Gastric Emptying in Formula-Fed and Breast-Fed Infants Measured with the ¹³C-Octanoic Acid Breath Test

Abstract

Background: 

The ¹³C-octanoic acid breath test, a noninvasive method for measuring gastric emptying, was used to compare the gastric-emptying rate of formula-fed and breast-fed infants. Octanoic acid, a medium-chain fatty acid marked with the stable isotope ¹³C is immediately absorbed in the duodenum. Because gastric emptying is the rate-limiting step for the absorption of medium-chain fatty acids, the fraction of ¹³C expired in the breath indicates the rate of gastric emptying.

Methods: 

Twenty-nine newborn infants (16 boys, 13 girls) were investigated, with parental consent. The infants had a mean gestational age at birth of 34.5 weeks (range, 27-41 weeks) and a birth weight of 2148 g (range, 960-4100 g). Their mean weight on the day of the test was 2496 g (range, 1998-4140 g), and their mean age was 23 days (range, 7-74 days). Each infant received a test meal after a maximum fasting period of 3 hours. Fourteen infants were fed formula milk (Nutrilon Premium, NV Nutricia, Zoetermeer, The Netherlands) with ¹³C-octanoic acid and 15 infants received expressed mother's milk mixed with ¹³C-octanoic acid. After obtaining two basal breath samples and the feeding, breath samples were collected using a nasal prong, every 5 minutes during the first half hour and every 15 minutes during the next 3.5 hours. Analysis of the expired ¹³C fraction in the breath samples was performed using isotope-ratio mass spectrometry, and the gastric emptying curve and gastric emptying parameters were determined.

Results: 

The mean half-emptying time determined by the ¹³C-octanoic acid breath test was 65 minutes (range, 27-98 minutes) for the formula fed infants and 47 minutes (range, 16-86 minutes) for the breast-fed infants. The difference between the half-emptying times is significant (t-test, p < 0.05).

Conclusions: 

The results of the ¹³C-octanoic acid breath test indicated faster gastric emptying of human milk than formula. Our findings are in accordance with those in earlier studies, using the invasive-dilution technique; noninvasive and detailed ultrasonography, which is not easily used because it is operator dependent and the observation time is short; or cineesophagogastroscintigraphy, which is less suitable for infants (because of the radiation involved). The ¹³C-octanoic acid breath test is a safe and noninvasive method for measuring gastric emptying in small infants and allows comparison of various feeding methods.

Our current knowledge of development of gastrointestinal motility and normal motility patterns in children of different age groups, particularly in preterm infants, is still limited ⁽¹⁾. To optimize the initiation of enteral feeding and to decrease the risk of necrotizing enterocolitis, it is important to investigate gastric emptying ⁽²⁾.

Various techniques exist for assessing the gastric-emptying rate. Radioscintigraphy is considered the gold standard in adults ^(3,4), but radiation and transportation of the patient limits its use in small children. Most of the available studies of gastric emptying in infants involve marker dilution techniques and sampling techniques ^(5-9). These methods are invasive, impractical, and difficult to perform. Ultrasonography is noninvasive and allows a detailed evaluation of antropyloric function, but it is not easily used because it is operator dependent and the observation time is short ^(10-12). Other techniques, such as impedance tomography, antroduodenal manometry, and magnetic resonance imaging, are, respectively, sensitive to movements of the patients, invasive, and expensive. Magnetic resonance imaging requires sedation in the neonate. Breath tests combine reliability, noninvasiveness, and feasibility. Gastric emptying can be measured using the ¹³C-octanoic acid breath test. Octanoic acid, a medium-chain fatty acid is labeled with the stable isotope ¹³C, which is harmless to infants. Octanoic acid is immediately absorbed when it is evacuated into the duodenum ⁽¹³⁾. Thus, gastric emptying is the rate-limiting step for the absorption of medium-chain fatty acids, and the fraction of ¹³C expired in the breath indicates the rate of gastric emptying.

The ¹³C-octanoic acid breath test has been validated in adults in comparison with radioscintigraphy ⁽¹⁴⁾. The technique has been successfully applied in preterm infants ⁽¹⁵⁾. This noninvasive breath test has also been used simultaneously with and has been validated in comparison with the invasive-dilution technique in young infants ⁽¹⁶⁾.

