tailed significance level p<0.05, per the Wilcoxon test.
Note that in Table 6, the Z scores for global pain show an increasing trend, and the mean scores for pain show a decreasing trend as the number of neurotherapy sessions increases.  One can infer from these trends that the scores at completion compared to the scores at baseline, are not different because of random fluctuations in the data or processes, but instead, are the result of a systematic process.  This strengthens the argument that neurotherapy is the effective process.

In this study, one would not expect to see perfect trend lines in the pain scores or Z scores because of the many confounds affecting the data.  For example, a few patients having significant flare-ups in symptoms from their other disorders can result in significant increases in group-mean-pain scores and significant decreases in the resulting Z scores.  Such fluctuations are seen in the data of Table 6, however the trends are clear. 

Fibromyalgia Pain

The physician would make estimates of average fibromyalgia pain in the upper and lower torso during each examination meeting.  These estimates were made on a scale of 0-3, with 0 representing no fibromyalgia pain, and 3 representing maximum fibromyalgia pain.  This estimate was made independently of other pain the patient may have been experiencing from other disorders.  The fibromyalgia pain reported in this study, and shown in Tables 2 and 6, was the average of these upper and lower torso estimates.

At the beginning of therapy, the patients had torso average fibromyalgia pain levels ranging from 0-3, with a group average of 1.4.  At the completion of therapy, these averages were reduced to 0.3, which is an overall reduction of 74% in fibromyalgia pain score.

Of these patients, none had worse fibromyalgia pain scores at completion, 11 were improved, and 4 showed no change.  For those with no change, three had 0?0 scores and one had 1?1 scores.  The pre- and post-test results are different at a one-tailed significance level p< 0.05, per the Wilcoxon test.

Note that in Table 6, the Z scores show an increasing trend and the fibromyalgia pain scores show a decreasing trend as the number of neurotherapy sessions increases.  This strengthens the argument that neurotherapy is the effective process.
Fatigue
At the start of each neurotherapy session, patients were also asked to rate their level of fatigue on a scale from 0 to 10, with 0 representing no fatigue, and 10 representing maximum fatigue.  The data are shown in Tables 3 and 6.  As with the pain scale, there is a large amount of scatter in the data, but there is also a trend towards decreasing fatigue as neurotherapy progresses.

At the beginning of therapy, patients had fatigue level scores ranging from 0-10, with an average of 5.5, and at completion of therapy, the range was 1-8, with an average of 3.3.  Thus the range changed from 0-10 to 1-8, which is an improvement, and the average fatigue score decreased from 5.5 to 3.3, which is an overall reduction of 40% in fatigue score for all patients.

Of these 15 patients, two reported an increase in fatigue by two points each at the completion of therapy, one patient reported the same fatigue level at completion of therapy (2?2), and the remaining 12 patients all reported a decrease in fatigue an average of 3.1 points each (range of 1-7), or an average reduction of 48% for the patients that improved.  As with global pain, the greatest decreases in fatigue score were reported by those starting neurotherapy with the greatest amount of fatigue.  The differences between the start and completion of neurotherapy were significant at a one-tailed significance level p< 0.05 per the Wilcoxon test.
Table 6 shows the trends in the scores, with the fatigue score decreasing and Z score increasing as the number of neurotherapy sessions increases.  From these trends one can infer that neurotherapy is the effective process affecting these variables. 

Mood/Depression

As with pain and fatigue, patients were asked at the start of each neurotherapy session to rate their mood or depression on a scale of 0-10, with 0 representing no mood or depression, and 10 representing maximum.  This data is shown if Tables 4 and 6.

 Prior to starting their neurofeedback program, patients had ratings on this scale ranging from 0-7, with an average of 3.7.  At completion, the range was 1-7, and the average was 2.7.  Thus the range did not change appreciably, but the overall average for all patients decreased from 3.7 to 2.7, which is an average mood/depression reduction of 27%.

