mise à jour du
5 février 2006
Clin Pediatr (Phila)
1994; 33; 11; 654-662
Behaviors associated with onset of gastroesophageal reflux episodes in infants
Prospective study using split-screen video and pH probe
Feranchak AP, Orenstein SR, Cohn JF.
Department of Pediatrics, Children's Hospital of Pittsburgh


To identify behaviors associated with the onset of gastroesophageal reflux episodes in infants both systematically and prospectively, each of 10 patients (aged 2 to 32 weeks) was studied during 2 hours of intraluminal esophageal pH probe monitoring, using a split-screen audiovisual recording technique. Videotape analysis of eight infants who had scoreable reflux events revealed six discrete behaviors closely associated temporally (P < .001 to < .05) with the onset of reflux events: "discomfort" (crying or frowning), "emission" (of liquid or gas, i.e., regurgitation, drooling, or burping), yawning, stridor, stretching, and mouthing. Three behaviors (hiccuping, sneezing, and thumb-sucking) were infrequent but were significantly associated with onset of reflux events in one or two patients each. A tenth behavior, coughing or gagging, was significantly associated with onset of reflux events in two patients, but not in the rest, despite relatively frequent occurrence. Exploration of temporal relations between reflux and each behavior suggested that discomfort, emission, mouthing, and cough-gag may have caused reflux episodes, and that all 10 of the behaviors may have been caused by reflux episodes. These findings and a "quiet period" immediately preceding episodes in six of the infants suggest interesting pathophysiologic mechanisms in infants which require further evaluation.
During the past several decades, while methods to identify nonregurgitant gastroesophageal reflux, such as pH probe, have come into wide use, investigators have noted some behaviors to be associated with reflux events in infants and older individuals. In many cases, these associations were general rather than temporal: patients with chronic cough, intractable hiccups, and severe stridor had abnormal quantities of reflux, and the behavior in question improved concurrently with treatment of their reflux. Additionally, there have been reports linking individual instances of esophageal acidification - either spontaneous or induced - to behaviors including cough, hiccups and stridor. The present study differs from both of these types of previous reports by being a systematic prospective approach to the question of behaviors associated with reflux episodes in unselected infants with reflux disease.
To record the behaviors and esophageal pH in a reliable manner susceptible to review and analysis, we utilized audiovideo recording. To assure the synchronization of the behavioral data and the reflux data, we used a split-screen technique. In a small sample of infants, we identified 10 discrete behaviors statistically associated with onset of reflux events.
The significant association of discrete behaviors with onset of reflux episodes in unselected subjects studied prospectively supports the more anecdotal information previously available for many of those behaviors. The association of these behaviors with onset of reflux is particularly striking when one considers the low frequency of many of the behaviors and the resultant low power to demonstrate the association. Only three of the behaviors (discomfort, cough-gag, and stretch) occurred more than 20 times during the entire study.
Three possible causal relationships may be surmised between reflux episodes and simultaneous discrete behaviors. The behavior may cause the reflux, the behavior may result from the reflux, or both may be caused by a third event.
Behaviors may cause reflux by increasing the abdominothoracic pressure gradient - i.e., by increasing abdominal positive pressure (cough, sneeze, crying) or by increasing thoracic negative pressure (stridor, hiccups). They may also cause reflux by inducing transient lower esophageal sphincter relaxations, as occurs with belching other factors which promote such relaxations are as yet poorly defined, but might include mouthing, thumb-sucking, yawning, or stretching, as suggested by work demonstrating myIohyoid activity at onset of 42% of such sphincter relaxations.
Behaviors may result from reflux if they are manifestations of airway soiling (cough, sneeze), of esophagitis (discomfort), or of reflex bronchospasm, laryngospasm, or diaphragm activity (cough, stridor, hiccups).
Behaviors may also be caused by reflux if they are used to clear refluxate from the esophagus (mouthing, salivation/ drooling, and possibly thumb-sucking, stretching, or yawning). These possible relationships are discussed in detail below.
Discomfort was associated with reflux episodes in all six subjects who manifested any associations between behaviors and reflux. Extrapolating from the experience described by adults, discomfort from heartburn is a likely result of acid contacting an inflamed esophagus in infants. Indeed, excessive crying, irritability and/or sleep disturbance often presenting as "colic" has been noted in 85% of a series of infants with esophagitis. Heartburn is the most likely explanation for the discomfort manifested by our subjects in the 15-second temporal segment immediately following the pH drop. For occasions when the discomfort occurred in the preceding temporal segment, the behavioral manifestations of discomfort may have produced the reflux event by increasing abdominal pressure, although in an earlier study, we found that crying per se did not seem to increase reflux frequency.
Emission of liquid (regurgitation, drooling) or gas (belching) was related to reflux episodes in three subjects. Because the recording technique did not definitely discriminate between regurgitation of apple juice and drooling of saliva, drooling and regurgitation were not distinguished. Additionally, emission of gas and liquid were scored together. Although liquid regurgitation occurred simultaneously with reflux, reflux was sometimes preceded by belching or followed by drooling. The association of belching and reflux has previously been explored; transient relaxation of the lower esophageal sphincter occurs with both. Acid reflux might also induce drooling, as stimulation of salivation is one part of the normal mechanism for clearance of refluxed material from the esophagus. The fact that only a minority of the subjects regurgitated during acid reflux underscores the protective roles that peristalsis and upper esophageal sphincter tone normally play in returning refluxed material to the stomach.
Yawning has not been linked to reflux previously but was significantly associated with onset of reflux in three of our subjects. In each instance, yawning followed reflux by a few seconds and was accompanied by a rise in pH, suggesting that yawning may aid clearance of refluxate. The function of the yawn has received little attention. Previously suggested functions have included prevention of atelectasis, augmentation of venous return and reduction of peripheral circulatory resistance and amplification of the sense of smell. Further study is needed to determine whether the yawn modifies gastroesophageal anatomy or pressure relationships to assist clearance of esophageal acid.
