e4-09: Biofeedback

Biofeedback is a behavioral therapy method that teaches a patient to gain greater awareness and control over physiologic functions, including some not normally under conscious control. This is achieved by using technology that presents to the patient in visual or auditory form the level of activity of a physiologic parameter, such as muscle tension, skin temperature, sweat gland activity, respiratory muscle activity, PCO₂, heart rate variability, or brain wave pattern.

Biofeedback began in the 1950s with the work of the experimental psychologist Dr. Neil Miller. He demonstrated that rats could learn to regulate autonomic functions when motivated by painful or pleasurable stimuli. The idea that autonomic function could be volitionally controlled was contrary to the scientific dogma of the time. In the 1960s, other researchers began to examine the human ability to consciously alter physiologic function.

Certification is obtained through the Biofeedback Certification International Alliance (BCIA) and is open to professionals from clinical health care areas including, but not limited to, psychology, nursing, and counseling. Requirements for certification are as follows: (1) 48 hours of didactic education; (2) 20 hours of contact spent with a BCIA-approved mentor, 50 patient/client sessions, and case conference presentations; and (3) successful completion of a written certification examination. There is also a continuing education requirement.

When treating a medical or psychological disorder, biofeedback practitioners are required to hold a current state license in a BCIA-approved health care field or, if without this license, work under the supervision of an appropriately credentialed health care professional. Separate, additional certificates are required for electroencephalographic (EEG) biofeedback (neurofeedback) and pelvic muscle dysfunction work.

Since biofeedback is a taught, self-regulatory skill, patients should be motivated; prepared to take an active role in their treatment; respond well to visible, quantifiable information; or have aptitude or familiarity with technology. Typical biofeedback sessions last 50 minutes, with half or more of that time in active monitoring. To begin a biofeedback session, the therapist applies sensors to the patient that monitor the physiologic parameters of interest. For example, in a case of chronic tension headache, an electromyographic (EMG) sensor may be applied to the skin of the forehead. Information from the sensor feeds back to the patient as sound or as an image on a computer screen in real time, so he or she can begin to learn to reduce tension in the forehead muscle. Multiple sessions are usually required, and the patient is asked to practice the learned skills between sessions. The goal of biofeedback is to help the patient train himself or herself to produce the desired response at other times. Some biofeedback software programs and sensors are available for use on a home computer or with portable electronic devices.

There are no reported adverse effects of biofeedback therapy. Benzodiazepines, opioids, other sedating medicines, and even caffeine can have undesirable effects on biofeedback training.

A. Hypertension

The AHA 2013 Scientific Statement paper on CIT for blood pressure lowering conferred a “Class IIB, level of evidence B” designation for the efficacy of biofeedback for treating hypertension, and stated that “biofeedback may be considered in clinical practice to lower blood pressure.” Studies on the blood pressure–lowering effect of biofeedback have been published for over 40 years. Meta-analyses consistently show that compared with wait-list or other inactive control groups, biofeedback can significantly lower patients' systolic and diastolic blood pressures (approximately 5–8 mm Hg systolic and 3–5 mm Hg diastolic pressure). In many of the older studies when biofeedback is compared with an active control group (eg, sham biofeedback or nonspecific behavioral control), it is usually not statistically superior in its effect on blood pressure except when coupled with relaxation training or cognitive therapy. On the other hand, follow-up studies suggest that biofeedback is indeed superior to active controls. For example, a 2012 randomized controlled trial of 43 prehypertensive persons compared heart rate variability biofeedback with slow abdominal breathing and treatment as usual. The biofeedback group experienced significantly greater blood pressure reductions than either control group as well as increased baroreceptor sensitivity. In general, the degree of blood pressure reduction with biofeedback strongly correlates with pretest blood pressure. Persons with higher baseline blood pressures show larger beneficial effects from biofeedback.

Because most patients in studies of nonpharmacologic treatment of hypertension have only mild hypertension, the effect of treatments like biofeedback are more difficult to detect. Also, it is worth noting that, as is the case for all antihypertensive therapies, small reductions in blood pressure yield important clinical benefits.

There is also research on respiratory biofeedback training for blood pressure reduction. A particular pattern of device-assisted respiration practiced for 15 minutes daily has been shown in seven of nine small studies to lower ambulatory blood pressure. A thorough 2012 meta-analysis concluded that both diastolic and systolic blood pressure may be reduced with short-term use of device-guided breathing. However, because five of the eight trials were sponsored by or involved the manufacturers of the device, they were excluded. With their exclusion, no overall effect was found. Longer-term, independent trials are required. A 2013 randomized controlled trial involving 48 type 2 diabetic patients found no blood pressure benefit to using this device-guided breathing technique. A 2015 review of 15 published studies of this device concluded that it does significantly lower blood pressure (but one of the review's authors may have had a financial conflict of interest).

