- Top of page
- Supporting Information
Rumination is the voluntary, albeit unconscious, contraction of the abdominal muscles forcing return of food to the mouth, followed by re-chewing, swallowing or spitting. Although these events are often described as ‘vomiting’, no violent retching is involved and stomach contents are usually returned to the mouth as a series of small volume events rather than one large volume expulsion. In most cases, repetitive belching is also a consequence of abnormal behaviour, either due to excessive swallowing of air (i.e. aerophagia), or suction of air into the oesophagus during forced inspiration (i.e. supra-gastric belching).[2-4]
The diagnosis of rumination and belching disorders can be made from clinical history (Table 1); however, there is often a delay due to lack of awareness of these conditions. As a result, those affected may see many doctors and undergo multiple investigations and treatments before the cause of symptoms is established. Objective diagnosis can be based on the close temporal association of typical symptoms with evidence of abnormal behaviour on physiological studies. Recent advances, such as high-resolution manometry (HRM) combined with impedance, facilitate the detection and diagnosis of dysmotility and dysfunction during and after meals.[6-11] This work has raised awareness of these conditions; however, the aetiology and classification of abnormal behavioural responses to digestive symptoms events have not been well defined. Specifically, the symptoms that provoke abnormal behaviour and the clinical utility of advanced physiological measurement in describing this response remain uncertain.
Table 1. Definition of rumination syndrome and aerophagia (Rome III criteria, must be fulfilled for last 3 months with symptom onset at least 6 months prior to diagnosis) and gastric and supra-gastric belching
|Rumination syndrome in adults|
|Diagnostic criteria. Must include both of the following:|
Persistent or recurrent regurgitation of recently ingested food into the mouth with subsequent spitting or remastication and swallowing
Regurgitation is not preceded by retching
| Supportive criteria |
Regurgitation events are usually not preceded by nausea
Cessation of the process when the regurgitated material becomes acidic
Regurgitant contains recognizable food with a pleasant taste
Troublesome repetitive belching at least several times a week
Air swallowing that is objectively observed or measured
|Venting of air from the stomach, with increase in intra-luminal impedance from distal to proximal esophagus|
|Anterograde movement of gas followed by followed by rapid expulsion (Rapid increase in impedance from proximal to distal, with rapid retrograde return to baseline)|
We propose that rumination and many cases of repeated belching are not distinct conditions, but are caused by a common behavioural response to abdominal pain or other, unpleasant digestive symptoms (Figure 1). This view is supported by several observations. Both rumination and repeated belching have been associated with chronic abdominal pain and the presence of psychiatric disorders.[12, 13] Cases of regurgitation and belching have been associated with abdomino-gastric strain and both conditions have been shown to respond to cognitive behavioural interventions.
Figure 1. Flowchart of trigger symptom, behavioural response and presenting symptom in rumination and its variations.
Download figure to PowerPoint
This hypothesis can be tested. If a common behavioural response is the cause of both conditions, then (i) the presenting complaint (i.e. return of food to the mouth or belching) should be produced by similar behavioural responses to a variety of digestive symptoms (e.g. epigastric pain, bloating, reflux), (ii) a generic behavioural intervention should provide effective management whether the presenting complaint is ‘vomiting/regurgitation’ or ‘belching’ and (iii) effective treatment can be directed either at the symptoms that trigger the behaviour or at the abnormal behaviour itself. We performed a retrospective review of consecutive cases with a diagnosis of rumination and repetitive belching made on HRM based on published diagnostic criteria[6-11] to assess whether these predictions were supported by clinical observation. Based on the results, a new classification system for these behavioural disorders is proposed, based not on the presence of repetitive regurgitation or belching or manometry (e.g. ‘R-wave’) alone, but rather on the underlying mechanism of disease.
- Top of page
- Supporting Information
This study provides evidence that rumination and many cases of repetitive belching are not distinct conditions, but represent common behavioural responses to a variety of digestive symptoms. Only rapid, supra-gastric belching independent of oral intake represents a distinct abnormal behaviour. The findings also demonstrate the clinical utility of advanced physiological measurement during a test meal to describe mechanism of disease and establish diagnosis in patients with chronic, unexplained symptoms. In addition, encouraging data are presented that even one, brief behavioural intervention can produce lasting clinical benefit in many cases.