Of the 15 patients, three reported an increase (0?3, 0?2, and 6?7), 5 reported the same level at completion (5?5, 2?2, 2?2, 3?3, and 0?0), and seven reported an improvement in their mood/depression an average of 2.9 points, and a range of 1-5 points, or an average reduction in mood/depression of 57% for the group of patients that improved.  The Wilcoxon test of statistical significance for this symptom showed the results to be non-significant.  Thus one cannot conclude that there was a significant difference between the start and completion groups for this symptom.

Looking at the Z scores for mood/depression in Table 6, one sees a general trend of increasing Z score as the number of neurotherapy sessions increases.  Although the Z score does not reach statistical significance at completion, one might speculate that if the trend were to continue with additional neurotherapy sessions, then the mood/depression score might also reach statistical significance.

Stiffness

At each physician meeting, the patient was asked to estimate the duration of their stiffness, in minutes per day, up to a maximum of 360 minutes.  The data, presented in Tables 5 and 6, show that prior to starting neurofeedback training, patients had an average stiffness duration of 213 minutes, and at the conclusion the average was reduced to 145 minutes.  As with the other measures, there is scatter in the data, and the range for both pre- test and conclusion was 0-360 minutes.  However, the group overall reduction was 32%, and this included eight who were improved, three with no change, and four who had more stiffness at completion.
For those patients that did improve on the stiffness scale, their group overall improvement was 89%.  However, these encouraging results are clouded by a non-significant result from the Wilcoxon test of statistical significance.

When looking for trends in the data of Table 6, one sees an initial trend of decreasing stiffness score, but the trend is reversed at 40 sessions and at completion.  Likewise the Z score shows meaningless trend information.  One can conclude that neurotherapy is having little to no effect on this measure of discomfort.
TOVA Test Results

As Table 11 demonstrates, none of the TOVA auditory (A) test results, comparing completion results to baseline results (0 neurotherapy sessions), were statistically significant as calculated by the Wilcoxon test.  A rough check on these significances was made by comparing the Wilcoxon results to those obtained from the t-Test.  In all cases, the two tests gave similar statistically non-significant results.  Thus the TOVA? auditory test does not currently appear to be a good monitor of neurotherapy treatment results for fibromyalgia.
The TOVA? visual (V) test gave statistically significant results (p<0.05) between start and completion of neurotherapy for the sub-scores of ADHD, Commission errors (impulsivity or disinhibition), Response Time Variability, and D Prime (perceptual sensitivity).  All changes were in the direction of improved CNS functioning.

Correlation of TOVA Scores to Fibromyalgia Symptom Scores

As shown in Table 12, the correlation of mean fibromyalgia pain to mean Commission errors is –0.85.  Thus the Commission errors standard score appears initially to be a good monitor of neurotherapy treatment progress for fibromyalgia.  However, when calculations were made of the correlation between these scores for individual patients, the results fluctuated widely.  As seen in Table 13, correlations for individual patients ranged from –0.99 to + 0.90.

If the TOVA score were to be of value in monitoring fibromyalgia treatment progress, the correlation for all patients should be negative, and close to -1.0.  The negative sign means that when the fibromyalgia symptom score goes down (an improvement), the TOVA? score goes up (an improvement).  Because individual patient correlations varied over such a wide range, the TOVA? visual test does not provide a monitor of fibromyalgia neurotherapy treatment progress.  Thus it is concluded that, as presently administered and interpreted, the TOVA? visual and auditory tests do not offer a reliable fibromyalgia neurotherapy treatment monitor.

References

American Electroencephalographic Society (1991).  American Electroencephalographic Society guidelines standard electrode position nomenclature.  Journal of Clinical Neurophysiology, 8, 200-202.

Ayers, M.  E.  (1999).  Chapter 9, Assessing and Treating Open Head Trauma, Coma, and Stroke Using Real-Time Digital EEG Neurofeedback.  In Evans, J.R., and Arbanel, A. (Eds), Introduction to Quantitative EEG and Neurofeedback (pp204-222).  San Diego: Academic Press.

