Gastroesophageal reflux (GER) is defined as the retrograde passage of gastric contents into the esophagus. In term and preterm infants, GER is usually a benign physiologic process, but it meets the definition of gastroesophageal reflux disease (GERD) if it causes clinical symptoms or complications. A multitude of gastrointestinal (GI), respiratory, and other symptoms, including apnea, worsening of lung disease, irritability, feeding intolerance, failure to thrive, and stridor, have been attributed to GERD. However, determining whether reflux is the cause of symptoms in an individual patient can be challenging. The approach to an infant with suspected GERD is further complicated by the paucity of available medications demonstrated to be safe or effective in this population.
Upper Gastrointestinal Motility and Physiology
An understanding of GER in infants begins with the physiology of the upper GI tract. Esophageal motor function is well developed in infants as early as 26 weeks’ gestational age. Swallowing triggers coordinated esophageal peristalsis and lower esophageal sphincter (LES) relaxation, as it does in more mature patients. However, the velocity of propagation is significantly faster in term infants than in preterm infants. Manometry has also documented that spontaneous esophageal activity unrelated to swallowing tends to take the form of incomplete or asynchronous waves; this type of nonperistaltic motor activity occurs more frequently in preterm infants than in adults.
The LES, which blocks GER, is made up of intrinsic esophageal smooth muscle and diaphragmatic skeletal muscle. Although premature infants were once thought to have impaired LES tone, several manometry studies have documented good LES tone, even in extremely low–birth weight infants. In term and preterm infants, as in older patients, transient LES relaxations (TLESRs) unrelated to swallowing are the major mechanism that allows GER by abruptly dropping lower esophageal pressure below gastric pressure. These TLESRs may occur several times per hour in preterm infants, although the majority of TLESR events are not associated with GER. Preterm infants with GERD and those without GERD experience a similar frequency of TLESRs, but infants with GERD have a higher percentage of acid GER events during TLESRs. It has been hypothesized that straining or other reasons for increased intra-abdominal pressure may increase the likelihood of a GER event during a TLESR. Although LES relaxations also occur during normal swallowing, these are less often associated with GER events than with isolated TLESR events.
In addition to the anatomic and physiologic factors described that increase the likelihood of the retrograde passage of gastric contents into the esophagus, infants ingest a much higher volume per kilogram of body weight, approximately 180 mL/kg/day, compared with older children and adults. In the neonatal intensive care unit (NICU) population, preterm and term patients with nasogastric or orogastric feeding tubes may experience more reflux episodes as a result of mechanical impairment of the competence of the LES.
Gastric emptying is also an important factor in the passage of fluids through the upper GI tract. One small study showed that between 25 and 30 weeks, gastric emptying time seems to be inversely and linearly correlated with gestational age at birth. This study also found that simultaneously decreasing the osmolality and increasing the volume of feeds accelerated gastric emptying, although changes in osmolality or volume alone did not have a significant effect. Emptying also occurs faster with feeding of human milk than of formula. Several small studies suggest that prebiotics, probiotics, and hydrolyzed formulas may speed gastric emptying time in formula-fed infants. Fortification of human milk may slow gastric emptying time. The clinical significance of these findings with regard to GER remains uncertain, however. Although it seems logical that slower gastric emptying would be associated with increased GER, a study on the relationship between gastric emptying and GER in preterm infants found no such association.
Diagnosis of Gastroesophageal Reflux and Gastroesophageal Reflux Disease
Although infants have a propensity to experience frequent GER, the majority of GER is physiologic and nonpathologic. GERD is defined as GER that causes complications. Unfortunately, in infants, particularly preterm infants, complications of GER are difficult to characterize. Clinicians disagree about which symptoms are caused by GER or GERD. There is mixed evidence in the literature to support or refute most of the proposed complications of GER, including apnea, worsening of lung disease, and failure to thrive, in infants. An ongoing problem, particularly in the preterm population, is that many of the putative symptoms of GERD also frequently occur for other reasons. For instance, preterm infants without GERD also frequently experience apnea, lung disease, or feeding intolerance.
