The rumination syndrome: Clinical features rather than manometric diagnosis

GASTROENTEROLOGY1995;108:1024-1029

The Rumination Syndrome: Clinical Features Rather Than Manometric Diagnosis MICHAEL D. O'BRIEN,* BARBARA K. BRUCE,* and MICHAEL CAMILLERI* *Gastroenterology Research Unit and tDivision of Psychology, Mayo Clinic and Mayo Foundation, Rochester, Minnesota

Background/Aims: Rumination is infrequent in adults

of normal mental capacity. Upper gastrointestinal manometry reportedly confirms the diagnosis. Clinical characteristics, treatment(s), and outcomes of these patients are unclear. Methods: We assessed 38 adults and adolescents with rumination between 1987 and 1994. Esophagogastroduodenal manometric recordings (n = 36; 3-hour fasting and 2-hour postprandially) were reviewed; follow-up information was obtained from mailed questionnaires. Results: Patients saw a mean of five physicians and had symptoms for a mean of 2,75 years before diagnosis. Features included daily, effortless regurgitation of undigested food starting within minutes of meals. Weight loss was substantial (mean, 29 Ib) in 42% of patients. Seventeen percent of female patients had a history of bulimia. Manometry confirmed the clinical diagnosis in 33% but was otherwise normal in all. Of 16 patients who respondedto our questionnaires of 29 with > 6 months of follow-up (average, 35 months), 12 reported subjective improvement. In 14, the behavior persists. Conclusions: The rumination syndrome is underdiagnosed. With typical clinical features, gastroduodenal manometry seems unnecessary to confirm the diagnosis. Diagnosis and reassurance are important in management to avoid unnecessary tests and treatments. mination is a normal component of the digestive rocess in ruminant animals such as cattle, sheep, and goats. Improved digestibility of ingested material is facilitated by regurgitation with subsequent rechewing and reswallowing of food particles. The rumination syndrome (also merycism or merycasm) is a term used to describe a similar behavior in humans. Rumination is defined as regurgitation of partially digested food that is subsequently reswallowed or ejected. The etiology and physiological mechanisms of rumination in humans is poorly understood. In animals, motility of the rumen and reticulum (the first two chambers of the four-chambered stomach in which fermentation occurs) is regulated by two centers in the lower brain stem) Postprandially, retrograde peristalsis in the rumen carries solid ingesta into the cardia region of the reticulum. Lower esophageal

sphincter relaxation ensues, and retrograde esophageal peristalsis carries the material back into the mouth. Rechewing diminishes particle size and enhances exposure of the food's surface for further digestion. In humans, reverse peristalsis in the esophagus has not been described. This has necessitated formulation of different hypotheses to explain the difference between animals and humans with respect to this unusual phenomenon. It has been suggested in manometric studies that rumination results from increased intra-abdominal pressure by abdominal walt muscle contraction and simultaneous voluntary lower esophageal sphincter relaxation.2 4 Others have postulated that forced inspiration against a closed glottis (or Miieller's maneuver) triggers the regurgitation) These hypotheses have not been proved, and it is equally unclear whether voluntary relaxation of the lower esophageal sphincter, a normally involuntary process, is posslble~ The rumination syndrome has been sited frequently in infants, children, and mentally deficient individuals. 6-13 It has also been speculated that the prevalence of this condition is higher in patients with bulimia nervosa. 14'15 This has led some psychiatrists to classify rumination syndrome as an atypical eating disorder. Rumination in adults of normal intelligence is reportedly rare, although difficulty in diagnosis and lack of awareness of its existence indicate that this may not be strictly so. Furthermore, little is known about the natural history or efficacy of treatment in these patients. Upper gastrointestinal manometry has been reported to show a characteristic pattern that confirms the diagnosis, 3 namely, the presence of simultaneous pressure waves at all levels of the recording (R waves), which indicates artifact caused by an abrupt increase in intra-abdominal pressure due to contraction of the muscles of the abdominal wall. R waves are associated with regurgitation and/or decreases in pH as measured by a pH probe placed in the distal esophagus. © 1995 by the AmericanGastroenterologicalAssociation 0016-5085/95/$3.00

April 1995

M a t e r i a l s and M e t h o d s We report on 38 adolescents (age, -----t3 years) and adults of normal mental capacity with the rumination syndrome seen at our institution between 1987 and 1994. Patient characteristics and clinical symptoms were obtained from carefully documented clinical records. Patients were screened for the presence of butimia nervosa, either active or quiescent. Natural history, treatment, and outcome of treatment data were gathered from 16 patients who responded to a followup questionnaire. Manometric tracings for 36 patients were reviewed, and the use of this test in facilitating the diagnosis was analyzed.

