Objectives. To characterize the clinical features, results of diagnostic testing, and treatment outcomes for children and adolescents with rumination syndrome.
Methods. Review of the medical records for all 147 patients ages 5 to 20 diagnosed with rumination syndrome at our institution between 1975 and 2000. Data are presented as mean ± the standard error of the mean.
Results. Sixty-eight percent were female. Age at diagnosis was 15.0 ± 0.3 years. Symptom duration before diagnosis was 2.2 ± 0.3 years, 73% missed school/work, and 46% had been hospitalized because of symptoms. Before diagnosis, 16 (11%) underwent surgery for evaluation or management of symptoms. Twenty-four (16%) had psychiatric disorders; 3.4% had anorexia or bulimia nervosa. All patients described postprandial regurgitation after almost every meal (2.7 ± 0.1 meals per day). Weight loss was described by 42.2% (median: 7 kg). Additional symptoms included: abdominal pain, 38%; constipation, 21%; nausea, 17%; and diarrhea, 8%. Structural studies were normal. Gastric emptying of solids at 4 hours was delayed in 26 of 56 patients. Esophageal pH testing in 24 patients showed reflux/regurgitation in 54%. Gastroduodenal manometry in 65 patients showed characteristic rumination-waves in 40%. Outcome data (at median follow-up 10 months) were available for 54 patients. Symptoms resolved in 16 (30%) and improved in 30 (56%).
Conclusions. Recognition of the clinical features of rumination syndrome in children and adolescents is essential; the diagnosis is often delayed and associated with morbidity. Extensive diagnostic testing is unnecessary. Early behavioral therapy is advocated, and patient outcomes are generally favorable.
Rumination syndrome is characterized by the effortless regurgitation into the mouth of recently ingested food followed by rechewing and reswallowing or expulsion.1,2 Individuals with classic rumination syndrome typically do not experience heartburn, abdominal pain, or nausea when the regurgitation occurs. The syndrome is most commonly seen in infants and the developmentally disabled. However, rumination syndrome does occur in children, adolescents, and adults with normal intelligence.3,4
Individuals with rumination syndrome are often misdiagnosed or undergo extensive, costly, and invasive testing before diagnosis. Insufficient awareness of the clinical features of rumination syndrome contributes to the underdiagnosis of this important medical condition. Rumination syndrome is frequently confused with bulimia nervosa, gastroesophageal reflux disease, and upper gastrointestinal motility disorders including gastroparesis and chronic intestinal pseudo-obstruction. Complications of rumination syndrome include weight loss, malnutrition, dental erosions, halitosis, electrolyte abnormalities, and significant functional disability.3
Rumination syndrome is a clinical diagnosis based on symptoms and the absence of structural disease.3 Although the Rome II diagnostic groups include childhood functional gastrointestinal disorders, only “infant rumination syndrome” is described in detail and pertains to infants with symptom onset before 8 months of age.5 Diagnostic criteria for children and adolescents with rumination beginning after infancy have not been defined.
The purpose of this study was to further characterize the clinical features, results of diagnostic testing, and treatment outcomes for children and adolescents between the ages of 5 and 20 with rumination syndrome. We perceived that review of a large cohort of patients in the 5- to 20-year age group was necessary to develop future consensus criteria for rumination syndrome in children and adolescents.
A computerized diagnostic index was used to identify children and adolescents between the ages of 5 and 20 diagnosed with rumination syndrome at Mayo Clinic Rochester during the 25-year period between 1975 and 2000. One hundred forty-seven patients were identified. Data extracted included demographics, clinical features, past medical history, social history, evaluation, treatment, and outcome. The study was approved by the Mayo Institutional Review Board, and authorization for the use of the medical records for research purposes was confirmed before access was provided to the medical records.
Descriptive data are displayed as number of patients (percent), or mean ± the standard error or the mean (n = number of patients with data available if less than the total number, 147). The Mann-Whitney U test was used for comparisons of means between groups. The χ2 test was used for comparisons of proportions.
Of the 147 patients identified, 68.0% were female, and 97.2% were Caucasian. The majority (70.8%) were from the midwestern United States (Minnesota, Iowa, Wisconsin, Nebraska, North Dakota, Illinois, Indiana, South Dakota, and Michigan). These demographics reflect the usual ethnic and geographic distribution of patients attending Mayo Clinic Rochester.