Physiologic Gastroesophageal Reflux
Nonpathologic GER occurs frequently in both preterm and term infants. Among 509 healthy asymptomatic infants aged 3 to 365 days monitored with an esophageal pH probe, the mean number of acid reflux episodes in 24 hours was 31.28, with a standard deviation of 20.68. The reflux index, the percent of time the esophageal pH was <4, ranged from <1 to 23, with the median and 95th percentile being 4 and 10, respectively. For this reason, a reflux index of 10 is often considered the threshold value for an abnormal study, but it must be remembered that none of the infants in this study were thought to suffer from symptomatic GERD, and clinical correlation with symptoms is required to make the diagnosis of GERD. Among the neonates in this study, the 95th percentile for the reflux index was as high as 13.
In a smaller study of 21 asymptomatic preterm neonates with a median postmenstrual age of 32 weeks, continuous combined esophageal pH and impedance monitoring detected refluxed fluid in the esophagus by impedance for a median of 0.73% (range 0.3%-1.22%) of the recording time, and acid exposure detected by pH monitoring for a median of 5.59% (range 0.04%-20.69%) of the recording time. When using combined pH and multichannel intraluminal impedance (MII) monitoring, detection of acid exposure may exceed volume exposure because the esophageal pH may remain depressed for a time after the majority of the bolus has been cleared, as well as for a variety of other technical reasons. Norms for acid and non–acid reflux are less well defined in preterm than in term infants because of the practical and ethical barriers involved in placing esophageal pH probes in a large number of asymptomatic preterm infants. However, the data from this small study make it clear that GER events occur frequently in asymptomatic infants and that a wide range of reflux measurements may be seen in healthy preterm infants without GERD.
In a study of otherwise healthy infants seen in general pediatric practice, half of all parents reported at least daily regurgitation at 0 to 3 months of age. The peak prevalence occurred at 4 months, with 67% reporting regurgitation, but thereafter it declined rapidly. Thus benign regurgitation was the norm in the first few months of life. Parents reported regurgitation to be a problem when it was associated with increased crying or fussiness, perceived pain, or back arching. The prevalence of regurgitation perceived as a problem peaked at 23% at 6 months but was down to 14% by 7 months. The majority of these children did not receive treatment for GERD from their pediatrician, suggesting that a diagnosis of GERD was only made in a minority of these patients. Infants who did as well as those who did not experience frequent regurgitation between 6 and 12 months of age were subsequently followed-up a year later. At this time, none of the parents described regurgitation as a current problem, and only one child experienced spitting at least daily. That child had not experienced frequent regurgitation at 6 to 12 months of age. Infants who had frequent spitting at 6 to 12 months of age did not experience more ear, sinus, or upper respiratory tract infections or more wheezing. In general, this cohort demonstrated that in the vast majority of infants, regurgitation is a benign process that is outgrown. However, it was noted that in the 1-year follow-up assessment, parents of infants experiencing frequent regurgitation at 6 to 12 months were more likely to report prolonged meal times (8% versus 0%) and frustration about feeding their child (14% versus 4%), even though regurgitation symptoms were no longer present. It is not clear whether this represents a true difference in feeding behavior or parental perception in a group likely to be sensitized to feeding issues.
Gastroesophageal Reflux Disease—Symptoms
Although the definition of GERD hinges on the presence of troublesome symptoms or complications, identifying whether the symptoms in infants are, in fact, caused by reflux can be challenging. Symptoms frequently attributed to GERD in infants include regurgitation, Sandifer posturing, worsening of lung disease, food refusal or intolerance, apnea, bradycardia, crying or fussiness, and stridor. Regurgitation may be a symptom of GERD in infants but, in itself, is not a sufficiently sensitive and specific finding to make a diagnosis. In addition, otherwise healthy infants without sequelae from their regurgitation, so-called happy spitters, do not require treatment. Clustering regurgitation with other symptoms may increase the accuracy of diagnosis, as demonstrated by the Infant Gastroesophageal Reflux Questionnaire–Revised (I-GERQ-R). However, the validity of such questionnaires has not been established in the NICU population, which includes preterm infants and sick term neonates who have multiple competing causes for the symptoms frequently attributed to GERD.