Clinical Data Thirty-eight patients (mean age, 27 years; range, 1 3 65 years; 29 female, 9 male) with the rumination syndrome were identified using a computerized diagnostic index at the Mayo Clinic (Rochester, MN). To obtain follow-up information, a questionnaire was sent to the 29 patients in whom at least 6 months (mean, 39 months; range, 6 - 7 4 months) had passed since initial evaluation at our institution. Of these 29 patients, 16 responded to the questionnaire. Results of biochemical and hematologic indices and radiographic and endoscopic evaluations were extracted from the medical records.

Manometric Studies Upper gastrointestinal manometry was performed in 36 patients with a clinical diagnosis of the rumination syndrome as described previously. ~6 In 20 patients, an esophageal pH probe was placed alongside the manometric tube and fixed with its tip 5 cm above the gastroesophageal junction. Patients were studied for 5 continuous hours, 3 hours fasting and 2 hours following ingestion of a standard 535-kcal meal consisting of chicken, potato, and tapioca pudding. During the study, the patients were instructed to press an event marker during any episodes of regurgitation or vomiting for later correlation with the manometric data. At the conclusion of the study, all patients underwent esophageal manometry using the stationary pull-through technique.

Analysis of Manometric Tracings The manometric tracings were analyzed visually. We determined the number and amplitude of synchronous pressure spikes that were recorded at all levels in the stomach and small intestine surveyed. These R waves were correlated with regurgitation as recorded by the patients and with decreases in pH in the 20 patients in whom pH was recorded concomitantly. All tracings were evaluated for the presence of an underlying motility disturbance; specifically, postprandial antral hypomotility was assessed to rule out gastroparesis as a cause of regurgitation in these patients. Esophageal manometry was used to assess upper and lower esophageal sphincter tone and relaxation and amplitude and propagation of contractions of the esophageal body.

CLINICAL FEATURES OF THE RUMINATION SYNDROME

1025

Results Clinical Data Results of biochemical studies performed in this group of patients were uniformly normal with the exception of liver function test abnormalities and the presence of thyroid autoantibodies in 2 individuals, 1 female with autoimmune hepatitis and 1 male with autoimmune thyroiditis. One patient had iron-deficiency anemia. Structural studies of the upper gastrointestinal tract by esophagogastroduodenoscopy and/or contrast radiographic studies were performed and did not show a structural abnormality in any patient. Patients were seen by an average of five physicians (range, 1 - 8 ) before diagnosis of the rumination syndrome. Patients reported having experienced symptoms for an average of 3.5 years (range, 0 . 5 - 1 3 years) before the diagnosis was made. The diagnoses most often confused with this entity were gastroesophageal reflux, chronic vomiting of undetermined etiology, hiatal hernia, and gastric emptying disorder (or gastroparesis). All patients reported experiencing episodes of repetitive regurgitation of undigested or partially digested food. The character of the regurgitant was uniformly described as lacking an acidic, bitter, or sour taste. Most reported that this activity started within 10 minutes after eating (n = 18). Only rarely were episodes reported to begin 30 minutes or more after the end of a meal (n = 4). In a few patients in whom the duration of episodes was documented, typical ruminating activity lasted between 1 and 2 hours postprandially. The number of episodes of rumination ranged from three times per week (n = 1) to daily (n = 37); in 15 patients, the regurgitation occurred after every meal. Weight loss was present in 42% of ruminators and averaged 29 lb (range, 6 - 1 5 0 lb). In the rest of the patients, weight loss was not documented in patient records. Other symptoms were rarely reported: heartburn (n = 5), nausea (n = 5), abdominal pain (n = 3), belching (n = 1), bloating (n = 1), chest pain (n = 1), hiccups (n = 1), and gas (n = 1). No one reported frank vomiting or retching. Interestingly, two individuals who reported heartburn noted that the onset of this s y m p t o m heralded the disappearance of regurgitation. Two patients were receiving enteral feeding, and 1 female patient was receiving central parenteral nutrition, which was complicated by 2 documented episodes of sepsis related to the central lines. Five of the 29 female patients (17%) were noted to have a history of bulimia nervosa, although all denied this behavior currently. Two individuals had a diagnosis of depression and were receiving therapy for this condition at the time of evaluation.