Baehr, E., Rosenfeld, J.  P., Baehr, R., and Earnest, C.  (1999).  Chapter 8, Clinical Use of an Alpha Asymmetry Neurofeedback Protocol in the Treatment of Mood Disorders.  In Evans, J.  R., and Arbanel, A.  (Eds),  Introduction to Quantitative EEG and Neurofeedback (pp 181-203).  San Diego: Academic Press. 

Barkley, R.  A.  (1992).  Is EEG Biofeedback Treatment Effective for ADHD Children? Proceed With Much Caution.  CH.A.D.D.er Box, April, 1992.

Bendat, J.  S., and Piersol, A.  G.  (1971).  Chapter 7, General Considerations in Data Acquisition and Processing.  In: Random Data: Analysis and Measurement Procedures, New York: Wiley-Interscience.

Bennett, R.  M.  (1987).  Fibromyalgia.  Journal of the American Medical Association, 257, 2802-2803.

Bennett, R.  M.  (1995).  Fibromyalgia: the Commonest Cause of Widespread Pain.  Comprehensive Pain, 21(6), 269-275.

Brown, V.  W.  (1995).  Neurofeedback and Lyme’s Disease: A Clinical Application of the Five Phase Model of CNS Functional Transformation and Integration.  Journal of Neurotherapy, 1, (2), Fall, 1995.

Brownback, T., and Mason, L. (1999).  Chapter 6, Neurotherapy in the treatment of Dissociation.  In Evans, J.  R., and Arbanel, A.  (Eds), Introduction to Quantitative EEG and Neurofeedback (pp 145-156).

Buskila, D., Neumann, L., Hershman, E., Gedalia, A., Press, J., and Sukenik, S. (1995).  Fibromyalgia syndrome in children: an outcome study.  Journal of Rheumatology,21, 525-528.

Caro, X.  J.  (1989).  Is There an Immunologic Component to the Fibrositis Syndrome?  Rheumatic Disease Clinics of North America, 15(1), 169-186.

Clauw, D.  J.  (1995).  The pathogenesis of chronic pain and fatigue syndromes, with special reference to fibromyalgia.  Medical Hypotheses, 44, 369-378.

Cornblatt, B.  A., Risch, N.  J., Faris, G., Friedman, D., and Erlenmeyer-Kimling, L.  (1988).  The Continuous Performance Test, identical pairs version (CPT-IP):I.  New findings about sustained attention in normal families.  Psychiatry Research, 26, 223-238.

Crook, J., Weir, R., and Tunks, E.  (1989).  An epidemiologic follow-up survey of persistent pain sufferers in a group family practice and specialty pain clinic.  Pain, 36, 49-61.

Demitrack, M.  A., and Crofford, L.  J.  (1998).  Evidence for and pathophysiologic implications of hypothalamic-pituitary-adrenal axis dysregulation in fibromyalgia and chronic fatigue syndrome.  Annals New York Academy of Science, 840, 684-697.

Elam, M., Johansson, G., and Wallin, B.  G.  (1992).  Do patients with primary fibromyalgia have an altered muscle sympathetic nerve activity?  Pain, 48, 371-375.

Epstein, H.  (1980).  EEG Developmental stages.  Developmental Psychobiology, 13, 629-631.

Felson, D.  T., and Goldenberg, D.  L.  (1986).  The natural history of Fibromyalgia.  Arthritis and Rheumatism, 29, 1522-1526. 

Goldenberg, D.  L.  (1989).  Diagnostic and Therapeutic Challenges of Fibromyalgia.  Hospital Practice, September, 30, 39-52.

Goldenberg, D.  L.  (1996a).  Fibromyalgia, chronic fatigue syndrome, and myofascial pain.   Current Opinion in Rheumatology, 8, 113-123.

