Biofeedback training in treatment of childhood constipation: a randomised controlled study

BACKGROUND: Because abnormal defaecation dynamics, which can be modified by biofeedback, are considered to be the underlying problem in constipation, biofeedback training may be a useful treatment for constipation. This treatment has mainly been studied in uncontrolled trials. We evaluated defaecation dynamics and clinical outcome in chronically constipated children in a randomised study comparing conventional treatment and conventional treatment with biofeedback training. METHODS: Patients, 5 to 16 years old, were referred to the Academic Medical Center in Amsterdam by general practitioners, school doctors, paediatricians, and psychiatrists. They had to fulfil at least two of four criteria for paediatric constipation and were included if they had been treated medically for at least one month before randomisation. Patients had a medical history, abdominal and rectal examination, and anorectal manometry at the start and end of the 6-week intervention period. The conventional group received laxative treatment with additional dietary advice, toilet training, and maintenance of a diary of bowel habits. The biofeedback group received the same conventional treatment and additionally five biofeedback training sessions. During the first 3 weeks, patients visited the outpatient clinic weekly; two subsequent visits were twice monthly. FINDINGS: 94 patients were randomised to conventional treatment (CT) and 98 to conventional treatment with additional biofeedback training (CT+BF). Normal defaecation dynamics increased in the CT group from 41% to 52% (not significant) and in the CT+BF group from 38% to 86% (p = 0.001). At 6 weeks, more patients in the CT+BF group showed normal defaecation dynamics, compared to the CT group (p < 0.001). This result was unaltered by controlling for baseline status in a logistic regression model. At 1 year, successful treatment (defaecation frequency > or = 3/week, soiling and/or encopresis < 2/month, and no laxatives) was accomplished in 59% of the CT and 50% of the CT+BF group (p = 0.24). The results were maintained after 1 1/2 years follow-up. No association was found between achievement of normal defaecation dynamics and clinical outcome. INTERPRETATION: Additional biofeedback training compared to conventional therapy did not result in higher success rates in chronically constipated children. Furthermore, achievement of normal defaecation dynamics was not associated with success: abnormal defaecation dynamics seem not to play a crucial role in the pathogenesis of childhood constipation. Intensive conventional laxative treatment should remain the first choice in chronically constipated children.