B. Coronary Artery Disease

There have been clinical studies of the effect of biofeedback on heart rate variability. Heart rate variability is considered to be an indicator of autonomic tone and has been postulated to be the mediator of the effects of cognitive and mind-body therapies on cardiovascular conditions. Cardiac electrophysiology studies have shown that decreased heart rate variability is predictive of both short-term and long-term mortality after myocardial infarction. One randomized controlled trial of 154 patients with established coronary artery disease showed that biofeedback significantly increases heart rate variability. This same research group conducted a second randomized controlled study looking at a 1-year follow-up of 222 patients and found fewer all-cause readmissions and all-cause emergency visits (13.33 vs 35.59%) than the control group. Another randomized controlled trial showed an 86% reduction in post-myocardial infarction mortality at 2 years from a psychosocial intervention that included biofeedback training of heart rate variability. However, this study did not show a significant change in heart rate variability with the intervention unless individuals with high heart rate variability at baseline were excluded. Thus, it remains unanswered whether improved heart rate variability from biofeedback training indeed affects important cardiovascular outcomes.

C. Raynaud Phenomenon

There have been eight randomized controlled trials of the efficacy of biofeedback on Raynaud phenomenon. Three show biofeedback to be superior to active control (eg, autogenic training), and four show benefit from biofeedback but equal to active control. One large study showed thermal biofeedback to be equivalent to EMG biofeedback treatment and to treatment with sustained-release nifedipine.

D. Headache

Biofeedback and other behavioral techniques have been extensively studied in prevention of recurrent migraine and tension headaches, with more than 50 studies including many randomized controlled trials. Several meta-analyses concur there is a significant reduction (30–55%) in headache frequency with biofeedback and other behavioral interventions. Other outcomes with significant improvement in several headache studies are anxiety, depression, self-efficacy, and medication usage. The American Academy of Neurology practice guidelines state that there is grade A evidence for the use of thermal biofeedback with relaxation or EMG biofeedback for prevention of migraine. Biofeedback compares favorably with pharmaceutical treatments in several studies. There are also studies showing enhanced efficacy from adding medications to biofeedback therapy. The American Academy of Neurology practice guidelines state that there is grade B evidence for behavioral therapy (eg, biofeedback and relaxation techniques) when combined with prophylactic medication therapy (eg, propranolol and amitriptyline) to achieve additional clinical improvement for migraine relief. Finally, a 2016 randomized controlled pilot study of 27 patients suggested that biofeedback may have a role in treating medication-overuse headache.

E. Urinary Incontinence

Pelvic floor muscle training is an effective treatment for incontinence in women. According to a 2011 Cochrane review of 24 trials, biofeedback-assisted pelvic floor muscle training yields even greater benefit for incontinence in women, but a 2012 review questioned whether observed effect could be related to the amount of health professional contact. A 2013 randomized controlled trial supported the use of biofeedback for urinary incontinence in premenopausal women. In 2016, a randomized controlled trial involving 58 women with overactive bladder symptoms showed significantly improved symptoms and quality of life with EMG biofeedback-assisted pelvic muscle therapy after 9 weeks of treatment. In men with urinary incontinence after prostatectomy, a 2016 systematic review and meta-analysis of 13 randomized controlled trials involving a total of 1108 patients found that the immediate-, intermediate-, and long-term effects of biofeedback-assisted pelvic floor muscle training on urinary incontinence were significantly better than pelvic floor muscle training alone. Small-to-moderate immediate- and intermediate-term effects were also observed on the quality of life. A 2012 randomized controlled trial of 52 post-prostatectomy men showed that pelvic floor biofeedback significantly improved erectile function.

