What Is Known
- Optimal feeding for preterm infants is of major importance for their short- and long-term health.
- Achievement of adequate enteral nutrition soon after birth is often difficult in premature infants because of feeding intolerance.
- Premature infants can receive enteral feeding either by continuous or intermittent bolus tube feeding, but both strategies are associated with possible benefits and risks.
- It is unknown which strategy is preferable.
What Is New
- Bolus and continuous feeding are equally suitable feeding strategies for preterm neonates.
- Intermittent bolus feeding may be preferable.
Inadequate nutrient supply during the sensitive period of development endangers body and organ growth, particularly in neurodevelopment (1–4)(1–4)(1–4)(1–4). Therefore, achieving adequate nutrition as soon as possible after birth is a major goal for all neonates.
Early enteral nutrition (EN) to low-birth-weight (LBW) infants is often difficult to achieve because of gastrointestinal immaturity and complicating medical conditions. Full feeds are often delayed because of feeding intolerance and temporary cessation of feeding. This necessitates prolonged parenteral nutrition (PN) and increases the risks of related complications including infections and cholestasis (5–7)(5–7)(5–7).
Early achievement of full enteral feeding reduces these risks and promotes the functional development of the gut and the mucosal immune system (7). Therefore, many hospitals around the world begin minimal enteral feeding (MEF) as soon as possible after birth and increase the amount of enteral feeding daily via tube feeding. Tube feeding can be either semicontinuous (CON) or as a bolus (BOL) every 2 to 3 hours. Neonatologists, however, are in equipoise regarding the best method because of conflicting evidence regarding the benefits and safety of both methods (8).
Continuous feeding is thought to improve energy efficiency, duodenal motor function, nutrient absorption, and splanchnic oxygenation (9–13)(9–13)(9–13)(9–13)(9–13). A substantial portion of the nutrients provided, however, could be lost to the delivery system (14). In contrast, bolus feeding may result in a more physiological release pattern of gastrointestinal tract hormones and may stimulate gastrointestinal tract development and enhance protein accretion (15–17)(15–17)(15–17). It, however, may also adversely affect pulmonary function (18) and be more difficult for the immature gastrointestinal tract to handle ultimately resulting in increased feeding intolerance and feeding-related apnoeas.
A Cochrane review (8) concluded that the present evidence is inadequate for determining an optimal feeding strategy because of the small sample sizes and methodological limitations. Therefore, the goal of this study was to determine which feeding strategy—continuous or intermittent nasogastric tube feeding—is the optimal approach for shortening the time to full enteral feeding among LBW infants.
The participating infants were enrolled between February 2007 and February 2009. Infants were eligible if they met the following criteria: birth weight (BW) <1750 g, gestational age <32 weeks, and born in or admitted to the neonatal intensive care unit (NICU) of the Sophia Children's Hospital of the Erasmus Medical Centre in Rotterdam within 24 hours of birth. Exclusion criteria were congenital gastrointestinal obstructions (eg, duodenal atresia, anal atresia) and suspected metabolic, endocrine, or renal disorders.
Informed consent was obtained from the parents when the infant was considered eligible for the study. The study design was approved by the institutional review board of the Erasmus Medical Centre in Rotterdam.
The study infants were enrolled within 24 hours of birth, stratified according to BW, and randomly assigned within each weight group to either intermittent BOL or CON feeding. The randomisation (using random numbers) used sealed opaque envelopes that were grouped in an even blocked design. The envelopes were prepared for every BW category with 10 participants per category every time. There were no minimum or maximum number of participants per BW category.
The feeding intervention ended in the event of withdrawal of parental consent, necrotising enterocolitis (NEC) stage II and beyond, or death. Patients completed the study if they were given either BOL or CON nasogastric milk feeding until a minimal enteral intake of 120 mL · kg−1 · day−1 or reached the postnatal age of 28 days. The endpoint of 120 mL · kg−1 · day−1 was selected because neonates who reached this amount were discharged from the NICU to a regional level II hospital whenever permitted by their clinical condition as dictated by the standard of care.