Past Medical History
Forty-eight (32.7%) patients had prior diagnoses of gastrointestinal illnesses; 49 (33.3%) had other medical diagnoses; and 25 (17%) had concomitant psychiatric disorders including depression (6.1%), anxiety disorder (3.4%), attention-deficit/hyperactivity disorder (3.4%), adjustment disorder (1.4%), obsessive compulsive disorder (0.7%), somatoform disorder (0.7%), and posttraumatic stress disorder (0.7%). Anorexia nervosa and bulimia nervosa were documented in 2 (1.4%) and 3 (2.0%) patients, respectively. A history of physical and/or sexual abuse was documented in 4 (2.7%) patients. Developmental disabilities were present in 7 patients including developmental delay in 5, autism in 1, and dyslexia in 1.
Past Surgical History
Before diagnosis of rumination syndrome, 16 (11%) patients had abdominal operations performed elsewhere for evaluation or management of associated symptoms: 6 appendectomies; 5 cholecystectomies; 3 laparotomies; 1 duodenal enterostomy; and 1 patient who underwent multiple operations including pyloroplasty, partial colectomy with ileostomy, and cholecystectomy. The clinical history of the latter patient was reported elsewhere.2 None of the patients had symptomatic improvement postoperatively.
Social History and Family History
Eighty-five percent of patients lived at home with their families, and 12% were college students who lived on campus (n = 131); the remaining 3% were married, living with a significant other, or living with a foster family. Relatively uncommon was the use of alcohol (13 patients), tobacco (14 patients), and recreational drugs (4 patients).
Regarding school performance, 90% of the 39 patients with data available earned A and B grade point averages. Using adjectival terms from the medical records to categorize scholastic performance (n = 73), 18.8% were excellent students, 58.9% were good students, 2.7% were fair students, and 9.6% were poor students. (Of note, data regarding school performance were not available for the 7 patients with developmental disabilities.)
Clinical Features of Rumination Syndrome
The age at diagnosis was 15.0 ± 0.3 years as shown in Fig 1. Symptom duration before diagnosis was 2.2 ± 0.3 years (n = 126). Age at symptom onset was 12.9 ± 0.4 years (n = 126). Males were younger at symptom onset (11.0 ± 0.8 vs 13.8 ± 0.5; P = .003) and at the time of diagnosis (13.2 ± 0.7 vs 15.8 ± 0.3 years; P = .001) than females.
The body mass index at the time of diagnosis was 21.5 ± 0.4 kg/m2 (n = 119); no gender differences for body mass index were detected (21.2 ± 0.9 kg/m2 for males vs. 21.7 ± 0.4 kg/m2 for females; P = .3).
Although ‘vomiting’ was often the presenting complaint, all but 1 of the 147 patients described “effortless” postprandial regurgitation. Episodes of regurgitation occurred after 2.7 ± 0.1 meals per day (n = 34). By the patients’ estimations, regurgitation commenced 21.1 ± 5.7 minutes postprandially, median 10 minutes (n = 23), and persisted 72.3 ± 9.4 minutes (n = 20). Reswallowing of regurgitated food was described by 25 (17.0%) patients. On the other hand, remastication of regurgitated food was documented in 4 (2.7%) patients. Nighttime episodes occurred in only 3 (2.0%) patients.
In 15 (10.2%) patients, specific stressors were identified just before symptom onset including: 3 with deaths in the immediate family; 3 starting school/college; 2 accidents; 2 onset of sports season; 1 parental divorce; 1 geographical relocation; 1 change in job for father; 1 change in church; and 1 going to camp.
Abdominal pain was an associated symptom in 56 (38.1%) patients. Other digestive tract symptoms included: constipation in 31 (21.1%), nausea in 25 (17.0%), diarrhea in 12 (8.2%), bloating in 6 (4.1%), and dental problems in 5 (3.4%).
Weight loss was reported in 62 (42.2%) patients; the average weight loss in this group was 9.6 ± 1.0 kg, (median: 7.1 kg; n = 56). Patients with weight loss were older at symptom onset (14.5 ± 0.6 years vs 11.6 ± 0.6 years without weight loss; P = .0003) and at the time of diagnosis (16.4 ± 0.4 years vs 13.8 ± 0.5 years without weight loss; P < .0001).