1026

O'BRIEN ET AL.

GASTROENTEROLOGY Vol. 108, No. 4

Similarly, esophageal manometry did not show any abnormalities of upper and lower esophageal sphincter function or esophageal body contractions during wet swallows using 5-mL water boluses.

pHprobe

antroduodenal

Follow-up

4

desc. duodenal distal duodenal proximal jejunal

regurgitation

II

t I II I 4 minutes

Figure 1. Antroduodenal manometric tracing and monitoring of pH in the distal esophagus in a patient with the rumination syndrome. Note the concurrence of regurgitation (arrows) with decreases in intraesophageal pH and R or simultaneous waves consistent with increased intra-abdominal pressure, However, two R waves (*) are unassociated with regurgitation or decreases in intraesophageal pH.

Data on follow-up (mean duration, 35 months; range, 7 - 7 4 months) were requested from 29 patients and obtained from the 16 patients who responded to our questionnaire regarding the therapeutic interventions recommended by their physicians at the Mayo Clinic and elsewhere. Six received prokinetics, 7 received antacid therapy, 3 underwent behavioral therapy, 2 underwent psychotherapy, and 2 received combined behavioral therapy and psychotherapy. Only 2 of 16 patients reported cessation of rumination symptoms; 1 received chiropractic therapy, and the other patient did not believe resolution of symptoms was related to any specific therapy. A total of 12 of the 16 patients who responded to our questionnaire reported subjective improvement in symptoms; however, 14 continued to experience rumination (mean, 3 years after initial consultation at the Mayo Clinic). Four patients stated that their symptoms were unchanged or worse. No therapy was deemed effective enough to predict any likelihood of a consistent response in these patients.

Discussion The preferred mode for handling the regurgitated contents in those (n = 26) reporting this was as follows: 50% (13 of 26) reswallowed boluses, 27% preferred spitting out boluses, and 23% engaged in both behaviors, mostly depending on the social situation at the time of rumination episodes.

Manometry Upper gastrointestinal manometry studies were analyzed visually for the presence of R waves as well as simultaneous decreases in pH in those with concomitant pH probe placement. R waves were correlated with regurgitation as reported by patients using the event marker. A visual representation of these events is shown in Figure 1. Results are summarized in Table 1. R waves were observed in 17 of the 36 patients who underwent manometric evaluation. However, R waves were associated with regurgitation or decrease in pH in only 12 of 36 (33%) patients. None of the manometric studies showed a pattern to suggest an intrinsic disturbance of normal gastric or proximal small bowel motor function (e.g., myopathic or neuropathic). Postprandial antral hypomotility was not observed in any manometric tracing.

The rumination syndrome is an underdiagnosed condition in adolescents and adults of normal mental capacity. Difficulties in arriving at the correct diagnosis may be caused by the lack of awareness of this entity among physicians. Patients may unintentionally add to this confusion by using such terms as vomiting when describing their symptoms. This underscores the importance of obtaining an accurate clinical history in patients complaining of chronic vomiting. Our study shows that clinical features that suggest

Table 1. Manometric Evaluation Upper gastrointestinal (n = 20) R waves associated R waves associated R waves associated R waves only Normal Upper gastrointestinal probe (n = 16) R waves associated R waves only Normal

manometry with esophageal pH probe with regurgitation and decreased pH with regurgitation only with decreased pH only