All of the patients received nasogastric tubes. Enteral feeding started on the day of birth according to our MEF regime. Every 4 hours, patients received MEF as a function of their BW—0.5, 1, or 2 mL (for BW 500–749 g, 750–1249 g, and 1250–1749 g, respectively). They were given their own mother's milk or formula, if mother's milk was not available. In the absence of asphyxia or patent ductus arteriosus (PDA) on the day after birth, feeds were started at 24 mL · kg−1 · day−1. Equal daily increments were given such that under ideal circumstances full-volume feeds (120 mL · kg−1 · day−1 to 100 kcal · kg−1 · day−1) were reached in 6 days. The PN was reduced inversely proportional to the increasing amount of feed administered. When 120 mL · kg−1 · day−1 of EN was reached, PN was stopped. Slightly adapted standardized feeding regimes were followed in case of asphyxia or in children small-for-gestational age (SGA, defined as weight for age <−2 SD) (19). In these adapted standardized feeding regime, the SGA infants or infants with PDA required 2 days longer to reach full-volume feeds because of a less rapid increase in feed volume. Infants with asphyxia required 1 day longer to reach full-volume feeds because of the delayed introduction of enteral feeds. During that period, the infants did receive MEF.
Every 3 hours, the infants enrolled in the BOL group received bolus feedings by gravity after tube priming for 15 minutes. Infants enrolled in the CON group received semicontinuous feeds; that is, they received a quarter of their hourly feeding volume by gravity via an open syringe that was filled with milk every 15 minutes throughout the day. In both groups, the gastric residuals were measured every 8 hours before the feedings. Feedings were postponed for 3 hours if the gastric residuals exceeded 3-fold the hourly volume or the volume of the preceding bolus in the CON and BOL groups. When meconium passage did not occur within the first 24 hours after birth, enemas (saline 0.9%) were administered as standard care. Enemas were repeated every day until spontaneous defecation occurred. This protocol was already in place at the start of the study and was not altered by the study team. The attending neonatologist made such decisions.
When available, expressed mother's milk was the nutrition of choice and was used undiluted. When mother's milk was unavailable or when the supply was insufficient to meet complete dietary needs, it was replaced by standard infant formula (Nenatal; Nutricia, Zoetermeer, The Netherlands). When human milk was used, breast milk fortifier (Nutricia, Zoetermeer) was added at 2.1 g/50 mL when enteral feeds reached 100 mL · kg−1 · day−1 or when enteral feeds were given for ≥7 days. Patients were recorded as receiving expressed own mother's milk when the minimal intake of mother's milk exceeded 75% of the prescribed intake on the day of reaching full enteral feeding.
Supplemental fluids were given on day 1 as solutions of glucose and amino acids and subsequently as standardized PN. Our goal was to provide glucose between 4 and 12 mg · kg−1 · min−1, amino acids between 3 and 4 g · kg−1 · day−1, and fat between 2.5 and 3 g · kg−1 · day−1.
The primary outcome was defined as days to reach full enteral feeding (120 mL · kg−1 · day−1). It was calculated from initiation of feeding. The following tolerance parameters and weight gain outcomes were defined as secondary outcomes: gastric residual volumes per day, number of apnoea episodes per day, number of patients with feeding interruptions, days to regain BW, days on mechanical ventilation, and adverse events (nosocomial infections, NEC, or death).
Data related to feeding tolerance (gastric residuals, number of apnoea episodes, and number of feeding interruptions) were collected daily until the endpoint (120 mL · kg−1 · day−1) was reached or for 28 days, whichever came first. Weight was measured every other day. Mechanical ventilation was divided into invasive and noninvasive ventilation (continuous positive airway pressure/nasal intermittent positive pressure ventilation). Oxygen supply by nasal prongs only was not considered ventilation. Adverse events were assessed daily.