F. Fecal Incontinence

There are more than 40 experimental trials of biofeedback for fecal incontinence in adults, and all but two of these report moderate to large benefit. However, the methodologic quality of some of these studies is low. A 2012 Cochrane review of biofeedback and sphincter exercises for fecal incontinence included 21 randomized trials and concluded that there was evidence that biofeedback enhanced treatment outcomes; however, because of methodological weaknesses of many of the trials, a definitive conclusion was not possible. A 2013 systematic review of 13 trials was more positive, noting that biofeedback plus electrical stimulation was particularly effective. A report of 513 consecutive patients seen at a tertiary colorectal referral clinic showed that more than 70% of patients received benefit. Skeptics have noted that most of the studies on biofeedback for fecal incontinence were performed in tertiary referral centers with specialization in this modality. However, a 2017 retrospective study of 205 patients with fecal incontinence using low-intensity biofeedback therapy at a secondary care center found success rates comparable to tertiary referral centers (71% response rate, with 47% of patients achieving continence). Good anal sphincter exercise technique, longer in-home practice, and more sessions were significantly associated with a favorable outcome. Another important real-world study in 2013 showed that less than half of patients with fecal incontinence referred for pelvic biofeedback actually received this therapy due to lack of insurance coverage or geographic distance to treatment facility but, of those who did ultimately receive biofeedback therapy, the success rate was high (80%), consistent with previous positive study reports. Another 2013 trial found significant quality of life benefit for fecal incontinence patients undergoing biofeedback therapy. Many authors and clinicians in this field consider pelvic biofeedback to be first- or second-line therapy for fecal incontinence. American College of Gastroenterology Practice Guidelines consider biofeedback safe and effective and recommend it for treatment of fecal incontinence.

G. Chronic Constipation

Some 30–50% of cases of chronic constipation are due to pelvic floor dyssynergia, a failure of the pelvic floor and anal muscles to relax during straining. Pelvic floor dyssynergia may be less amenable to laxative and fiber treatment but appears to be successfully treated with biofeedback. This indication has been studied in adults in at least 11 randomized controlled trials. In six trials, biofeedback was superior to sham biofeedback, polyethylene glycol, diazepam, usual care, and placebo. Because of its success rate (70–80%) and safety, biofeedback is widely recommended as first-line therapy alone or in combination with other measures. A 2012 report of 226 patients referred to a tertiary care center also showed that over two-thirds of patients with dyssynergic defecation had improved symptoms with pelvic biofeedback therapy. A randomized controlled trial of 157 patients with a related phenomenon, the levator ani syndrome, showed similar results; patients responded more often to biofeedback than to massage or electrostimulation.

H. Stroke Rehabilitation

Biofeedback for stroke rehabilitation has a sound theoretical basis but has been difficult to evaluate for several reasons, including nonstandardized outcome measures, multiple different versions of biofeedback therapies, and variability in the time post-stroke at which biofeedback is offered. There are more than 30 randomized controlled trials, all with small sample sizes, and although the majority of studies show benefits, such as improved motor strength, gait quality, and functional recovery, results are mixed. A Cochrane review concludes that EMG biofeedback cannot be recommended as a routine treatment but, due to its safety and possible efficacy, its use can be reasonably considered. A comprehensive 2009 review on motor recovery from stroke similarly points out the limitations of the biofeedback literature but concludes there are positive results for recovery of arm function and standing balance but not for sit-to-stand transfer ability. At least four trials have concluded that there is significant benefit of biofeedback for gait training after stroke. A 2013 randomized controlled trial showed improvement in hand function with biofeedback, and a 2014 randomized controlled trial showed improvement in lower extremity function. In two 2016 trials, visual biofeedback did not enhance outcomes of gait symmetry and postural balance beyond what was achieved with conventional treadmill training and physical therapy, but in a third trial conducted in patients with poststroke dysphagia, laryngeal elevation training combined with game-based biofeedback increased the removal rate of nasogastric tubes.

I. Psychiatric Disorders

EEG biofeedback, also known as neurofeedback, was first used to reduce seizure frequency in epilepsy in the 1970s. Since then, it has been applied to psychoses, substance use disorders, anxiety disorders, affective disorders, traumatic brain injury, attention-deficit/hyperactivity disorders, and other psychiatric conditions. However, there are few methodologically rigorous trials investigating this form of biofeedback in adults. A 2012 case series report of 51 patients with schizophrenia showed neurofeedback led to improvements in the Positive and Negative Syndrome Scale score in the majority of patients. A 2016 randomized controlled trial of 84 patients with psychotic symptoms who underwent heart rate variability biofeedback sessions showed that no group differences occurred in paranoid symptoms or subjective stress, although paranoia was diminished in the subset of participants who were breathing per protocol. A small 2013 randomized trial of 20 people with opiate addiction undergoing neurofeedback showed improvement in somatic symptoms, mental health, relief from withdrawal, and reduced desire to use opioids. A 2016 review of neurofeedback for PTSD examined five studies and found that neurofeedback had a statistically significant effect in three studies.