Breast-feeding is the optimal form of nutrition for newborns. It contains the essential nutrients and, compared with formula feedings, it contains extra growth factors, hormones, and other nutrients. Moreover, human milk may have the advantage that the stomach-emptying process occurs at a faster rate. This may be important because gastroesophageal reflux (GER), which often affects infants, may be caused by delayed gastric emptying ^(17,18).

Our purpose was to compare the rate of gastric emptying in infants fed formula milk and in those fed breast milk using this new, noninvasive technique.

PATIENTS AND METHODS

Patient Selection

Twenty-nine healthy newborns were selected from the nurseries of the University Hospitals Leuven and the St Augustinus Hospital Antwerp, Belgium. Entry criteria required that the newborns have no medical problems, weigh at least 2000 g, and receive formula or human milk. No therapeutic feedings were allowed. None of these infants had gastrointestinal, cardiovascular, or respiratory disorders. Also, none of the mothers had been treated with antepartum corticosteroids, because it has been demonstrated that antenatal betamethasone administration to mothers of infants born between 26 and 32 weeks' gestation is associated with increased duodenal contraction rate, number of contractions per burst, and luminal peak pressure ⁽¹⁹⁾. Consent was obtained from the parents, and results of the gastric emptying test were mailed to them.

Breath Test

The Test Meal

The natural content of ¹³C in the test meals was measured. Test meals containing a high ¹³C content were excluded from use in the breath test. The formula test meal (50 ml) was prepared as follows: 50 ml of water was brought to boiling. Subsequently, 50 µl ¹³C-octanoic acid (Isotec, Centerville, OH, U.S.A.) and 1 g polyethyleneglycol 3350 (Norgine NV/SA, Zaventem, Belgium) were solubilized and cooked. Then, an adequate amount of formula powder (Nutrilon Premium; NV Nutricia, Zoetermeer, The Netherlands) was added.

The test meal was prepared on the morning of the test or the day before the test and kept in the refrigerator overnight. Immediately before administration, the bottle was reheated and then cooled to adequate temperature (37°C). The expressed breast milk (50 ml) was heated to 50°C. Fifty microliters ¹³C-octanoic acid and 1 g polyethyleneglycol were added. The milk was reheated and cooled for consumption. This test meal was prepared on the day of the test.

The test meals were administered orally within a time limit of 15 minutes at the infant's usual feeding time, allowing for at least 3 hours of fasting.

Breath Sampling

Breath samples were collected using a 6-French gastric tube (Argyle, Sherwood Medical, Petit Rechain, Belgium) that was positioned in the nasopharynx. A syringe was attached to the tube, and 10-ml samples of expired air were aspirated in parts of 1 ml per expiration and were stored in vacuum exetainers.

Two breath samples were obtained in basal conditions. After the infant consumed the formula or human milk test meal, breath samples were collected every 5 minutes during the first 30 minutes and every 15 minutes during the next 3.5 hours. Three hours after the test was begun, the infant received a usual feeding.

Analysis

Analysis of the expired ¹³C fraction in breath samples was performed using isotope-ratio mass spectrometry (IRMS; ABCA, Europa Scientific, Crewe, UK). The results were expressed as the percentage of ¹³C recovery per hour and percentage of cumulative ¹³C dose after 4 hours.

The CO₂-production, necessary for the calculations of the percentage ¹³C dose/hour is almost constant in preterm infants, as reported earlier. We used 20 mmol/kg per hour in our calculations ⁽²⁰⁾. Mathematical analysis resulted in two models: the first model is expressed as D = at^be^-ct, where D is percentage of ¹³C dose per hour, t is time, and a, b, and c are constants. The parameter derived from this model is the (gastric-emptying coefficient, GEC): GEC = ln a. The second model, D = mkβ^{- kt}(1 - e^-kt)^β-1, where D is the percentage ¹³C dose per hour, t is time, and m, k, and β are constants, results in the t_½ (half-emptying time): t_½ = [(-1/k)·ln(1 - 2^-1/β)]. Both parameters are independent of CO₂ production ⁽¹⁴⁾.

