Objectives: Chronic constipation is a common childhood problem and often caused or worsened by abnormal dynamics of defecation. The aim of this study was to assess the benefit of pelvic floor physical therapy (PFPT), a novel treatment in pediatrics for the treatment of chronic constipation with dyssynergic defecation. Methods: This was a retrospective study of 69 children seen at a pediatric neurogastroenterology program of a large tertiary referral center for chronic constipation and dyssynergic defecation, determined by anorectal manometry and balloon expulsion testing. We compared the clinical outcome of patients who underwent PFPT (n = 49) to control patients (n = 20) whom received only medical treatment (laxatives/stool softeners). Additionally, characteristics of the treatment group were analyzed in relation to therapeutic response. Results: Thirty-seven (76%) of the patients who received physical therapy had improvement in constipation symptoms, compared to 5 (25%) of the patients on conservative treatment (p < 0.01). Additionally, patients who received pelvic physical therapy had fewer hospitalizations for cleanouts (4 vs. 25%, p = 0.01) and colonic surgery than those that were treated with medical therapy exclusively (0 vs. 10%, p = 0.03). Among the patients who received physical therapy, those that suffered from anxiety and/or low muscle tone had a higher response rate (100%). There were no adverse effects from the intervention. Conclusion: The new field of pediatric PFPT is a safe and effective intervention for children with dyssynergic defecation causing or contributing to chronic constipation, particularly in children whose comorbidities include anxiety and low muscle tone.
Constipation is a common problem in childhood, accounting for 5% of pediatrician visits and 25% of gastroenterology visits . Although constipation has been extensively studied, much remains unknown about its pathophysiology with various hypothesized etiologies based on age and symptoms . Abnormal defecation dynamics associated with pelvic outlet obstruction are suspected to be a frequent cause of constipation, in addition to slow transit [3, 4]. Patients with dyssynergic constipation have poor relaxation of the pelvic floor muscles and/or insufficient abdominal pressure thrust, thus making it difficult to pass stool. Over time, as more stool is retained due to difficulties with expulsion, the rectal stool burden increases, which may lead to painful straining and eventual fecal incontinence from overflow.
Dyssynergic defecation, or abnormal muscular coordination of defecation, has been shown to be a factor in persistence of constipation  and can be diagnosed using anorectal manometry (ARM) [4, 6, 7]. The ARM studies defecation dynamics using a catheter with a rectal balloon at its tip and pressure sensors to detect anal canal pressure/relaxation patterns while a patient simulates defecatory movements (bear-down maneuver). Recently, 3-D ARM has come into use providing additional understanding of the sphincter anatomy  and the pelvic floor muscular function, including the puborectalis sling . With these advances in technology, understanding of childhood constipation may be further elucidated, and therapy can now refocus on possible intervention and treatment of defecation dynamics. However, it should be noted that there is recent concern that this newer probe may provide inaccurate results, and further understanding of the benefits and utility of this newer equipment is being evaluated currently .
The current management for childhood constipation involves a combination of medication, dietary interventions, and behavioral modification . Medications, including laxatives and stimulants, are used to soften the stool and promote colonic contractions to induce increased frequency of soft bowel movements. Additionally, patients are also instructed to complete scheduled toilet sitting and to perform Valsalva techniques such as blowing a balloon while defecating.
Multiple studies on the use of biofeedback and pelvic floor physical therapy (PFPT) to regain control of the pelvic muscles and achieve normal defecation dynamics have proven efficacious in the adult population [12, 13]. Therefore, in adults with chronic constipation, it is now the standard of care to undergo ARM evaluation, followed by therapy for dyssynergic defecation prior to further medical or surgical interventions . PFPT has not specifically been studied in children, but as part of the PFPT arsenal, it may include biofeedback. Unfortunately, studies conducted on the use of biofeedback in the pediatric population almost 2 decades ago were unconvincing as initial short-term studies showed promising results [15, 16], but long-term studies showed no difference in efficacy compared to laxatives alone . Thus, based on the negative results from these long-term studies, the current standard of care in pediatric constipation does not include biofeedback or PFPT as treatment of any form of pelvic outlet obstruction .