Stretching followed reflux in the one baby in whom it was associated with reflux. Like yawning, which often accompanies it, stretching may aid refluxate clearance by modifying the gastroesophageal junction. It may also be a nonfunctional response to irritation due to reflux.
Stridor was associated with reflux in two subjects and followed the reflux in both cases. Neither of these infants presented for reflux evaluation because of stridor. In each case, the episodes of stridor were rather subtle, consisting of audible inspiration occurring after reflux for only a few breaths. Stridor has been associated with gastroesophageal reflux in a few published case reports but the underlying mechanism remains unclear. Our data conform to the concept that reflux may cause stridor by stimulating an inflamed mucosa to produce reflex laryngospasm. The temporal sequence in our two infants did not suggest the converse, that stridor caused reflux (by augmenting the gastroesophageal pressure gradient), although this has also been proposed to occur.
Hiccups previously have been linked to esophageal irritation, and possible mechanisms have been proposed. Hiccups might induce reflux through an effect on esophageal motility or by augmenting the gastroesophageal pressure gradient (inspiration against a closed glottis increasing both the negative intrathoracic pressure and the positive intragastric pressure). Conversely, reflux might induce hiccups; a single report documents the production of hiccups by esophageal acidification in a patient with peptic esophagitis.2 Hiccups are particularly prominent in infants; a group of premature infants hiccuped during 2.5% of monitored time. Many parents of infants we have evaluated for reflux have described frequent postprandial hiccups. The two infants in the present study who displayed a significant association between hiccups and spontaneous esophageal acidification further support this association. Since reflux preceded their hiccups, reflux most likely caused their hiccups.
Coughing or gagging occurred after reflux in both infants in whom it occurred with reflux, suggesting the protective function of clearing the larynx or more distal airway. This protective function is also implicated in reports of chronic cough ameliorated by therapy of reflux and by other reports of cough following reflux. One of our subjects also refluxed after coughing, suggesting that the increased abdominal pressure may have provoked reflux.
Sneezing followed reflux in one baby. Like cough, sneezing is protective, clearing the nasopharynx of refluxed material. Conversely, although not suggested by the temporal relationship in our patient, sneezing might cause reflux by increasing abdominal pressure.
Mouthing, associated with reflux in one infant, occurred both before and after reflux episodes, suggesting bidirectional causality. Thus, after reflux, mouthing may represent a clearance attempt; preceding reflux, it may induce lower esophageal sphincter relaxation.
Thumb-sucking occurred after reflux in the one infant manifesting this association. Such sucking might increase refluxate clearance by stimulation of salivation and swallowing, or it might be a nonspecific self-soothing behavior due to the discomfort of reflux. It is unlikely that such sucking causes reflux, in spite of the lower esophageal sphincter relaxation which occurs with swallowing, since an earlier study did not demonstrate an increase in reflux frequency to be caused by pacifier-sucking in seated infants.
The quiet period we observed preceding some episodes of reflux may correspond to a similar phenomenon previously noted by others. However, other studies have linked movement with reflux episodes. These differences may be due to the underlying behavior state (e.g., if quiet precedes reflux awake, but movement precedes reflux during sleep), or to different types of reflux produced (e.g., if quiet precedes nonregurgitant reflux, but movement precedes regurgitant reflux) and will be interesting to study in the future.
The diversity of behaviors in response to reflux shown by our subjects is consistent with our current understanding of the diverse responses to reflux disease: some individuals with mild esophagitis have severe chest pain and others have no pain - despite severe esophagitis; some have bronchospasm and others do not; some have marked regurgitation and others do not.
This study does not address the issue of whether these behaviors are more common in infants with reflux disease; it simply shows that the behaviors are more likely to occur during the onset of reflux episodes than during periods without esophageal acidification.
The apple juice feedings we used differ from standard milk feedings in their pH; thus, our observations may be most readily generalized to acid reflux occurring late postprandially.
Although the study's conclusions might be questioned because the person coding the behaviors was not formally prevented from observing pH values on the other side of the screen, the practical demands of the observational task of coding in real time preclude this from happening.
The repeated statistical testing introduces the risk of a type I error. When the subjects are aggregated, the symptom "discomfort" is the only one which clearly reaches significance when the Bonferroni correction is used to protect against such an error, although the conservative nature of the Bonferroni correction suggests that perhaps mouthing, emission, and yawning might also reasonably be considered significant. Considering the study's design as a single subject study with eight replications, and again using the Bonferroni correction within subjects, the data in Table 3 can be conservatively considered significant for discomfort in two patients, yawn in one patient, stridor in one patient, thumb-sucking in one patient, and cough/gag in one patient. Thus, the seven behaviors which this study most clearly associates with onset of reflux episodes are discomfort, emission, yawning, stridor, mouthing, thumb-sucking, and cough/gag. Sneezing and hiccuping occurred a total of only one and five times, respectively, during the entire study, so the power to detect an association with reflux onsets was minimal.
The implications of this study are twofold. First, recognition of these associations may allow clinicians to detect subtle signs of reflux episodes in infants who do not spit up. Second, the associations may illuminate the pathophysiology of reflux episodes or the causes of some of the behaviors, such as yawns or hiccups. Our findings from this preliminary study should be used to generate more specific hypotheses for further, more focused controlled study.
In summary, we have found a temporal association of 10 behaviors with the onset of reflux episodes in infants. Further study is needed to confirm these findings, to elaborate the direction of causality, and to clarify the underlying mechanisms.