Medications Before Diagnosis
Before evaluation at Mayo Clinic Rochester, 78 (53.1%) patients were prescribed medications specifically for symptoms associated with rumination; on average, these patients received 1.8 ± 0.1 types of medications. Of note, no patients had significant symptomatic improvement with medical therapy. Sixty-six (45.0%) patients had been treated with acid-blocking medications (histamine [H2] blockers and/or proton pump inhibitors) and 56 (38.1%) with prokinetic medications (cisapride, metoclopramide, or erythromycin). Forty-three (29.3%) patients received both acid-blocking and prokinetic medications. Twelve (8.2%) patients had been treated with anti-emetic medications, 10 (6.8%) with antidepressants, and 3 (2.0%) with narcotic pain medications.
Before evaluation at Mayo Clinic Rochester, 5 (3.4%) patients had required supplemental enteral (via tube) or parenteral nutrition for management of symptoms, dehydration, or nutritional support.
Of the 44 patients with data available, 32 (72.7%) had missed school or work because of symptoms. In these 44 patients, rates of school/work absenteeism were not significantly higher in patients with abdominal pain; 83.3% of patients with abdominal pain had missed school/work compared with 60.0% of patients without abdominal pain (P = .2).
Of the 76 patients with data on hospitalizations available in the medical record, 35 (46.1%) had been hospitalized for evaluation of symptoms or treatment of complications related to rumination. Patients with prior hospitalizations were older at symptom onset (15.0 ± 0.8 vs 12.3 ± 0.8 years; P = .02) and older at the time of diagnosis (16.4 ± 0.6 vs 14.6 ± 0.6 years, P = .02) than patients who had not been hospitalized. Of the 74 patients with historical data available regarding weight loss and hospitalizations, the prevalence of weight loss was significantly higher in patients who had been hospitalized (58.8% vs 30.0% in patients without prior hospitalizations; P = .02).
Evaluation of Symptoms Before Diagnosis of Rumination Syndrome
Laboratory studies were performed in 93 (63.3%) patients; results were normal with the exception of hypokalemia in 1 patient who had been previously diagnosed with Bartter’s syndrome.
Results from structural studies performed both at Mayo Medical Center and elsewhere before consultation at Mayo including upper gastrointestinal barium series (performed on 53.7%), esophagogastroduodenoscopy (53.7%), abdominal ultrasound (31.3%), small bowel follow-through (29.3%), abdominal computed tomography (CT) scan (18.3%), barium enema (4.9%), flexible sigmoidoscopy (4.1%), video swallow evaluation (2.0%), and colonoscopy (1.4%) were noncontributory. Brain imaging was done in 30 (20.5%) patients; results were normal by CT scan in 16, magnetic resonance imaging scan in 11, and CT and magnetic resonance imaging scans in 3.
Scintigraphic gastric emptying at 4 hours was delayed in 26 of the 56 patients assessed. Orocecal transit at 6 hours, a surrogate marker for small bowel transit, was assessed in 28 patients and was delayed in 13 patients. However, 7 of the 13 patients with delayed orocecal transit also had delayed gastric emptying at 4 hours. Scintigraphic colonic transit was generally performed in patients with constipation or diarrhea. The method and normal values have been described elsewhere.6 Colonic transit was abnormally slow in 5, and accelerated in 1 of the 12 patients tested.
Esophageal pH testing was performed in 24 patients; results were normal in 45.8%, and showed reflux/regurgitation in 54.2%. None of the patients tested had significant nocturnal or supine reflux, and the majority with reflux/regurgitation had numerous, brief postprandial episodes. Regurgitation/reflux occurred >200 times after a single meal in 1 patient. Gastroduodenal manometry was performed in 65 patients; 55.4% were normal, 40% had rumination-waves, and 4.6% had antral hypomotility. Regurgitation/vomiting during the gastroduodenal manometry assessment occurred only in patients with rumination-waves on manometry. Esophageal manometry testing performed using the station-pull-through technique at the conclusion of the gastroduodenal manometry assessment was normal in all but 2 patients, 1 of whom had evidence of a hypertensive lower esophageal sphincter and the other a hypotonic lower esophageal sphincter.
The average number of the above diagnostic tests performed per patient was 3.3 ± 0.2 (median: 3; range: 0–8).