7 2 2 2 7

manometry without esophageal pH with regurgitation

1 3 12

April 1995

rumination are typical and include repetitive regurgitation of gastric contents starting within minutes of a meal. These episodes may last 1 - 2 hours, but they contrast with the typical timing of postprandial vomiting in patients with gastroparesis, who tend to start vomiting later postprandially. The regurgitant consists of undigested, easily recognizable, or partially digested food, but it is not sour or bitter in taste. The phenomenon reported by 2 patients, who noted that heartburn signaled the abatement of rumination symptoms, has been noted previously by others .7'18 but has also been refuted by another group. 19 Rumination, unlike vomiting, is rarely associated with retching and is described by patients as seemingly effortless in many instances. Furthermore, patients must make a conscious decision regarding the regurgitant once it is present in the oropharynx. The choice to reswallow the food is more common than spitting out but is not the only choice made by these patients. Some patients engage in both activities, depending on the social situation at the time of rumination. Rumination is typically an everyday and often an every-meal occurrence. Weight loss, possibly severe, may be a prominent feature in ruminators seen in a tertiary referral practice. However, weight loss was often absent in previous series and was not documented in the medical records in the majority (58%) of our patients. It is unclear whether some patients may be able to learn this behavior as a measure to control their weight, much like patients with bulimia nervosa. Seventeen percent of the female patients had a history of bulimia; this leads one to speculate that in some cases it may be a learned behavior in which patients are able to purge themselves without digitally inducing frank vomiting. Alternatively, rumination might better be thought of as a variant of bulimia nervosa or an atypical eating disorder. The group of patients evaluated at our tertiary referral center shows features that suggest referral bias. For example, the high frequency of weight loss and the large number of physicians previously consulted suggest that our patients were particularly affected by the condition. Are there any factors that seem to precipitate this phenomenon? Previous investigators have reported that an acute illness accompanied by nausea and vomiting might have set a reflex sequence in predisposed individuals among their patients. 3 We found no obvious, common precipitating factors in our patients. Thus, it remains unclear whether definable precipitants exist. Rumination seems to be more prevalent in young adults. In addition, females seem to be more commonly affected than males; in our study, the ratio was 3 to 1, which is lower than the typical sex ratio of 9 - 9 . 5 to 1 reported for bulimia nervosa. 2°,21

CLINICAL FEATURES OF THE RUMINATION SYNDROME

1027

Because this syndrome may be undiagnosed for months to years, patients often are subjected to expensive and invasive procedures. Patients may be given enteral or parenteral nutrition, ~ which causes risks of infection and sepsis, as well as undergo surgical interventionsY 7 Excessive testing seems unwarranted in those with a characteristic history for rumination. Because the correct diagnosis can be established clinically, we believe that the use of gastrointestinal manometry for diagnostic purposes is unfounded in those with an accurate history. Diagnostic patterns reported previously, 3 including an R wave that coincides with actual regurgitation, were found in only 33% of our patients. This should be reassuring to primary care physicians and gastroenterologists who may not have access to such diagnostic tools. Equally important, no other manometric indicators of intrinsic motor pathology would have been missed if manometry had not been performed in our patients. The physiological mechanism of rumination remains an enigma, and further physiological studies are needed to address this. It is unclear how the gastric contents are actually retropulsed through the esophagus into the oropharynx because antiperistalsis does not seem to occur in the human esophagus. Hence, it would seem that the retropulsing force might originate in the abdomen as a result of increased intra-abdominal or intragastric pressure. One possible explanation is simultaneous relaxation of the lower esophageal sphincter at the time of increased intra-abdominal pressure, as might occur during a Mtillet's maneuver. This could explain why contraction of the abdominal wall muscles alone is insufficient to bring about rumination. Thus, the observation of an R wave, which suggests a generalized increase in intra-abdominal pressure, is of no pathophysiological significance except as a marker of the event. Because patients do not perform the Mtiller's maneuver while we observe them regurgitating food effortlessly in the laboratory, and because many episodes of regurgitation occur without R waves, we believe that an alternative explanation is likely. Rumination might be the consequence of a learned, voluntary relaxation of the lower esophageal sphincter 2 or diaphragmatic crura, allowing the increased intragastric hydrostatic pressure postprandially and normal phasic and tonic contractions of the proximal stomach to overcome the resistance to regurgitation usually provided by the antireflux mechanisms at the esophagogastric junction. As the regurgitant reaches the distal esophagus, an explanation of why secondary peristalsis is not stimulated to clear the food back into the stomach is needed. It is conceivable that the force and speed of the regurgitating process preclude secondary peristalsis because food reaches the