Over the past several decades, there have been advances in the treatment of outlet obstruction constipation in pediatrics, often drawing on the techniques used in adults . Pediatric PFPT has and continues to be developed with modifications that address the unique traits of children and the involvement of parental support. Additionally, biofeedback technology now uses external rather than internal sensors – an important advancement in the psychological impact of children who are often traumatized and highly anxious when confronted with anal instrumentation. The aim of this study is to examine the effect of PFPT involving novel techniques to address and treat outlet obstruction and dyssynergic defecation in the pediatric population.
We completed a retrospective chart review over 15 months (September 2012 – December 2013) of patients aged 5–18 years with chronic constipation based on the ROME III criteria . Specifically, we focused on those who were medication-dependent or unresponsive to medical intervention with continued symptoms for more than 3 months. The cohort included only those patients who completed a clinically indicated 3-D ARM, with results that were highly suggestive of dyssynergic defecation and were thus referred to PFPT. We received approval from the Institutional Review Board prior to data collection.
The solid-state 10 mm in diameter 3D manometry probe (ManoScan 3D; Sierra Scientific Instruments, Los Angeles, CA, USA) was used, which has a pressure sensing segment of 64 mm in length, composed of 256 sensors spaced at 2 mm intervals arrayed as a continuous grid. An additional sensor is contained within the rectal balloon and detects rectal pressure. The probe is inserted into the rectum of the patient with pressure recorded in the rectum, anal canal, and atmosphere. An indicator on the probe was used to align the probe appropriately throughout the entire study.
Each study consisted of multiple maneuvers with the awake unsedated child in the left lateral decubitus position. A child life specialist was present to help alleviate the child’s fear and anxiety. The examination included recording of the resting anorectal pressure and squeeze attempts. The beardown maneuver (simulated defecation) was completed, with repeated measurements at increasing balloon inflation, from 0 up to 60 mL, focusing on canal relaxation and rectal thrust pressures. The rectoanal inhibitory reflex was then elicited by balloon distention, followed by internal rectal sensation evaluation. The probe was removed and a balloon (balloon expulsion catheter, MUI corporation) was inserted and insufflated to 60 mL to evaluate ability to pass into the commode within 1 min . Pelvic floor dyssynergia was diagnosed based on criteria initially determined in the adult population, when: (1) no relaxation of the anal canal during beardown despite progressive anal canal insufflations with the balloon and/or [21, 22] (2) insufficient increase in rectal pressure during beardown , or (3) failure to expel balloon into commode . Additionally, patients were classified based on pediatric data as being hypersensitive by initial sensing of balloon inflation <25 mL and hyposensitive when sensing balloon inflation at >80 mL inflation [24‒26].
The patients were classified into 2 categories: treatment and nontreatment group. Those in the treatment group received at least 3 sessions of physical therapy. Based on medical provider experience, patients needed several sessions in order to receive benefit from the intervention. Those in the nontreatment group were treated with medical treatment only. Both groups of patients had previously received at least 3 months of medication therapy with continued symptoms and remained on the medications during the ARM. Those who received physical therapy for only 1 or 2 sessions were excluded from the study.
Physical Therapy and Biofeedback
The physical therapy intervention consisted of an intake session where the patient was evaluated for motor control, strength, and endurance of the primary muscle groups involved with defecation: respiratory diaphragm, transversus abdominus, and the pelvic floor musculature. The follow-up visits included goals of maximizing the patient’s ability to correctly engage the diaphragm during respiration, maximize transversus abdominus strength and endurance, and master isolation of the pelvic floor musculature. This intervention occasionally included utilization of surface electromyography (EMG) via the pathway TR series EMG (The Prometheus Group, Dover, NH, USA). Patients were also assessed for diastasis abdominus rectus, a separation of the abdominus muscle, which is frequently associated with low muscle tone.
The child was also trained in normal toileting mechanics, by engaging the transversus abdominus and elongating the pelvic floor. Patients were encouraged to complete between 3 and 8 sessions of PFPT. A complete description can be found in the adjoining online supplementary material (see www.karger.com/doi/10.1159/000500121).
Patients were seen weekly or bi-weekly for follow-up physical therapy sessions. The physical therapist evaluated the patient based on exercises completed during the session, patient symptoms, and the parent’s input. Their assessment of the patient’s progress was documented in the patient’s chart. During the treatment period, the patients in both groups continued to have periodic follow-up with the gastroenterologist with medication adjustments if necessary.