Treatment and Outcomes
Behavioral treatment was recommended for 127 of the 147 patients, and was undertaken at Mayo Clinic Rochester in 71 patients. Medical treatment was recommended for 25 of the 147 patients including: H2-blocker in 9, proton pump inhibitor in 3, prokinetic medications (cisapride, erythromycin, or metoclopramide) in 11, tricyclic antidepressants in 3, hyoscyamine in 1, and anti-emetic medication in 1. None of the patients were treated with supplemental enteral (via tube) or parenteral nutrition.
Outcome data were available for 54 (36.7%) patients. The mean duration of follow-up was 10.2 ± 1.4 months. The baseline characteristics and clinical features of rumination for these 54 patients were similar to those of patients without outcome data available. The only significant differences between the 2 groups were a slightly older age at diagnosis (15.8 ± 0.5 vs 14.5 ± 0.4 years; P = .02) and a higher prevalence of surgeries before diagnosis of rumination syndrome (19.2% vs 5.6%; P = .02) for patients with outcome data available. Therefore, the outcome data are likely representative of the group as a whole.
Fifty-two of the 54 patients with outcome data available had received behavioral treatment consisting of biofeedback, relaxation training, instruction in diaphragmatic breathing, and/or cognitive behavioral therapy. Of the 46 patients in this group who underwent behavioral treatment at Mayo Clinic Rochester, the number of treatment sessions with a behavioral medicine professional was 2.7 ± 0.5 (median: 1; range: 1–19). Overall, symptoms resolved in 16 (29.6%), and improved in 30 (55.5%); therefore, a positive impact on symptoms was noted in 85.1%. In 12.9%, no improvement in symptoms was noted at the time of follow-up. The 2 patients who did not receive formal behavioral therapy consultation received reassurance and an explanation of the nature of their condition, and were noted to have improvement of symptoms at follow-up. One patient who also suffered from Bartter’s syndrome died during the course of follow-up.
Patients with symptomatic improvement had significantly higher body mass indices at the time of diagnosis than those whose symptoms remained unchanged (body mass index: 21.9 kg/m2 vs. 19.0 kg/m2; n = 46; P = .02). No other baseline characteristics or clinical features had a significant impact on outcome.
In adults, the diagnosis of rumination syndrome is usually based on eliciting classical symptoms in the absence of structural disease.3 Unfortunately, our experience at a tertiary care center suggests that pediatric and adolescent patients with rumination syndrome often undergo extensive, costly, and invasive testing, and are frequently misdiagnosed as having gastroesophageal reflux disease or gastroparesis. We believe that insufficient awareness of the clinical features of rumination syndrome in pediatric and adolescent patients contributes to the difficulty in diagnosing this important medical condition. Formal diagnostic criteria for rumination syndrome occurring in children and adolescents with onset beyond infancy have not been defined in the consensus criteria.1
In the current study, gastric emptying was delayed in 46% of patients assessed. However, there are a number of caveats in the interpretation of scintigraphic gastric emptying tests in children and adolescents with rumination syndrome. Delayed gastric emptying may result from regurgitation and reswallowing of food, which is then delivered intermittently or later to the distal stomach for digestion and mixing. Another caveat in the interpretation of the scintigraphic gastric emptying test is the potential for expulsion of the radiolabeled meal; when the majority of the test meal is expelled, the study cannot be completed. Although an assessment of gastric emptying may help to exclude a significant stomach dysmotility and a normal result may reassure the patient and family, abnormal results should be interpreted carefully in the setting of rumination.
In contrast to the relatively common observation of delayed gastric emptying in the patient population reported, antral hypomotility was extremely rare on formal testing with manometry. Although gastroduodenal manometry has been advocated as a diagnostic test in the setting of rumination,7 only 40% of the 65 patients assessed in our study had characteristic R-waves or simultaneous pressure spikes across all sensors, suggesting an extraintestinal stimulus and contraction of the abdominal musculature. In addition, gastroduodenal manometry is an invasive test requiring significant technical and interpretive expertise, and is only available in tertiary care centers.