Goldenberg, D.  L.  (1996b).  What is the future of fibromyalgia?  Rheumatic Disease Clinics of North America, 22, (2), 393-406.

Granges, G., Zilko, P., and Littlejohn, G.  O.  (1994).  Fibromyalgia syndrome: Assessment of the severity of the condition 2 years after diagnosis.  Journal of Rheumatology, 21, 5213-529.

Greenberg, L.  M., and Kindschi, C.  L.  (1996).   T.O.V.A? Clinical Guide (pg 20).  Universal Attention Disorders, Inc.  Los Alamitos, CA.

Griep, E.  N., Boersma, J.  W., Lentjes, E.   G., Prins, A.  P., van der Korst, J.  K., and de Kloet, E.  R.  (1998).  Function of the hypothalamic-pituitary-adrenal axis in patients with fibromyalgia and low back pain.  Journal of Rheumatology, 25(7), 1374-1381.

Halperin, J.  M., Wolf, L., Greenblatt, E.  R., and Young, G.  (1991).  Subtype analysis of commission errors on the continuous performance test in children.  Developmental Neuropsychology, 7(2), 207-217.

Harding, S.  M.  (1998).  Sleep in Fibromyalgia Patients: Subjective and Objective Findings.  The American Journal of the Medical Sciences, 315(6), 367-376.

Hardt, J.  V., and Kamiya, J.  (1978).  Anxiety change through electroencephalo-graphic alpha feedback seen only in high anxiety subjects.  Science, 7, 201, 79-81.

Hauri, P.  J., Percy, L., Hellekson, C., Hartmann, E., and Russ, D.  (1982).  The treatment of psychophysiologic insomnia with biofeedback: a replication study.  Biofeedback and Self-Regulation, 7(2), 223-235

Hoffman, D.  A., Stockdale, S., Hicks, L.  L., and Schwaninger, B.  A.  (1995).  Diagnosis and Treatment of Head Injury.  Journal of Neurotherapy, 1, 1, Summer 1995.

Hudspeth, W.  (1985).  Developmental neuropsychology: functional implications of quantitative EEG maturation.  Journal of Clinical and Experimental Neuropsychology, 7, 606.

Hudspeth, W., and Pribram, K.  (1990).  Stages of brain and cognitive maturation.  Journal of  Educational Psychology, 82, 881-884.

Hudspeth, W., and Pribram, K.  (1991).  Physiological indices of cerebral maturation.  International Journal of Psychophysiology, 12, 19-29.

James, L.  C., and Folen, R.  A.  (1996).  EEG biofeedback as a treatment for chronic fatigue syndrome: a controlled case report.  Behav Med, Summer, 22(2), 77-81.

Kamiya, J.  (1968).  Conscious Control of Brainwaves.  Psychology Today, 1(11), 57-60.

Kamiya, J., et al.  (1969).  Visual evoked responses in subjects trained to control alpha rhythms.  Psychophysiology,  5(6), 683-695.

Kennedy, M.  J., Goldenberg, D.  L., and Felson, D.  T.  (1994).  A prospective long-term study of  Fibromyalgia (abstract).  Arthritis and Rheumatism, 37, S213.

Laibow, R.  (1999).  Chapter 4, Medical Applications of  NeuroBioFeedback.  In Evans, J.  R., and Arbanel, A, (Eds), Introduction to Quantitative EEG and Neurofeedback (pp 83-102).  San Diego: Academic Press.

    Landers, D.  M., Petruzello, S.  J., Salazar, W., Crews, D.  J., Kubitz, K.  A., Gannon, T.  I., and Han, M.  (1991).  The influence of electrocortical biofeedback on performance in pre-elite archers.  Med Sci Sports Exerc, 23(1), 123-129.

Leark, R.  A., Dupuy, T.  R., Greenberg, L.  M., Corman, C.  L., and Kindschi, C.  L.  (1996).  T.O.V.A? Test Of Variables Of Attention, Professional Manual, Version 7.0  Universal Attention Disorders, Inc.  Los Alamitos, CA.