Statistical analysis was performed according to the intention-to-treat approach for all of the patients who entered the study. For continuous measures, the mean or median differences were reported with 95% confidence intervals (CIs) as appropriate. Incidences were analysed using the risk ratio with 95% CIs. The Cox proportional hazards regression analysis was conducted to assess the effect of bolus feeding on the primary endpoint “days to reach full enteral feeding.” Hazard ratios were adjusted for the BW strata. Sex and SGA/appropriate-for-gestational age (AGA) were entered to test for effect modification. A P value <0.05 was considered statistically significant.
The sample size of 250 infants was based on a power analysis performed using historical data. The normal time to achieve full enteral nutrition for preterm infants in the NICU of the Sophia Children's Hospital was 13 days (SD 5 days) at the time of the study. Other studies suggested a range of 15 to 35 days to reach full enteral feeding. CON feeding was the standard feeding regime at our NICU before the start of the study. We hypothesised that BOL feedings would result in a 2-day decrease (effect size 0.4 SD) in the number of days required to reach full enteral feedings (primary outcome) versus CON feedings. The sample size needed to verify this hypothesis was 220 patients based on an α error of 0.05 and a β of 0.10. We included an additional 30 infants to allow for dropouts.
A total of 250 children underwent randomisation, and 246 children were included in the analysis (Fig. 1); 4 withdrew consent. Demographic and clinical characteristics of the 246 infants at the onset of the study are provided in Table 1. The number of participants in each BW category was 500 to 749 g (n = 39), 750 to 999 g (n = 72), 1000 to 1249 g (n = 61), and 1250 to 1750 g (n = 74). No significant differences were found between the 2 groups.
Table 2 presents the primary and secondary outcomes. There were 23 patients (13 in the BOL group and 10 in the CON group) who did not reach the endpoint of 120 mL · kg−1 · day−1 before 28 days. Six were lost to follow-up because of early discharge from the NICU to a regional hospital; intervention was discontinued in 3 patients because of NEC, and 13 died (Fig. 1). Furthermore, 1 infant from the BOL group failed to reach full enteral feeding within 28 days because of milk curd syndrome.
The number of days to achieve full feeding calculated from the initiation of feeding was not significantly different between the groups. The mean daily gastric residual volumes were significantly lower in the BOL group than in the CON group, as was the total number of patients with feeding interruptions. No differences in the other tolerance parameters or weight gain were found between groups (Table 2). Stratification for BW revealed that BW was inversely related to days to reach full enteral feeding.
In the Cox regression analysis, the sex and SGA/AGA values were entered to test for effect modification. A significant effect modification (P = 0.048) was observed for SGA versus AGA infants indicating a modified effect of bolus feeding on days required to reach full enteral feeding between the 2 groups. Because of this effect modification, we looked at the separate results for the AGA and SGA groups. The stratum-specific hazard ratios were 1.18 (0.88–1.59) for the AGA infants (n = 195) and 0.61 (0.36–1.09) for the SGA infants (n = 51). This implied that we cannot determine whether there is a clinical implication for this group.
The primary and secondary outcomes for children ≥1000 and <1000 g were similar to those in the general population (Table 2). The incidences of adverse effects were similar in both groups (Table 3). We stress that this study, however, was not sufficiently powered to determine the safety of either feeding mode with respect to the incidence of NEC or death.
This randomised controlled trial demonstrates that for preterm LBW infants, there are no differences in days to reach full enteral feeding, weight gain, or safety between CON or BOL feeding. Optimising enteral nutrition is very important for improving subsequent health, but very few studies have investigated different feeding strategies. No previous trial has compared bolus versus continuous feeding in a cohort of >171 patients. Therefore, a Cochrane meta-analysis with only 511 infants concluded that because of small sample size and methodological limitations, the existing evidence was insufficient to determine an optimal feeding strategy(8). Our study, which is the largest study to date, was designed to overcome these issues.