Although GERD and bronchopulmonary dysplasia (BPD) seem to be associated, the presence or direction of causality have not been determined. Patients with increased work of breathing may generate more negative intrathoracic pressures, thereby promoting the passage of gastric contents into the esophagus. Conversely, aspirated refluxate could injure the lungs. Finally, there may be no causal link in the majority of patients, with immaturity and severity of illness predisposing them to both conditions. In addition, part of the apparent association between BPD and GERD may be caused by an increased index of suspicion for GERD in patients with BPD, leading to increased rates of diagnosis.
A similar issue exists with regard to apnea in premature infants. Although esophageal stimulation may trigger airway protective reflexes in animal models, there is insufficient evidence in human infants to confirm that reflux causes apnea. In addition, apnea may itself trigger reflux. Finally, it may be that immature infants are simply prone to both apnea and reflux, and there is no causal association. In a cohort of infants referred for overnight esophageal and respiratory monitoring for suspicion of GERD as a cause of apnea, desaturation, or bradycardia, fewer than 3% of all cardiorespiratory events were preceded by a reflux event. The infant with the highest percentage had 4 of 21 cardiorespiratory events preceded by GER. Conversely, 9.1% of reflux events were preceded by a cardiorespiratory event. This study shows that it is more common for a cardiorespiratory event to precede reflux than for reflux to precede a cardiorespiratory event. Cardiorespiratory events preceded by reflux were not more severe than those not preceded by reflux. Furthermore, even in this population referred for suspicion of GER-triggering cardiorespiratory events, only a small minority of cardiorespiratory events were, in fact, preceded by reflux. This suggests that even if all of these temporally related events were also causally related, and even if a treatment were completely efficacious at eliminating GERD, the majority of cardiorespiratory events would not be eliminated by GERD treatment. However, data from small or moderately sized research cohorts cannot rule out the possibility that reflux can trigger the majority of cardiorespiratory events in a small subset of patients. Because bedside recording of apnea events is known to be inaccurate, correlation of apnea with feeding or reflux events in a specific patient requires formal simultaneous respiratory and esophageal monitoring studies.
It is unclear what component of the refluxate triggers complications. Infants experience less acid GER or GERD compared with older children or adults, in large part because of frequent buffering of gastric contents by milk. Although the majority of GER events in infants are nonacid events, at least some preterm infants are able to experience significant acid GER, often defined as an esophageal pH <4 for more than 10% of the recording. Acid GER predominates preprandially, and nonacid GER postprandially in infants. However, it is not clear whether acidity is the mechanism by which reflux causes complications in infants. The other characteristics of the refluxate that have been postulated to be associated with symptoms include the height of the bolus in the esophagus, the volume of the bolus, or the pressure exerted on the esophagus.
Gastroesophageal Reflux Disease—Diagnostic Tests
Numerous tests exist to measure acid and nonacid GER in infants. Esophageal pH probes measure acid reflux, and esophageal MII measures the presence of fluid in the esophagus regardless of pH. Impedance and pH sensors can be combined in one esophageal probe to give the most information about the frequency and timing about both acid and nonacid GER. Many systems have the capacity to be run in conjunction with respiratory monitoring or, for a family member or health care provider to mark the timing of a clinical symptom, to attempt to temporally correlate symptoms and GER events. An upper GI radiographic series is useful for assessing anatomic abnormalities that may contributing to or mimic GER but is a poor measure of the frequency or severity of GER because it only captures a brief window in time. A nuclear medicine scintigraphy study can identify postprandial reflux and aspiration and quantify gastric emptying time. There is no current gold standard diagnostic modality for GERD in infants. Part of the reason is that it is still not clear what component of reflux, such as its frequency, volume, acidity, or height, is most likely to cause complications in infants, and each test measures different parameters. An international consensus statement on GERD described that no single diagnostic test can prove or exclude extraesophageal presentations of GERD in pediatric patients. Furthermore, many NICU patients are too small for endoscopy for direct assessment of esophagitis, so esophageal symptoms can only be inferred from vague symptoms, such as food refusal or fussiness. Finally, because the diagnosis of GERD relies on the presence of clinical complications, no physiologic test that only characterizes the frequency or characteristics of GER events in a patient can, by itself, confirm the diagnosis of GERD ( Table 1.1 ).