Ledingham, J., Doherty, S., and Doherty, M.  (1993).  Primary fibromyalgia syndrome-an outcome study.  British Journal of Rheumatology, 32, 139-142. 

Lubar, J.  F.  (1995).  Chapter 20. Neurofeedback for the Management of Attention-Deficit/Hyperactivity Disorders.  In Schwartz, M.  S. and Associates (Eds.), Biofeedback, a Practitioners Guide  (pp. 493-522).  New York: The Guilford Press.

Lubar, J.  F., and Bahler, W.  W.  (1976).  Behavioral management of epileptic seizures following EEG biofeedback training of the sensorimotor rhythm.  Biofeedback and Self-Regulation, 7, 77-104.

Lubar, J.  F., and Lubar, J.  O.  (1999).  Chapter 5, Neurofeedback Assessment and Treatment for ADD/HD.  In Evans, J.  R., and Arbanel, A., (Eds), Introduction to Quantitative EEG and Neurofeedback (pp103-143).  San Diego: Academic Press.

Lubar, J.  F., and Shouse, M.  N.  (1976).  EEG and behavioral changes in a hyperactive child concurrent with training of the sensorimotor rhythm (SMR).  A preliminary report.  Biofeedback and Self-regulation, 1, 293-306.

Lubar, J.  F., and Shouse, M.  N.  (1977).  Use of biofeedback in the treatment of  seizure disorders and hyperactivity.  In B.  B. Lahey and A.  E. Kazdin (Eds.), Advances in clinical child psychology  (pp. 203-265).  New York: Plenum Press.

MacFarlane, G.  J., Thomas, E., Papageorgion, A.  C., Schollum, J., Croft, P.  R., and Silman, A.  J. (1996).  The natural history of chronic pain in the community: a better prognosis than in the clinic.  Journal of Rheumatology, 23, 1617-1620. 

Matousek, M., and Petersen, L.  (1973).  Frequency analysis of the EEG background activity by means of age dependent EEG quotients.  In Kellaway and Petersen (Eds.), Automation of Clinical Electroencephalography, New York: Raven Press, New York.

Nahmias, J., Tansey, M.  A., and Karetsky, M.  S.  (1994).  Asthmatic extrathoracic upper airway obstruction: laryngeal dyskinesis.  N. J. Med., 91(9), 616-620.

Nørregaard, J., Bulow, P.  M., Prescott, E., Jacobsen, S., and Danneskiold-Samsoe, B.  (1993).  A four-year follow-up study in fibromyalgia.  Relationship to chronic fatigue syndrome.  Scandinavien Journal of Rheumatology, 22(1), 35-38.

Norris, S.  L., and Currieri, M.  (1999).  Chapter 10, Performance Enhancement Training Through Neurofeedback.  In Evans, J.R., and Arbanel, A., (Eds), Introduction to Quantitative EEG and Neurofeedback. (pp 224-240).  San Diego: Academic Press.

O’Dougherty, M., Neuchterlein, K.  H., and Drew, B.  (1984).  Hyperactive and hypoxic children: signal detection, sustained attention, and hyperactivity.  Children’s Hospital, Columbus, Ohio. 

Othmer, S.  (1998a).  EEG Biofeedback Training: A Response to Russell Barkley.  In Training Syllabus, Volume Three, Attention Deficit Disorder. (pp 129-132).  EEG Spectrum, Encino, California.

Othmer, S.  (1998b).  EEG Changes During Biofeedback.  In Training Syllabus, Volume One, Attention Deficit Disorder. (page 4-22).  EEG Spectrum, Encino, California

Othmer, S., Othmer, S.  F., and Kaiser, D.  A.  (1999).  Chapter 11, EEG Biofeedback: An Emerging Model for Its Global Efficacy.  In Evans, J.  R., and Arbanel, A., (Eds), Introduction to Quantitative EEG and Neurofeedback (pp 243-310).  San Diego: Academic Press.