Patient charts were reviewed for initial symptoms of constipation, time course of treatment, and reports from the physical therapist and physician on the patient’s symptoms and clinical response. Within the physical therapy reports, presenting symptoms, physical exam, and response to treatment were also obtained. Outcomes that were assessed included response defined by: decreased straining, decreased fecal incontinence, or reduction in the amount of medication use per parental and patient report as documented in the physical therapist notes at the completion of treatment course, and by the gastroenterologist note, both occurring 2–3 months after completion of the ARM. In nontreatment patients, the outcome was based on the gastroenterologist’s reports. Additional outcome measures analyzed included hospitalizations for stool cleanouts and surgical interventions including cecostomy placement, appendicostomy, diverting ileostomy, or colectomy. Follow-up of patients was completed for up to 1 year via physician reports. Characteristics of patients were obtained, including history of previously diagnosed anxiety disorder, autism, and learning disorders.
Continuous variables were presented as median, and categorical variables were presented as number with percent. Given the nonnormality of the continuous variable, Wilcoxon rank-sum test was used to evaluate the continuous variable, and chi-square test was used to evaluate categorical variables using the Stata program. A significance level of p < 0.05 was used for all statistical analyses.
One hundred and fifty-three patients were screened using ARM, and 73 patients were found to have dyssynergic defecation. Of these, 4 were excluded for completing <3 sessions of physical therapy. The treatment group consisted of 49 patients, and the nontreatment group had 20 patients. The 2 groups were not statistically different in terms of age, gender, or time of follow-up. The majority of patients in both groups were classified as having dyssynergic defecation due to poor relaxation of the anal canal. A smaller percentage of patients had a normal ARM but failed the balloon expulsion test. These types of dyssynergia had a similar frequency in both groups (Table 1). There was no statistical difference between the percentage of patients with rectal hypersensitivity and hyposensitivity in each group.
The patients presented with various symptoms of constipation and treatment history. There was no statistical difference between the 2 groups in the percentage of patients with fecal incontinence, straining, pain, or urinary symptoms. There were a similar number of patients in each group on osmotic and stimulant medications at the time of initiation of PFPT (Table 1).
In the PFPT treatment group, we found that 38/49 (76%) responded, while only 5/20 (25%) of those in the non-PFPT treatment group had a response documented at their most recent visit (p < 0.01; Table 2). Additionally, the treatment group had a statistically significant lower number of patients who proceeded to hospitalization (p= 0.014) and to surgery (p = 0.026) as compared to the treatment group. Overall, with all 3 outcomes measures analyzed, the treatment group had a better response to the PFPT intervention, as compared to the conservatively managed group. Frequency of follow-up care or use of EMG training was not a predictor of response to intervention.
Within the treatment group, the patients were analyzed to assess for specific characteristics of those who responded, including psychiatric and physical comorbidities. Patients completed an average of 5.5 sessions of physical therapy (range 3–14 sessions). Out of the 49 who received the intervention, 38 (76%) responded and 11 (24%) did not. Those with anxiety had a 100% response rate to the physical therapy, as did those with diastasis rectus abdominus, defined as a separation of the abdominus rectus muscle of >2 cm. About half of the patients with autism spectrum disorder who were able to respond to commands and those with learning disorders responded to the intervention (Table 3).
The follow-up time with a physician ranged from 8 to 52 weeks in 21 of the patients. All of the responders seen for follow-up continued to have overall improvement; however, 5 (24%) patients had worsening of symptoms since discontinuation of the physical therapy sessions (Fig. 1). Four out of 5 (20%) of these patients who had worsening of symptoms were no longer completing their physical therapy exercises at home. There were no adverse events related to physical therapy treatment.
This study examined the use of PFPT for the treatment of outlet obstruction constipation in patients with dyssynergic defecation diagnosed using 3D-ARM. We found that there was a significant difference in response, hospitalization rates, and need for colonic surgery between patients who underwent physical therapy and those that did not. In our experience, physical therapy was most useful for patients with anxiety as well as with diastasis abdominus rectus, often associated with low muscle tone. Additionally, patients with autism, as well as those with learning disorders, were frequently able to benefit from the therapy. Although we did not formally study these outcomes, the reports suggest that due to a decrease in fecal soiling, hospitalizations, and surgeries, we hypothesize that PFPT could possibly also lead to an improvement in functioning and decreased medical cost burden.