The clinical presentation, investigation, treatment and outcome of 46 adult patients with rumination and other belching/regurgitation disorders identified by HRM were reviewed. This patient group was typical of those in previous studies.[5, 6, 10, 20] The majority were female with a long history of functional gastrointestinal symptoms. As reported in children, a proportion of patients reported onset of symptoms following acute infection (not always gastrointestinal), surgery or psychosocial stress; however, the presenting symptoms of these individuals was otherwise no different from the group as a whole. In almost all cases, extensive investigation, sometimes including conventional manometry, had failed to establish a diagnosis and a variety of empirical treatments had failed to improve symptoms. Referral letters noted a variety of symptoms (Table 2), but rarely included rumination syndrome in the differential diagnosis and never mentioned supra-gastric belching. Since the study investigators also provide therapy for these conditions in our region, this finding is almost certainly due to low awareness and not because other physicians are making the diagnosis based on clinical presentation alone.
Physiological measurement with concurrent documentation of symptoms and clinical events during a test meal provided objective evidence of behavioural disorders. HRM with impedance is considered an accurate test[6-11]; however, we suspect false-negative results in a small number of patients with typical symptoms that had a normal ‘stationary manometry’ but findings on ambulatory pH-impedance that may represent ‘rumination’ (e.g. repeated, symptomatic non-acid reflux after meals) or ‘supra-gastric belching’ (e.g. aerophagia followed by expulsion). These individuals were excluded as our analysis focused on the mechanism of disease and required a definitive description of the physiological events. False-positive results are also possible; however, further investigation did not change the diagnosis for any individual during median 5-month follow-up. During HRM studies, the majority of patients [33/46 (72%)] spontaneously reported dyspeptic symptoms before diagnostic pressure events and the return of gastric or oesophageal contents to the mouth. The handful that reported no postprandial symptoms prior to the onset of rumination tended to have a long history (often since childhood). In these cases, the abnormal behaviour may be so well established that even the normal sensation of fullness after a meal could trigger this response. About half the patients had diagnostic results after 10 water swallows; however, the yield was doubled by the inclusion of free drinking (200-mL water) and a test meal. The close temporal association of abdominal symptoms before the appearance of rumination or belching confirms the behavioural aetiology of these conditions. Moreover, it provides patients with a clear explanation of the cause of symptoms that many found extremely helpful in coming to terms with the diagnosis and engaging with behavioural treatment.
Consistent with the study hypothesis, the symptoms that preceded the onset of rumination and supra-gastric belching were varied; however, the range of behavioural responses was very limited. Almost all the rumination and belching events were preceded by voluntary, albeit unconscious, contraction of the abdominal wall. In the majority with dyspeptic symptoms after the meal, this behaviour resulted in typical rumination of gastric contents; however, the same response could force out oesophageal contents if it occurred during eating. In other patients, abdomino-gastric strain occurred exclusively in response to typical reflux events and, as lower oesophageal sphincter relaxation reduced the resistance to retrograde flow, such individuals tended to eject large volumes of gastric contents. These observations with combined HRM impedance technology build on those of Rommel et al. and show that the timing of abdominal strain in relation to drinking and eating can determine the clinical presentation. Specifically, what was present in the lumen at the time the abdominal muscles contracted determined whether air, liquid or food returns to the mouth. Thus, the same mechanism can result in ‘gastric rumination’, ‘reflux rumination’, ‘supra-gastric rumination’ (i.e. return of oesophageal contents) and, in cases of aerophagia, ‘supra-gastric belching’ (Table 3). In many cases, these ‘rumination variations’ could be observed at different times in the same patient. More unusual ‘variations’ were observed in specific individuals (Table 4), such as cough to increase abdominal pressure; however, the basic mechanism of disease was present in almost all cases. The exception was rapid, repetitive belching that occurs independent of meals as described by Bredenoord et al. This cannot be produced by aerophagia and/or straining. Rather, it is achieved by suction of air into the oesophagus through an open UOS during forced inspiration with immediate release on expiration. In our experience, this behaviour can produce such a rapid succession of belches that it caused breathlessness and distress reminiscent of panic attacks.