|Able to Quantify GER Frequency
|Able to Diagnose GERD
|Upper GI series
|Can identify anatomic causes or mimics of GERD, such as a GI obstruction
|GER is measured over a brief period, so an assessment of the overall frequency of GER cannot be established
Correlation with symptoms is necessary
|Esophageal pH monitoring
|A 12- to 24-hour study allows for a better quantification of the amount of GER experienced by the infant
No radiation exposure
|It is not clear whether acid or nonacid refluxate triggers certain symptoms
Co-monitoring with a probe with both a pH and impedance channel may be a useful option
Because of technical limitations, some acidic events may not be detected by impedance alone
Correlation with symptoms is necessary
Many systems may be used in conjunction with cardiorespiratory monitoring or allow for marking of clinical symptoms to try to define the temporal association between symptoms and GER events
|Esophageal MII monitoring
|Allows for the detection of fluid boluses in the esophagus, regardless of the fluid pH
A 12- to 24-hour study allows for a better quantification of the amount of GER experienced by the infant
No radiation exposure
|Nuclear medicine scintigraphy study (also known as a “milk scan”)
|Can quantify gastric emptying time
Delayed images (usually 12-24 hours) allow for assessment aspirate in the lungs
Less radiation than a fluoroscopy-based study
|GER is measured over a short total period, even if repeat images are taken, so an assessment of the overall frequency of GER cannot be established
Does not distinguish between acid and nonacid GER
|Yes and no Correlation with symptoms is generally necessary, but a finding of aspiration could result in definitive proof of adverse sequelae
Gastroesophageal Reflux Disease—Treatment
Nonpharmacologic therapies for GERD include positioning, thickening of feeds, eliminating exposure to cow’s milk protein through maternal elimination diets or use of elemental formulas, reducing exposure to environmental tobacco smoke, and decreasing the volume of feeds while increasing the frequency. In a 2013 clinical report, the American Academy of Pediatrics (AAP) Section on Gastroenterology, Hepatology, and Nutrition endorsed these potential lifestyle modifications as potentially beneficial.
The type of milk an infant is consuming may affect GERD symptoms. In the Infant Feeding Practices Study II, direct breastfeeding was protective against reflux compared with formula feeding, but mothers were more likely to wean their babies with reflux. In preterm infants with symptoms of GERD and feeding intolerance, hydrolyzed formula has also been shown to decrease acid reflux measured with an esophageal pH probe, but not regurgitation events measured by impedance. When milk protein allergy is thought to be mimicking or triggering GERD, changing to a more elemental formula may also be appropriate. In the run-in period for a randomized control trial of a pharmacotherapeutic intervention for GERD, the majority of infants seemed to improve over a 2-week period with such a multipronged conservative management strategy, although this effect simply could also be attributed to time and maturation. Thickening of feeds has been shown to decrease episodes of clinical vomiting, although it does not seem to decrease physiologic measures of GER. Thickening can be achieved by adding thickeners or by using commercial GERD formulas that thicken in response to gastric acid. However, thickening of feeds can be challenging for preterm infants as, to date, commercial thickened feeds are nutritionally targeted for term infants and not preterm infants, commercial thickeners may be associated with a risk for necrotizing enterocolitis (NEC), and concerns have been raised about arsenic levels in the infant rice cereals often used for thickening.