RESULTS

Twenty-nine newborns (16 boys 13 girls) were studied, with parental consent. The infants had a mean gestational age at birth of 34.5 weeks (range, 27-41 weeks) and a birth weight of 2148 g (range, 960-4100 g). Their mean weight on the day of the test was 2496 g (range, 1998-4140 g) and their mean age was 23 days (range, 7-74 days). Fifteen infants received their mother's milk, and 14 infants received formula milk. There were no significant differences between the breast-fed and formula-fed neonates in gestational age or birth weight. None of the children vomited after the test meal. Individual patient data are shown in Tables 1 and 2.

TABLE 1: Individual data of infants fed formula milk

TABLE 2: Individual data of infants fed breast milk

Results of gastric emptying parameters are shown in Table 3, in comparison with results of other investigators. The mean t_½ was 47 minutes in infants who consumed their mothers' expressed breast milk and 65 minutes in those who received formula.

TABLE 3: Comparison of gastric emptying in breast-fed and formula-fed infants

The difference in t_½ was significant (p < 0.05; t-test). The range and mean of t_½ are shown in Figure 1. The mean gastric emptying curves with standard errors are illustrated for both groups in Figure 2. There clearly was a consistent pattern and a clear difference between the individual curves from both groups during the first 90 minutes.

FIG. 1: Gastric half-emptying times (t_½) for human milk (expressed breast milk) and formula milk.

FIG. 2: Mean (SEM) gastric-emptying curves of breast-fed (BF) infants and formula-fed (FF) infants.

The GEC was 3.21 in infants fed breast milk and 3.20 in infants fed formula milk. The GEC is a mathematical parameter that is influenced by the rates of appearance and disappearance of the label in breath, because the GEC is directly related to the ascending and descending slopes of the curve. The similar GECs with both milks are probably caused by a mean value that compensates for steep and flat slopes. The difference was not significant. No correlation was found between t_½ and gestational age (r = -0.2455), birth weight (r = -0.2484), weight at time of investigation (r = -0.1417) and age (r = 0.2951).

DISCUSSION

Our current understanding of gastric emptying is based on adult physiology, and little is known about developmental differences in infants that may predispose to the various gastrointestinal problems frequently occurring in this age group.

Gastric emptying of swallowed amniotic fluid has been noted in utero as early as 26 weeks' gestation using a linear-array ultrasound scanner ⁽²¹⁾ and as early as 30 weeks' gestation using amniography ⁽²²⁾. However, there appear to be some differences: Preterm infants seem to have a slightly slower emptying rate than do full-term infants. In our study group, there was a 3-week difference in gestational age. The mean gestational age in the breast-fed infants was 36 weeks and in the formula-fed group, 33 weeks. The postpartum age, however, was less in the breast-fed than in the formula-fed infants. Thus, the postconception age was identical in both groups. Mean birth weight was less in the formula-fed group than in the breast-fed group (1859 g and 2532 g, respectively). Our results showing slower gastric emptying in formula-fed infants could be interpreted partly as a consequence of their greater prematurity. This would be in accordance with the widely believed notion that gastric emptying may be delayed in preterm neonates ⁽²³⁾. However, results in a few studies that were performed in infants of at least 32 weeks' gestation demonstrate little difference in gastric-emptying rate in full-term infants ^(24-26).

Almost all full-term infants readily tolerate breast or bottle feedings. In contrast, preterm infants do not always have coordinated rhythmic sucking, swallowing, and breathing and are dependent on nasogastric tube feedings. Early tube feeding can be dangerous for the preterm infant, because it may be one of the precipitating factors for necrotizing enterocolitis. From 34 weeks on, preterm infants usually tolerate small breast or bottle feedings.

Breast-feeding is unequivocally the optimal form of nutrition for newborns all over the world. Breast-milk provides the essential nutrients such as protein, fat, carbohydrates, minerals, and vitamins. Human milk also contains a variety of biologically active substances which serve as mediators for growth and differentiation of tissues and organs and multiple other factors enhancing defense against foreign antigens and infectious agents. Human milk contains significantly higher levels of lactoferrin, nucleotides, polyamines, hormones, and growth factors (epidermal growth factor, insulin-like growth factor, growth hormone-releasing factor, luteinizing hormone-releasing factor) than do conventional infant formulas.