In general, we do not advocate esophageal pH testing in patients with rumination, because quantitatively abnormal ‘reflux’ is a consequence rather than a cause of the symptoms. When prolonged pH testing is performed for suspected gastroesophageal reflux disease, rumination syndrome is characterized by the predominance of postprandial regurgitation of acid, rather than supine or nocturnal reflux, both of which occur in classical gastroesophageal reflux disease.8,9
Abnormal relaxation of the lower esophageal sphincter, a common cause of classical gastroesophageal reflux disease in pediatric patients,10 is also involved in the process of rumination. Thurmshirn et al11 showed the mechanism of rumination involves relaxation of the lower esophageal sphincter in response to lower pressures in the fundus, and increased gastric sensitivity. In healthy adults, there is active contraction at the esophagogastric junction and increased lower esophageal sphincter pressure during periods of increased intraabdominal pressure;12 tonic contraction of the crural diaphragm is the proposed mechanism for this response. This mechanism is altered in the setting of rumination syndrome as transient lower esophageal sphincter relaxations occur after abdominal straining events.13
In addition to postprandial regurgitation, many patients in our study had other gastrointestinal symptoms including abdominal pain (38.1%), constipation (21.1%), and nausea (17.0%). The presence of additional gastrointestinal symptoms may induce physicians to recommend more extensive diagnostic testing. However, our study demonstrates that such testing is not helpful in the presence of typical clinical features of rumination.
Weight loss was also commonly associated with rumination, despite normal body mass indices at the time of diagnosis. Considering the female predominance of the condition and the frequent occurrence of weight loss, classical eating disorders such as anorexia nervosa and bulimia nervosa should be considered in the differential diagnosis. However, in our study population, eating disorders had been diagnosed in only 3.4% of patients. Previous studies have described a history of eating disorders (primarily bulimia nervosa) in a larger proportion of adult patients with rumination syndrome.14–16 Although weight loss is a concerning symptom in children and adolescents, we do not believe that weight loss is an indication for more exhaustive diagnostic testing in the presence of classical clinical features of rumination syndrome.
In general, rumination syndrome is a “benign” condition.17 However, our study demonstrates significant functional disability related to weight loss, school and work absenteeism, hospitalization, and extensive diagnostic testing in pediatric and adolescent patients with rumination. Early recognition of the clinical features of rumination and referral for behavioral treatment help to reduce adverse consequences in this patient population.
The current behavioral treatment for rumination syndrome at our medical center consists of habit reversal using diaphragmatic breathing as the competing response.18 Habit reversal is an empirically supported behavioral paradigm wherein a problematic or target behavior can be eliminated by the consistent use of an incompatible or competing behavior. In this instance, one cannot perform the target behavior and the competing response at the same time. In the case of rumination, consistent cued practice of diaphragmatic breathing during rumination effectively eliminates rumination, after proper training in both habit reversal and diaphragmatic breathing. The vast majority of patients in our study had significant symptomatic improvement, and many had resolution of symptoms after behavioral treatment. The median number of treatment sessions required was 1. Rumination complicated by comorbid medical, psychological, or psychiatric conditions may require additional therapeutic interventions.
There is a need for consensus clinical criteria for the diagnosis of rumination syndrome in children and adolescents. We would propose the criteria in Table 1. With typical clinical features, extensive diagnostic testing including gastroduodenal manometry and esophageal pH testing is unnecessary. Early intervention with behavioral modification is advocated. Outcomes in children and adolescents who have received behavioral therapy have been so favorable at our institution (>80% success) that a formal, controlled clinical trial has not been pursued. Collaboration between gastroenterologists, pediatricians, and psychologists in addition to educating patients and family members are key elements to a successful outcome.
This work was supported in part by grants R01-DK54681 and K24-DK02638 (to Dr Camilleri) and by General Clinical Research Center grant (#RR00585) from the National Institutes of Health.
We thank Cindy Stanislav for excellent secretarial assistance.
- ↵Clouse RE, Richter JE, Heading RC, Janssens J, Wilson JA. Functional esophageal disorders. Gut.1999;45(suppl 2) :II31– II36
- ↵Rasquin-Weber A, Hyman PE, Cucchiara S, et al. Childhood functional gastrointestinal disorders. Gut.1999;45(suppl 2) :II60– II68
- ↵Tytgat GN, Nio CY, Schotborgh RH. Reflux esophagitis. Scand J Gastroenterol Suppl.1990;175 :1– 12
- Larocca FE. Rumination in patients with eating disorders. Am J Psychiatry.1988;145 :1610
- ↵Levine DF, Wingate DL, Pfeffer JM, Butcher P. Habitual rumination: a benign disorder. Br Med J (Clin Res Ed).1983;287 :255– 256
- Copyright © 2003 by the American Academy of Pediatrics