Positioning has also historically been a mainstay of GERD management. Although typical positioning precautions for an infant with a diagnosis of GERD include elevating the head of the bed, there is no advantage to supine-upright positioning versus supine-flat positioning. Prone positioning seems to be associated with fewer GER events compared with supine positioning but is generally contraindicated because of the increased risk of sudden infant death. Lateral positioning with the right side down results in more frequent reflux events compared with left lateral positioning, but it is not clear whether this results in more symptoms. In a small study of intubated infants, elevating the head of the bed decreased the detection of gastric pepsin in the airway, presumably indicating a reduction of GER-related aspiration events with positioning.
Medications for the treatment of GERD are among the most common drugs prescribed in the NICU. In the United States, pharmacotherapy primarily consists of drugs to decrease gastric acidity, such as the histamine-2 (H 2 ) receptor antagonists and proton pump inhibitors (PPIs), and prokinetics, such as metoclopramide and erythromycin ( Table 1.2 ). However, an increasing number of studies have questioned the efficacy and safety of these medications in infants. In 2015 the AAP Section on Perinatal Pediatrics participated in the Choosing Wisely campaign, which charged medical societies with identifying unnecessary tests and treatments that contributed to health care waste. The Section included the following on its list: “Avoid routine use of anti-reflux medications for treatment of symptomatic GERD or for treatment of apnea and desaturation in preterm infants.” Nevertheless, because of the wide use of pharmacologic management and its benefit for the exceptional patient, this chapter will review how to gauge response to therapy and the pros and cons of each class of medication. Because both GER and the symptoms commonly linked to GERD, such as feeding intolerance and apnea, change rapidly as functions of time and maturation, valid studies of GERD in infants must account for this effect in their study designs. A study that simply measures symptoms before and after a therapy is likely to find improvement as a result of maturational effects, whether or not the therapy was truly efficacious. In addition, although many studies have demonstrated physiologic changes in response to pharmacotherapy, the gold standard for the treatment of GERD must be improvement in the symptoms that define the disease. Several recent well-conducted studied accounting for maturational changes have raised further questions about the efficacy and safety of common GERD drugs.
|Mechanisms of Action
|Examples of Putative Adverse Effects
|H 2 receptor antagonists
|Antagonists of the H 2 receptor in acid-producing gastric parietal cells
Suppresses basal and meal-induced acid production
|Suppression of gastric acidity is thought to mitigate damage to the esophageal mucosa or upper airway caused by the acidity of the refluxate
Use of acid suppression is based on the theory that the acidity of the refluxate is the trigger of certain complications, such as fussiness, food refusal, or stridor
Head rubbing or headache
Decreased calcium absorption
|Proton pump inhibitors
|Irreversibly block the gastric hydrogen/potassium adenosine triphosphatase in parietal cells
PPIs are thought to more strongly suppress gastric acidity than H 2 antagonists, leading to the potential for either greater efficacy or more adverse effects
|Antagonist of the dopamine-2 receptor subtype
|Motility agents are used to promote esophageal clearance of fluids and enhance lower esophageal sphincter tone
May also decrease the gastric emptying time
Gynecomastia and lactation
|Motilin receptor agonist
Promotes motility throughout the GI tract by promoting migrating motor complexes
Because of the difficulties in proving that a possible complication of GER is, indeed, caused by reflux and because of the questionable efficacy of available GERD medications, when treating an individual patient, it must be remembered that treatment failure may stem from either inappropriate application of drugs to treat symptoms not caused by GERD or failure of pharmacotherapy to improve true GERD. Apparent treatment successes may result either from a true treatment effect or from natural maturational changes in GERD or symptoms misclassified as resulting from GERD ( Table 1.3 ). Pharmacotherapy should be stopped if there is no improvement with therapy. If an improvement is seen, trial of cessation of therapy in several weeks should be considered because maturational changes may have been the cause of the initial apparent response or may obviate the need for therapy in the near future.