1028

O'BRIEN ET AL.

pharynx rapidly. AlternatiVely, inhibition of secondary peristalsis may be part of the learned process that results in inhibition of lower esophageal sphincter tone or of the diaphragmatic crura. We believe that a more attractive hypothesis is that rumination is a learned adaptation of the belch reflex. It was originally suggested that belching is produced by swallowing air, evoking a propagated esophageal contraction with reflex relaxation of the lower esophageal sphincter; the elevated intragastric pressure caused by the gas content or abdominal compression thus equilibrates across the relaxed gastroesophageal junction. 22 More recently, others have observed that transient lower esophageal sphincter relaxations were nondeglutitive and were not associated with abdominal straining but occurred more frequently in the upright than the supine position, unlike spontaneous reflux. 23 The duration of transient lower esophageal sphincter relaxations during gas reflux or belching is much longer (mean, 12 seconds) than that associated with deglutition (range, 2 - 4 seconds). 23 Wyman et al. postulated that gastric distention activates a vagal reflex to transiently relax the lower esophageal sphincter during belching. 23 Our hypothesis that rumination represents a learned adaptation of the belch reflex is based on the facts that most regurgitations observed in our patients were not associated with abdominal wall straining and that the durations of the lower esophageal sphincter-c'rus relaxations evoked by the belch reflex seem to be long enough to allow particulate food to be retropulsed from the stomach tO' the pharynx. Thus, the critical factor in the induction of rumination seems to be the resistance: provided by the lower esophageal sphinCter-crura complex. Insights into the pathophysiology of rumination will require an approach that simultaneously measures intragastric pressure, lower esophageal sphincter tone with a Dent sleeve, crural contraction and relaxation, and an event marker for nondeglutitive swallows as with a mylohyoid or submental electromyogram. The intraluminal apparatus also must be anchored well beyond the stomach to ensure it is not repetitively displaced in an orad direction by each episode of rumination and regurgitation. The most efficacious treatment of rumination in adolescents and adults is unclear. There seems to be no consistently effective medical therapy, and there appears tO be a h i g h placebo effect, as with other gastrointestinal functional disorders. Perhaps one of the best clinical approaches is patien t education and reassurance regarding the benign nature of the condition and, in view of the reported success of behavioral modification in limited numbers of adults 'of normal intelligence, ls'24 referral for

GASTROENTEROLOGY Vol. 108, No. 4

behavioral therapy. Equally reassuring to physicians is that 75% of the respondents to o u r questionnaire re~ ported improvement in symptoms at follow-up despite the wide range of therapies and the persistence of rumination in the vast majority of patients. Only 2 of 16 patients had complete remissions at follow-up. It is also possible that improvement may occur without treatment but with careful explanation of the condition, as noted by one individual who had complete resolution of symptoms. A more formal approach similar to that used in patients with eating disorders based on the Diagnostic and Statistical Manual of Mental Disorders, 4th edition, may be more effective than current treatment, but a controlled trial seems necessary before this recommendation can be made. In summary, the rumination syndrome is an underdiagnosed condition in adults and adolescents with normal intelligence. It has a characteristic clinical history, which is most often sufficient for diagnosis. The precise physiological mechanisms that allow regurgitation are not well understood. Treatment strategies have not been tested in patients with normal mental capacity; many patients have some symptom relief over time, but the behavior often persists over several years. If formal behavioral modification, such as that Used with other eating disorders, is to be recommended, a controlled trial is warranted for this disorder, which rarely results in any significant nutritional disturbances.

