Psychological Treatments in Functional Gastrointestinal Disorders

A Primer for the Gastroenterologist

Olafur S. Palsson; William E. Whitehead


Clin Gastroenterol Hepatol. 2012;11(3):208-216. 

In This Article


Biofeedback is a form of behavioral training that uses continuous visual or auditory feedback from recordings of specific physiological activity to enable patients to learn to voluntarily control those body functions. For example, in a patient who is constipated because she paradoxically contracts her pelvic floor muscles when having a bowel movement, the electromyographic activity of her pelvic floor muscles might be shown as a dynamic graph on a computer screen, while she simulates defecation to help teach her how to relax the pelvic floor muscles instead of contracting them. The therapist would provide verbal instructions and encouragement during her attempts to relax the muscles. In a patient with fecal incontinence, on the other hand, biofeedback might be used to teach patients how to more effectively contract an external anal sphincter that is very weak because of an obstetrical injury or other causes; in this case, biofeedback would be used to teach the patients an appropriate pelvic floor muscle exercise to practice at home to gradually increase the strength of the muscle. Biofeedback can also be used for sensory training (ie, to improve the patient's ability to detect and respond appropriately to physiological sensations such as stool or gas suddenly filling up the rectum). This type of training would be used in a patient who is unable to recognize when it is necessary to contract the pelvic floor muscles to prevent leaking gas or liquid stool because of a nerve injury. Usually 4–6 training sessions spaced 1–2 weeks apart are used whether the indication is constipation or fecal incontinence.

Pelvic floor biofeedback as described above is distinctly different from the other forms of psychological treatment for FGIDs discussed above because it is not used to cause changes in thoughts or feelings; rather, it is used to help patients learn to overcome specific physiological deficits directly. A different type of biofeedback is sometimes used to teach patients how to relax all the muscles of their body or to reduce autonomic arousal to counteract stress, but those techniques are very different and are rarely used to treat gastrointestinal disorders.

Biofeedback has been tested in randomized studies (Supplementary Table 5) almost exclusively as a therapy for functional constipation (16 RCTs) and fecal incontinence (9 RCTs). Comparison conditions have varied, including medical management, sham or non-anorectal biofeedback, balloon defecation training, polyethylene glycol, behavioral modification, diazepam, placebo, Botox, and surgery. Six of 9 RCTs in constipated adults found biofeedback patients to have significantly better outcomes than control patients (Supplementary Table 5). Of the 3 negative studies, one was methodologically flawed because it did not limit enrollment to patients with evidence of pelvic floor dyssynergia;[35] it is now evident that biofeedback does not work for constipation unless patients have dyssynergic defecation.[36] The other 2 negative trials came from a single group of researchers in Egypt[37,38] who found biofeedback to be less effective compared with Botox injection and, in one of the studies, compared with puborectalis surgery as well. In short, it seems that biofeedback is effective in adults with dyssynergic functional constipation, although it may possibly not be as effective as Botox injection or surgery. The largest and methodologically strongest trials all show a clinically significant improvement in constipation that is substantially greater in patients treated with biofeedback than in control patients (70%–86% vs 22%–48%),[39–41] and therapeutic gains are well maintained for at least 1 year after treatment.

In contrast to the predominantly positive studies in adults, only 2 of 7 published controlled trials testing biofeedback for constipation in children found an advantage for biofeedback compared with the control conditions. This unfavorable outcome picture may be due in part to 2 of the trials not limiting testing to dyssynergic patients. It has also been suggested that children may lack the ability to concentrate on the biofeedback task for sustained periods of training.[42] In any case, on the basis of the available evidence, biofeedback cannot be recommended for treating constipation in children.

Biofeedback has been advocated for the treatment of fecal incontinence since the first case series was described in 1974.[43] However, as shown in Supplementary Table 5, only 3 of 9 RCTs showed superior outcomes of biofeedback treatment compared with control conditions. Both of the 2 trials of biofeedback for encopresis in children yielded negative results.[44,45] The 3 studies in adults that compared biofeedback with conservative management that combined pelvic floor exercises with education[46–48] likewise found no advantage for biofeedback, which suggests that biofeedback may not generally help fecal incontinence patients more than those more widely available interventions for fecal incontinence. Much better results were found for biofeedback in adults in a study by Heymen et al,[49] who only enrolled nonresponders to a run-in education and medical management intervention in a randomized trial of either biofeedback or pelvic floor exercises alone. Biofeedback showed clear superiority under these circumstances (77% of patients reported adequate relief vs 48% in the pelvic floor exercise group), and improvements were well maintained for up to 12 months. These findings suggest that biofeedback is likely to be useful for improving outcomes for adult fecal incontinence patients who have not had satisfactory response to conservative management.

Biofeedback has also been tested for anorectal pain and functional dyspepsia in single RCTs. In a large trial of 157 individuals with chronic idiopathic anorectal pain,[50] biofeedback resulted in much higher rates of pain relief (87%) than were seen for electrogalvanic stimulation (45%) and levator muscle massage (22%) conditions, and treatment benefits were maintained at 1-year follow-up. Thus, biofeedback holds strong promise for the future as a possible advancement in the management of this difficult-to-treat problem. The only trial of biofeedback for functional dyspepsia[51] did not report the effect of this treatment on dyspepsia symptoms (but reported biofeedback to result in greater drinking capacity and quality of life improvement compared with no treatment) and therefore provided little direct information about the potential value of this therapy for improving the clinical symptoms of dyspepsia.

As this summary of these 5 psychological treatments illustrates, a vast amount of published RCT data support the value of psychological therapies as interventions for FGIDs. It should be acknowledged that many of the trials in this domain have methodological shortcomings. The great majority of them have been small; many have used waiting lists or mere symptom monitoring as control groups, both of which are likely to produce negative expectation of improvement that may exaggerate outcome contrast with the active treatment. Therapies are sometimes poorly described, making it hard to know exactly what therapy was tested or how to replicate it. Results have also been measured in numerous different ways that make it hard to compare outcomes across trials. Nonetheless, the preponderance of evidence, buttressed by some methodologically strong trials for each psychological modality tested, makes a strong case for psychological treatment as adjunctive therapy options to consider for FGID patients.