Once the diagnosis is established, behavioural treatment is the mainstay of treatment.[21, 22] It was shown that even a single, brief intervention can suppress rumination and belching. However, if the study hypothesis is correct and abnormal behaviour is a response to digestive symptoms, then it should be possible to direct therapy either at those symptoms or at the response itself. The majority of patients with rumination and supra-gastric belching had functional dyspepsia. In addition to behavioural management, these patients were prescribed low-dose antidepressants that reduce gastric hypersensitivity and visceral pain in this condition.[24, 25] Control of rumination and/or belching could be achieved quickly with behavioural management; however, initially, some individuals struggled to maintain control because of on-going dyspeptic symptoms. This became easier as the abdominal pain settled on the medication (typically over 4–8 weeks). These observations are consistent with the study hypothesis; however, in a small number of cases (Table 5), investigations revealed specific, treatable causes of symptoms. In one patient, abdominal pain was the result of a pancreatic pseudocyst and drainage resulted in immediate relief. In another, adhesiolysis around the proximal jejunum during insertion of a feeding tube released occult obstruction with immediate improvement in both dyspepsia and rumination such that the jejunostomy never had to be used. In another patient, rumination occurred exclusively in response to ‘typical’ reflux events and fundoplication provided excellent control of her symptoms. The immediate effect of specific treatment in these instructive cases supports the view that relief of unpleasant digestive symptoms can be sufficient to suppress also abnormal behaviours related to these symptoms.
Rumination is most often associated with dyspepsia; however, abnormal behavioural responses associated with reflux disease may also be quite common.[26, 27] Direct observations by HRM with impedance can document whether abdomino-gastric strain is forcing gastric acid into the oesophagus or whether spontaneous reflux events trigger the abnormal behavioural response. If rumination is the cause of, rather than a response to, acid reflux, then pH-studies will often produce false-positive results (see above). In these cases, fundoplication may physically prevent rumination, but dyspeptic symptoms are likely to increase and patient behaviour will adapt to the new circumstances. Conversely, if reflux is the trigger for rumination, then reflux suppression with the GABAB receptor agonist baclofen or anti-reflux surgery may be effective options. If surgery is considered, then patient selection is critical. In our case series, 5 patients had rumination syndrome diagnosed after anti-reflux surgery. In two cases, this behaviour commenced de novo years after surgery following an acute physical or psychological stress. However, in the three other cases, although fundoplication suppressed rumination, dyspeptic symptoms persisted and abnormal behaviour either overcame the wrap or resulted in supra-gastric rumination and belching.
On the basis of these observations, a classification of rumination and other regurgitation/belching disorders can be proposed (Table 6).
Table 6. Suggested classification of rumination and belching
|Classification of rumination and belching|
|Primary (‘classic’)||Rumination with or without belching during/after meal|
|Secondary or reflux associated rumination||Rumination secondary to trigger, such as reflux|
|Supra-gastric belching||Supra-gastric belching independent of meals|
Previous attempts at classification have been descriptive, based on the presence and timing of abdominal strain and association with retrograde flow of liquid and gas. In contrast, this system identifies three groups with distinct mechanisms of disease that may respond to specific management. Classic or “primary” rumination is most often triggered by dyspeptic symptoms. In this condition, abdomino-gastric strain results in the return of food or belching (depending on what is in the lumen), from the stomach or oesophagus (depending on the timing of contraction). Reflux or secondary rumination has a similar mechanism, but is triggered by reflux events. Other causes of "secondary" rumination also exist as detailed in Table 5. Both may respond to behavioural therapy directed at abdominal wall relaxation; however, the events that trigger this behaviour are different and may respond to specific therapy. In contrast, rapid, repetitive supra-gastric belching is produced by a distinct behavioural abnormality and may require specific therapy focused on the upper oesophageal sphincter.
This study has the limitations of most case reviews. Clinical data and follow-up were not always complete, especially in out-of-area referrals, and medical treatment was not provided in a systematic manner. In particular, although all patients received at least one session of behavioural instruction, only a minority received physiotherapy in the community. As a result, this study almost certainly underestimates the potential benefits of this approach. However, even in cases where behavioural therapy was not effective and individuals continued to have recurrent rumination or belching, a definitive diagnosis was helpful as it helped to avoid further investigation and inappropriate treatment.
In conclusion, these findings support the hypothesis that rumination and many of its variations represent common behavioural responses to digestive symptoms after meals and that a simple, generic behavioural intervention can provide effective management whether the presenting complaint is ‘vomiting/regurgitation’ or ‘belching’. Furthermore, this study demonstrates the clinical utility of HRM studies during a test meal in a group of patients with medically unexplained, treatment-resistant symptoms. Advanced physiological measurement identifies three groups with distinct mechanisms of disease that are likely to respond to specific management. Moreover, the vivid, visual demonstration of oesophageal function provided by HRM can help patients (and their doctors) understand the cause of their symptoms, and enhance patient acceptance of the diagnosis and the effectiveness of behavioural therapy.