Packard, R.  C., and Ham, L.  P.  (1996).  EEG Biofeedback in the treatment of Lyme Disease: A Case Study.  Journal of Neurotherapy, 1(3), Winter, 1996. 

Peniston, E.  G., and Kulkosky, P.  J.  (1989).  Alpha-theta brainwave training and beta-endorphin levels in alcoholics.  Alcohol Clin Exp Res, 13(2), 271-279.

Penniston, E.  G., and Kulkosky, P.  J.  (1999).  Chapter 7, Neurofeedback in the Treatment of Dissociation.  In Evans, J.  R., and Arbanel, A., (Eds), Introduction to Quantitative EEG and Neurofeedback.  (pp 145-156).  San Diego: Academic Press.

Qiao, Z.  G., Vaerøy, H., and Mørkrid, L.  (1991).  Electrodermal and microcirculatory activity in patients with fibromyalgia during baseline, acoustic stimulation and cold pressor tests.  Journal of Rheumatology, 18, 1383-1389. 

Rhind, V.  M., Unsworth, A., and Haslock, I.  (1987).  Assessment of Stiffness in Rheumatology: The Use of Rating Scales.  British Journal of Rheumatology, 26, 126-130. 

Rozelle, G.  R., and Budzynski, T.  H.  (1995).  Neurotherapy for stroke rehabilitation: a single case study.  Biofeedback Self Regulation, 20(3), 211-228.

Russell, I.  J.  (1998).  Advances in Fibromyalgia: Possible Role for Central Neurochemicals.  The American Journal of the Medical Sciences,315 (6), 377-384.

Saxby, E., and Peniston, E.  G.  (1995).  Alpha-theta brainwave neurofeedback training: an effective treatment for male and female alcoholics with depressive symptoms.  J Clin Psychol, 51(5), 685-693.

Seifert, A.  R., and Lubar, J.  F.  (1975).  Reduction of epileptic seizures through EEG biofeedback training.  Biol Psychol, 3(3), 157-184.

Sichel, A.  G., Fehmi, L.  G., and Goldstein, D.  M.  (1995).  Positive Outcome With Neurofeedback Treatment In a Case of Mild Autism.  Journal of Neurotherapy, 1(1) Summer, 1995.

Slotkoff, A.  T., and Clauw, D.  J.  (1996).  Fibromyalgia: When thinking is impaired.  J. Musculoskeletal Medicine, September, 32-36.

Solomon, D. H., and Liang, M. H. (1997).  Fibromyalgia: scourge of humankind or bane of a rheumatologist’s existence?  Arthritis and Rheumatism, 40(9), 1553-1555.

St. Amand, R.  P., and Marek, C.  C.  (1999).  What your doctor may not tell you about fibromyalgia.: the revolutionary treatment that can reverse the disease.  New York, Warner Books.

Stephens, S.  S.  (1962).  The surprising simplicity of sensory metrics.  American Psychologist, 17, 29-39.

Sterman, M.  B.  (1973).  Neurophysiologic and clinical studies of sensorimotor EEG biofeedback training: some effects on epilepsy.  Seminars in Psychiatry, 5(4), 507-525.

Sterman, M.  B. and Friar, L.  (1972).  Suppression of seizures in an epileptic following sensorimotor EEG feedback training.  Electroencephalography and Clinical Neurophysiology, 33, 89-95.

Swingle, P.  G.  (1998).  Neurofeedback treatment of pseudoseizure disorder.  Biol. Psychiatry, 44 (11), 1196-1199.

Tansey, M.  A.  (1986).  A simple and a complex tic (Gilles de la Tourette’s syndrome): their response to EEG sensorimotor rhythm biofeedback training.  Int J Psycholphysiol, 4(2), 91-97.