In line with the Cochrane meta-analysis(8), our study reveals that there is no difference in days to reach full enteral feeding between BOL and CON feeding. We did observe, however, a significant difference in feeding tolerance parameters between the 2 feeding strategies. A study by Schanler et al (20) supports these findings. They found that the percentage gastric residuals exceeding 50% of the 3-hour feeding volume was nearly twice as high in the continuous versus the bolus group. In contrast, they did observe a difference in time to reach full feeding in infants with a gestation age of 26 to 27 weeks between the 2 feeding strategies (20). We could not confirm this, however, in our group with a BW <1000 g.
Some limitations of our study design should be noted. The primary outcome of this study—“the days to achieve full enteral feeding”—was chosen because it is closely related to the combined effects of feeding tolerance, weight gain, respiratory support and complications. Bias, however, may have occurred because it was not feasible to blind the caregivers to treatment allocation and outcome.
Second, semicontinuous feeding was the standard feeding regime at our NICU before the start of the study. This is a very laborious method for the nursing staff; therefore, other NICU centres may use different continuous feeding methods. Whether the method of continuous feeding influences tolerance and growth remains unknown.
Third, we found statistical differences in the secondary outcomes “gastric residual volume” and “number of patients with feeding interruptions” between both feeding groups. Both outcomes are tightly linked because the feeding protocol dictated a feeding interruption, if gastric residuals exceeded 3-fold the hourly volume or the volume of the preceding bolus in the CON and BOL groups. It is inherent to the feeding method that infants fed via CON have small amounts of milk in their stomach versus the BOL group; the mean difference in gastric residuals was only 0.3 mL every 8 hours.
Gastric residual volume is often used as a marker to evaluate feeding tolerance, but it also serves as a warning sign for NEC (21,22)(21,22). Recent studies have demonstrated that only hemorrhagic or bloodstained residuals increase the risk of NEC. There is no literature consensus about the amount of residue that is clinically relevant (22–24)(22–24)(22–24). Hence, the importance of gastric residuals as a diagnostic sign is declining. Our present study supports this decline because the primary outcome did not differ between the groups. Moreover, the amount of residue that is clinically relevant remains unclear (22–24)(22–24)(22–24). It is questionable whether the difference found in our study has any clinical implications other than a possible increase in the number of patients with feeding interruptions. It is not possible to conclude from this study whether the higher amount of residuals inherent to CON feeding can solely explain the higher number of patients with feeding interruptions.
Furthermore, in this study, the primary outcome was reaching an intake of 120 mL · kg−1 · day−1, which is lower than the outcome used in comparable studies. Ideally, in all of the studies “full enteral feeding” should be defined as enteral feeding of ≥150 mL · kg−1 · day−1 for ≥72 uninterrupted hours. This would ensure tolerance and enable proper comparisons among studies. We, however, chose this lower cut off point because neonates who reached this amount were discharged from our NICU to a regional level II hospital that hosted a high care unit as soon as permitted by their clinical condition. In this way, we avoided losing patients to follow-up. Consequently, the incidence of clinical morbidities—particularly NEC—that could have occurred after achieving full enteral feeding may be underreported. Therefore, we also separately reported the occurrence of NEC after achieving full enteral feeding in our population.
The incidence of NEC in our study population including infants who became ill after the study period is relatively high (25,26)(25,26). It is a common practice in the Netherlands to refer low-risk premature infants to level II hospitals and only admit those infants who require complicated neonatal care to a level III hospital NICU. This may partly explain the relatively high incidence of adverse events, particularly NEC. Our study was neither designed nor powered to detect a difference in NEC rates between groups. Although no significant difference was observed, a meta-analysis of the larger groups should be conducted to further investigate this matter.
Finally, there is a considerable difference in time to reach full feeding in this population versus data from the historical cohort used for the power analysis. At the time of the start of the study, we used a feeding advancement protocol that was already in practice. It is well known from previous studies that participating in a nutritional research study improves adherence to the set protocol. This likely explains the shorter time required to reach full feeding versus historical data. There were no other changes with regard to the feeding regime.
We conclude that bolus feeding and continuous feeding are both suitable feeding strategies for LBW infants. BOL feeding, however, may be preferable because it allows more parental involvement in feeding.