References 1. Leek DF, Harding RH. Sensory nervo~sreceptorstf~ the ruminant stomach and reflex control of reticula-ruminal motility. In: McDonald IW, Warner ACI, eds. Digestion and metabolism in the ruminant. Armidale, Australia: University of New England, 1 9 7 5 : 6 0 76. 2. Smout AJPM, Breumelhof R. Voluntary induction of transient lower esophageal sphincter relaxation in an adult patient with the rumination syndrome. Am J Gastroenterol 1 9 9 0 ; 8 5 : 1 6 2 1 1625. 3. Amarnath RP, Abell TL, Malagelada J-R. The rumination syndrome in adults. Ann Intern Med 1 9 8 6 ; 1 0 5 : 5 1 3 - 5 1 8 . 4. Breumelhof R, Smout AJPM, Depler ACTM. The rumination syndrome in an adult patient. J Clin Gastroenterol 1 9 9 0 ; 1 2 : 2 3 2 234. 5. Reynolds RPE, Lloyd DA~ Manometric study of a ruminator. J Clin Gastroenterol 1 9 8 6 ; 8 : 1 2 7 - 1 3 0 . 6. Rogers B, Stratton P, Victor J, Kennedy B, Andres M. Chronic regurgitation among persons with mental retardation: a need for combined medical and interdisciplinary strategies. Am J Mental Retard 1 9 9 2 ; 9 6 : 5 2 2 - 5 2 7 . 7. McKeegan GF, Estill K, Campbell B. Elimination of rumination by controlled eating and differential reinforcement. J Behav Ther Exp Psychiatry 1987; 1 8 : 1 4 3 - 1 4 8 . 8. Winton ASW, Singh NN. Rumination in pediatric populations: a behavioral analysis. J Am-AcadChikJ Psychiatry 1 9 8 3 ; 2 2 : 2 6 9 275. 9. Chatoor I, Dickson L, Einhorn A. Rumination: etiology and treatment. Pediatr Ann 1 9 8 4 ; 1 3 : 9 2 4 - 9 2 9 .

April 1995

10. Rast J, Ellinger-Alien JA, Johnston JM. Dietary management of rumination: four case studies. Am J Clin Nutr 1985;42:95-101. 11. Foxx RM, Snyder MS, Schroeder F. A food satiation and oral hygiene punishment program to suppress chronic rumination by retarded persons. J Autism Dev Disord 1979;9:399-412. 12. Mestre JR, Resnick RJ, Berman WF. Behavior modification in the treatment of rumination. Clin Pediatr 1983;22:488-491. 13. Starin SP, Fuqua RW. Rumination and vomiting: a critical review of the behavioral, medical and psychiatric treatment research. Res Dev Disabil 1987;8:575-605. 14. LaRecsa FEF, Della-Fera MA. Rumination: its significance in adults with bulimia nervosa. Psychosomatics 1986;27:209212. 15. Fairburn CG, Cooper PJ. Rumination in bulimia nervosa. BMJ 1984; 288:826-827. 16. Malagelada J-R, Stanghellini V. Manometric evaluation of functional upper gut symptoms. Gastroenterology 1985;88:12231231. 17. Brown WR. Rumination in the adult. Gastroenterology 1968;54: 933-939. 18. Shay SS, Johnson LF, Wong RKH, Curtis DJ, Rosenthal R, Lomett JR, Owensby LC. Rumination, heartburn and daytime gastroesophageal reflux. J Clin Gastroenterol 1986;8:115-126.

CLINICAL FEATURES OF THE RUMINATION SYNDROME

1029

19. Levine DF, Wingate DL, Pfeffer JM, Butcher P. Habitual rumination: a benign disorder. BMJ 1983;287:255-256. 20. Pope HG Jr, Hudson Jl. Eating disorders. In: Kaplan HI, Sadeck BJ, eds. Comprehensive textbook of psychiatry. 5th ed. Baltimore, MD: Williams & Wilkins, 1989:1854-1864. 21. Norman K. Eating disorders. In: Goldman HH, ed. Review of general psychiatry. Norwalk, CT: Appleton & Lange, 1992:327-337. 22. McNally EF, Kelly JE Jr, Ingelfinger FJ. Mechanism of belching: effects of gastric distension with air. Gastroenterology 1964; 64:254-259. 23. Wyman JB, Dent J, Heddle R, Dodds WJ, Toulli J, Downton J. Control of belching by the lower esophageal sphincter. Gut 1990; 31:639-646. 24. Johnson WG, Corrigan SA, Crusco AH, Janell MP. Behavioral assessment and treatment of postprandial regurgitation. J Clin Gastroenterol 1987;9:679-684.

Received July 20, 1994. Accepted November 16, 1994. Address requests for reprints to: Michael Camilleri, M.D., Gastroenterology Research Unit, Mayo Clinic, Rochester, Minnesota 55905.