PFPT is focused on outlet obstruction constipation by improving defecation dynamics. In the adult population, it has been shown that slow transit constipation may be secondary or exacerbated due to prolonged outlet obstruction . Additionally, studies in pediatric patients have shown a correlation between poor muscle tone, pelvic floor dysfunction, and slow transit constipation [28, 29]. These together with our findings suggest that more emphasis should be placed on the treatment of dyssynergic defecation focusing on relaxation of appropriate pelvic muscle groups, increased force in the abdominal muscles, and proper coordination of the diverse muscular groups involved. Current standard therapy using stimulant and laxative medication can promote motility and improve stool consistency; however, this may be only partially beneficial until issues with defecation dynamics are treated.
Furthermore, Rao  found in adult patients with dyssynergic defecation that 31% reported symptoms of constipation since childhood, with an additional 41% of adult-onset patients, associating the initiation of defecation difficulty with a musculo-skeletal event such as trauma or back injury. Additionally, multiple studies have shown the benefit of biofeedback and pelvic muscle training for adult outlet obstruction constipation, making this intervention the standard of care in the adult population [13, 14]. This raises the question that if this type of treatment was started earlier, in a younger population, could it prevent years of constipation-associated symptoms and morbidity? Perhaps, patients who respond to pelvic floor treatment as adults could have achieved resolution of constipation in childhood had they received the intervention earlier and, thus, avoided prolonged symptoms into adulthood.
The treatment of fecal incontinence and abnormal defecation dynamics in pediatrics via biofeedback exclusively and with the use of internal probes received attention almost 2 decades ago, but the overall evidence for the long-term benefit was lacking [11, 30]. There have been 4 randomized controlled long-term studies published examining the effect of biofeedback therapy in the pediatric population; however, they have heterogeneous protocols with nonuniform biofeedback interventions and thus may be difficult to interpret together. The largest study by van der Plas et al.  randomizing 192 chronically constipated patients to receive laxatives or biofeedback therapy showed no difference between the 2 groups after 1 year, while Loening-Baucke  studied 129 chronically constipated and encopretic patients showing initial response to biofeedback but convergence to the medication only group over time. A Cochrane review published in 2014 shows insufficient evidence to allow for conclusions of efficacy in adults . The ANMS-EMS consensus statement concludes that biofeedback is beneficial in the adult population, but although some pediatric patients responded, it does not have added benefit over conventional treatment and therefore cannot be recommended . It is important to note that none of these studies on management of abnormal defecation dynamics in children studied the effect of PFPT.
Since these pediatric clinical trial studies have been published, there have been advances in our examination, understanding, and intervention of the pelvic floor. Although adult pelvic physical therapy has been used for more than a decade, pediatric PFPT is a new, emerging field, adapted from the adult intervention with a focus on pediatric needs and inclusion of parental support and teaching. Compared to biofeedback, focused mainly on pressure exertion along the anal canal, PFPT is a more comprehensive approach that teaches the specific appropriate muscle techniques to allow for improved strength and relaxation resulting in enhanced expulsion of stool. As it consists only of physical exercises and no internal anal canal intervention, there are few side effects and thus may be more acceptable to families than medications. An added possible advantage in the pediatric population is that parents are taught how to continue the treatment and exercises at home, thus the effect may be more durable.
Recent improvements in diagnostic technology such as 3-D ARM, pelvic MRI, and defecography can lead to further classification of defecation dynamics including extent of the puborectalis recruitment and delineation of the anatomy of the internal and external sphincter, thus providing additional insight into specific muscle regions that should be targeted during the treatment. The therapists also have access to a surface EMG for muscle localization/mapping rather than use of biofeedback alone, which is thought to be more sensitive in analyzing muscle contraction and relaxation. There are various engaging activities that the physical therapist may use, such as interactive computer games used to encourage dissociation of the pelvic muscle group from the abdominal muscles to produce appropriate muscle dynamics.