Tansey, M.  A.  (1993).  Ten-year stability of EEG biofeedback results for a hyperactive boy who failed fourth grade perceptually impaired class.  Biofeedback and Self Regulation, 18(1), 33-44.

Tansey, M.  A., and Brunner, R.  L.  1983).  EMG and EEG biofeedback in the treatment of a 10-year-old hyperactive boy with a developmental reading disorder.  Biofeedback and Self-Regulation, 8(1), 25-37.

Thatcher, R.  W.  (1980).  Neurolinguistics: theoretical and evolutionary perspectives.  Brain and Language, 11 235-260.

Thatcher, R.  W.  (1999).  Chapter 2, EEG Database-Guided Neurotherapy.  In Evans, J.  R., and Arbanel, A., (Eds), Introduction to Quantitative EEG and Neurofeedback (pp29-64).  San Diego: Academic Press.

Thomas, J.  E., and Sattleberger, B.  A.  (1997).  Treatment of Chronic Anxiety Disorder with Neurotherapy: A Case Study.  Journal of Neurotherapy, 2(1), Spring-Summer, 1997.

van Denderen, J.  C., Boersma, J.  W., Zeinstra, P., Hollander, A.  P., and van Neerbos, B.  R.  (1992).  Physiological effects of exhaustive physical exercise in primary fibromyalgia syndrome (PFS): is PFS a disorder of neuroendocrine reactivity?  Scandinavian Journal of Rheumatology, 21, 35-37.

Vlieland, T.  P.  M.  V., Zwinderman, A.  H., Breedveld, F.  C., and Hazes, J.  M.  W.  (1997).  Measurement of Morning Stiffness in Rheumatoid Arthritis Clinical Trails.  J. Clinical Epidemiology, 50(7), 757-763.

Wallace, D.  J.  (1997).  The Fibromyalgia Syndrome.  Annals of Medicine, 29, 
9-21.

Weigent, D.  A., Bradley, L.  A., Blalock, J.  E., and Alarcón, G.  S.  1998).  Current concepts in the Pathophysiology of Abnormal Pain Perception in Fibromyalgia.  The American Journal of the Medical Sciences, 315(6), 405-412.

Wolfe, F., and Pincus, T. (1995).  Data collection in the clinic.  Rheumatic Disease Clinics of North America, 21(2), 321-358.

Wolfe, F., Hawley, D.  J., Cathey, M.  A., Caro, X.  J., and Russell, I.  J.  (1985).  Fibrositis: Symptom Frequency and Criteria for Diagnosis.  The Journal of Rheumatology, 12(6), 1159-1163.

Wolfe, F., Anderson, J., Harkness, D., Bennett, R.  M., Caro, X.  J., Goldenberg, D.  L., Russell, I.  J., and Yunus, M.  B.  (1997a).  A Prospective, Longitudinal, Multicenter Study of Service Utilization and Costs in Fibromyalgia.  Arthritis and Rheumatism, 40(9), 1560-1570.

Wolfe, F., Anderson, J., Harkness, D., Bennett, R.  M., Caro, X.  J., Goldenberg, D.  L., Russell, I.  J., and Yunus, M.  B.  (1997b).  Health Status and Disease Severity in Fibromyalgia.  Arthritis and Rheumatism, 40(9), 1571-1579.

Wolf, F., Smythe, H.  A., Yunus, M.  B., Bennett, R.  M., Bombardier, C., Goldenberg, D.  L., Tugwell, P., Campbell, S.  M., Abeles, M., Clark, P., Fam, A.  G., Farber, S.  J., Feichtner, J.  J., Franklin, C.  M., Gatter, R.  A., Hamaty, D., Lessard, J., Lichtbroun, A.  S., Masi, A.  T., McCain, G.  A., Reynolds, W.  J., Romano, T.  J., Russell, I.  J., Sheon, R.  P.  (1990).  The American College of Rheumatology 1990 criteria for the classification of fibromyalgia: report of the multicenter criteria committee.  Arthritis Rheum., 33, 160-172.