None of the current formulas provides the unique mixture of food ingredients found in human milk. During the past decades we have seen many changes in the preparation of infant formula milk, made with the intent to increase their resemblance to human milk. The composition of energy sources has been adapted, vitamins and mineral composition have been modified, and substances such as taurine and carnitine have been added. Infant formula milk is the best substitute for human milk, if breast-feeding is not possible.

Gastroesophageal reflux is common in infants and children. In studies examining the relation between delayed gastric emptying and GER in children and infants, investigators have arrived at opposite conclusions. Several groups have found an association between delayed gastric emptying and GER ^{(17,18,27,28)}, but others have failed to do so. Ewer et al. ⁽²⁹⁾ found no association between gastric emptying and GER in preterm infants, nor did Rosen and Treves ⁽³⁰⁾ in a group of pediatric patients aged 0 to 16 years. No difference in gastric emptying was observed in children aged less than 3 years, regardless of the presence or absence of GER (DiLorenzo et al. ^[27]), However, Hillemeier et al. ⁽¹⁷⁾ also studied infants less than 2 years of age. Their data indicate that infants with severe GER are likely to have significantly delayed gastric emptying.

Heacock et al. ⁽³¹⁾ showed significant differences in physiological GER in newborn neonates who were breast fed compared with GER in those who were formula fed. The duration of GER episodes was significantly less in the breast-fed neonates during active sleep. This may mean that breast milk results in faster gastric emptying and less GER, whereas formula milk induces a longer gastric-emptying process and more GER. However, GER was recorded by a pH microelectrode for 4 hours after the infants' morning milk feeding, which is too short a registration period. It should also be remembered that all GER studies refer exclusively to acid reflux. No other studies correlating gastric emptying and GER have been described in neonates. The ¹³C-octanoic acid breath test will allow gastric-emptying studies combined with pH monitoring in very young infants.

Gastric emptying is influenced by various factors, of which only a few have been classified: Increasing caloric density can slow gastric emptying ⁽⁷⁾, and changes in carbohydrate composition of feeds can delay emptying ⁽⁸⁾. Thus, gastric emptying is dependent on the type of feeding. Earlier studies using more invasive techniques showed that gastric emptying is faster in children fed human milk than in children receiving infant formula. Cavell ⁽³²⁾ found a mean t_½ of 48 minutes in infants fed human milk and of 78 minutes in those fed formula milk using the marker-dilution technique. This method is probably unreliable, because the insertion of a tube may affect gastric motility. Tomomasa et al. ⁽³³⁾ investigated the pattern-based formula. Contractions were recorded manometrically during 3 postprandial hours. The migrating myoelectric complex appeared in 75% of breast-fed infants but only in 17% of formula-fed infants within the observation period. The fasting state recurred more rapidly in breast-fed infants. These findings suggest that human milk contains a factor accelerating the transport or digestion of intraluminal contents or that infant formula contains a factor that delays transport or digestion. An ultrasonic technique was used by Ewer et al. ⁽³⁴⁾ to compare gastric emptying after a feeding of expressed breast milk and formula milk. The data show that breast milk has a major effect on gastric emptying, which may have important implications for preterm infants who have feeding intolerance due to delayed gastric emptying. The mean t_½s (50% antral cross-sectional area) were significantly different, the values for breast milk and formula milk were 36 minutes and 72 minutes, respectively.

A study was performed to determine whether the addition of a fortifier to expressed breast milk affects gastric emptying in low-birth-weight infants. McClure and Newell ⁽³⁵⁾ showed, using ultrasonography, that fortifying breast milk does not affect gastric emptying. Another study by Ewer and Yu ⁽³⁶⁾ contradicts the conclusions of McClure and Newell; they demonstrated ultrasonically that breast milk emptied faster than fortified breast milk in 11 preterm infants. Fifty percent antral cross-sectional area occurred at 21 minutes with unfortified breast milk and at 48 minutes with fortified breast milk. However, not all fortifiers have the same caloric content, and the effects may change if the increased caloric content is inhibiting gastric emptying. The breast-milk fortifier used by McClure and Newell (Cow and Gate; Trowbridge, UK) raises the nutritional content of every 100 ml by 10 kcal of energy, whereas the FM-85 fortifier (Nestlé, Vevey, Switzerland) of Ewer and Yu increases the nutritional content of 100 ml by 18 kcal, which is almost twice as much. The higher caloric content is probably the reason why the addition of fortifier slowed gastric emptying in the study of Ewer and Yu and not in that of McClure and Newell. Temperature differences (6°C, 24°C, 37°C) of the milk (formula milk) do not seem to have any effect on gastric emptying ⁽³⁷⁾.