Based on the positive report of this and other studies [12, 13] addressing the pelvic floor and outlet obstruction, it may be beneficial to target specific patient populations more likely to succeed with the intervention. In our study analysis, we found that constipated patients with anxiety and/or low muscle tone demonstrated higher response rates than the general cohort. As physical therapy focuses on abdominal wall strengthening, it is clear why those with low muscle tone can derive a benefit as demonstrated in our study. In children, it is common that withholding behavior and outlet obstruction can be perpetuated by fear and stress about using the toilet . Therefore, those suffering from a comorbidity of anxiety may have a greater benefit from learning skills on appropriate muscle groups to reduce contraction and stimulate relaxation, which can be facilitated by a physical therapist. It should be noted though that the physical therapy patients had more frequent follow-up with providers than those on conventional therapy, which may be a contributing factor to their improvement .
Several of the prior studies addressing treatment of defecation dynamics with biofeedback alone showed an initial beneficial response with the intervention which over time was lost, leading to convergence with the control group and ultimately resulting in no difference between groups . Similarly, in our study, there were several patients who had initial responses that decreased over time. This may be due to the fact that the child returns to the same environment or behavior that originally led to the obstructive conduct. Thus, without the reinforcement of physical therapy, the patient may revert to the poor behavior habits as he had before. Additionally, some of the children may have impaired ability to retain information over time as compared to adults. This suggests that in order to establish long-term efficacy patients may need intermittent PFPT follow-up sessions to reinforce the techniques and skills that were taught during the initial training period.
There are several limitations to this study. It is a retrospective study and therefore subject to confounders, particularly when examining differences between the treatment and nontreatment groups. Additionally, the groups were small and therefore lacked power, particularly any complex regression model analysis of the treatment group. Furthermore, the outcome of response is based on chart review of the therapists/physician documentation of their perception of response, which may not correlate with the family’s thoughts. Prior studies have shown that the skill of the therapist and the motivation of the patient may be linked to outcome which was not assessed . Finally, many patients did not follow-up for an extended period of time after the intervention. Although this may be due to a persistent response, it is unclear how long many of the patients remain well over 1 year. It should also be noted that there is recent concern regarding the size of the 3D catheter, suggesting that due to the large diameter and rigidity of the probe, possible increased anxiety is caused during testing, and dyssynergic defecation may be overdiagnosed in patients with 3D ARMs . Unfortunately, there are no pediatric studies at this time comparing diagnosis of dyssynergic defecation in healthy children and those with bowel symptoms, and thus understanding the accuracy of pediatric testing is not known. Although this concern should be examined further, it should be noted that in our study, all patients (control and study group) were evaluated the same way, and abnormalities of defecation that were encountered were consistent between groups, with <50% of all patients meeting criteria for dyssynergic defecation.
Overall, the results of this study suggest that there are benefits of pelvic floor therapy in pediatric patients with outlet obstruction. The encouraging retrospective data reinforce the need to construct a randomized controlled study utilizing the novel pediatric PFPT techniques. A larger sample size to reexamine the long-term effects of pelvic floor training with physical therapy, including reinforcement sessions, should be performed with better patient characterization, particularly focusing on the treatment groups that may have better response such as those with psychiatric comorbidities and those with weak muscle tone.
Statement of Ethics
Subjects (or their parents or guardians) have given their written informed consent. The study protocol has been approved by the research institute’s committee on human research. Animal experiments conform to internationally accepted standards and have been approved by the appropriate institutional review body.
The authors declare no conflicts of interest. This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
C.Z.-K.: Massachusetts General Hospital; Pediatric Gastroenterology. Conceptualized and designed the study, acquired and interpreted data, drafted the initial manuscript, and approved the final manuscript as submitted. B.K.: Massachusetts General Hospital; Gastroenterology. Provided information study protocol, methods, and statistical analysis. Critically reviewed the manuscript and approved final submission. A.B. and E.C.: Massachusetts General Hospital, Department of physical therapy. Edited and revised work, provided information relating to physical therapy intervention including in the appendix. Approved final submission. J.B.-G.: Massachusetts General Hospital; Pediatric Gastroenterology. Analyzed the ARM results. Reviewed and revised the manuscript and approved the final manuscript as submitted.