Billeaud et al. ⁽³⁸⁾ state that gastric emptying does not vary with age or gender during the first year of life but is mainly influenced by the type of milk. Using cinoesophagogastroscintigraphy in a quantitative study of the transit of nutrients marked with a short-life radioisotope, they measured percentage retention after 30 and 120 minutes. Human milk, acidified cow's milk, and adapted formula containing whey-hydrolysed proteins were readily emptied from the stomach, and emptying of human milk was faster than emptying of formula. Caseinpredominant formulas resulted in a slower gastric-emptying rate. Follow-up formula and cow's milk were the slowest to evacuate from the stomach. Based on these findings, it was suggested that higher contents of proteins and caseins delay gastric emptying.

Another possible technique for measuring gastric emptying is epigastric impedance ^(26,39). A small current is passed through circular skin electrodes, and the impedance signal detected by two other electrodes is amplified, converted, and analyzed. An emptying signal and an impedance gastrogram are obtained with this technique. This is a simple investigation but very sensitive to spontaneous movements and displacement of the stomach in children, resulting in nonvalid measurements in approximately a quarter of the infants. Another limiting factor of this technique is the necessity of using simple, liquid test meals.

Applied potential tomography (APT) is also an electrical impedance measurement using tomography. Nour et al ⁽²⁵⁾ measured gastric emptying using APT in 29 term and 53 preterm infants less than 3 months of age. The first feeding was formula milk, human milk, or a rehydration solution. They found no difference in gastric emptying time and rate between formula and breast milk (however, only 6 infants were fed human milk) in these groups of infants. Applied potential tomography showed a slower emptying of milk feedings than of the aqueous solution. However, the impedance technique is sensitive to spontaneous movements.

In our study we report the comparison between the gastric-emptying rate of infants receiving breast milk or formula milk using the ¹³C-octanoic breath test. This novel test has been shown reliable and adequate ^(14,15). A group of preterm infants have already been studied, and gastric emptying resulted in a t_½ of 56 minutes ⁽¹⁵⁾. In this study population, none of the mothers had taken corticosteroids, which could influence gastric emptying.

We clearly demonstrated faster gastric emptying of breast milk (t_½ of 47 minutes) than that of formula milk (t_½ of 65 minutes). In contrast to those used in previous studies, the method that we used is acceptable and reliable. Our findings confirm that physiological gastric emptying of breast-fed neonates is faster than gastric evacuation of infants receiving adapted formula. However, it is not yet clear whether the slower rate of gastric emptying in infants receiving formula has a clinical significance. The subjects that we studied had no excessive regurgitations.

In conclusion, the ¹³C-octanoic acid breath test demonstrates significantly faster gastric emptying of human milk than of formula milk. Of all the methods available, the ¹³C-octanoic acid breath test is the most suitable: It is a safe, nonradiologic, and noninvasive technique. The test can be performed on an ambulatory basis or at the bedside (no transportation of the infant is necessary). Samples can be sent to a center that has a mass spectrometer, and the breath test is relatively inexpensive. We propose that this method is the most appropriate manner for measuring gastric emptying in small infants and preterm infants, allowing comparison between various feeding methods and conditions.

Acknowledgment: The authors thank the mothers, their infants, and the staff of the Pediatrics Department of the University Hospitals Leuven and the Hospital St. Augustinus Antwerp for their help and participation, and the Gastrointestinal Research Center for the helpful and kind collaboration of the staff. The activities of MVD were supported by Nutricia Leerstoel (NV Nutricia, Zoetermeer, The Netherlands).