The authors thank the parents of the study participants for consenting to the participation of their infants in this study. The authors also thank all of the help provided by the doctors and nurses of the neonatal ward of the Sophia Children's Hospital while conducting this study.
1. Ehrenkranz RA, Dusick AM, Vohr BR, et al. Growth in the neonatal intensive care unit influences neurodevelopmental and growth outcomes of extremely low birth weight infants. Pediatrics
2. Stephens BE, Walden RV, Gargus RA, et al. First-week protein and energy intakes are associated with 18-month developmental outcomes in extremely low birth weight infants. Pediatrics
3. Lucas A, Morley R, Cole TJ. Randomised trial of early diet in preterm babies and later intelligence quotient. BMJ
4. Embleton NE, Pang N, Cooke RJ. Postnatal malnutrition and growth retardation: an inevitable consequence of current recommendations in preterm infants? Pediatrics
5. Gargasz A. Neonatal and pediatric parenteral nutrition
. AACN Adv Crit Care
6. Kelly DA. Liver complications of pediatric parenteral nutrition
7. Jacobi SK, Odle J. Nutritional factors influencing intestinal health of the neonate. Adv Nutr
8. Premji SS, Chessell L. Continuous nasogastric milk feeding versus intermittent bolus milk feeding for premature infants less than 1500 grams. Cochrane Database Syst Rev
9. Grant J, Denne SC. Effect of intermittent versus continuous enteral feeding
on energy expenditure in premature infants. J Pediatr
10. Toce SS, Keenan WJ, Homan SM. Enteral feeding
in very-low-birth-weight infants
. A comparison of two nasogastric methods. Am J Dis Child
11. de Ville K, Knapp E, Al-Tawil Y, et al. Slow infusion feedings enhance duodenal motor responses and gastric emptying in preterm infants. Am J Clin Nutr
12. Baker JH, Berseth CL. Duodenal motor responses in preterm infants fed formula with varying concentrations and rates of infusion. Pediatr Res
13. Dani C, Pratesi S, Barp J. Continuous milk feeding versus intermittent bolus feeding in preterm infants. Early Hum Dev
2013; (89 suppl 2):S11–S12.
14. Rogers SP, Hicks PD, Hamzo M, et al. Continuous feedings of fortified human milk lead to nutrient losses of fat, calcium and phosphorous. Nutrients
15. Aynsley-Green A. New insights into the nutritional management of newborn infants derived from studies of metabolic and endocrine inter-relations during the adaptation to post-natal life. Proc Nutr Soc
16. Shulman RJ, Redel CA, Stathos TH. Bolus versus continuous feedings stimulate small-intestinal growth and development in the newborn pig. J Pediatr Gastroenterol Nutr
17. El-Kadi SW, Suryawan A, Gazzaneo MC, et al. Anabolic signaling and protein deposition are enhanced by intermittent compared with continuous feeding in skeletal muscle of neonates. Am J Physiol Endocrinol Metab
18. Blondheim O, Abbasi S, Fox WW, et al. Effect of enteral gavage feeding rate on pulmonary functions of very low birth weight infants. J Pediatr
1993; 122 (5 pt 1):751–755.
19. Usher R, McLean F. Intrauterine growth of live-born Caucasian infants at sea level: standards obtained from measurements in 7 dimensions of infants born between 25 and 44 weeks of gestation. J Pediatr
20. Schanler RJ, Shulman RJ, Lau C, et al. Feeding strategies for premature infants: randomized trial of gastrointestinal priming and tube-feeding method. Pediatrics
21. Lucchini R, Bizzarri B, Giampietro S, et al. Feeding intolerance in preterm infants. How to understand the warning signs. J Matern Fetal Neonatal Med
2011; 24 (suppl 1):72–74.
22. Mihatsch WA, von Schoenaich P, Fahnenstich H, et al. The significance of gastric residuals in the early enteral feeding
advancement of extremely low birth weight infants. Pediatrics
23. Bertino E, Giuliani F, Prandi G, et al. Necrotizing enterocolitis: risk factor analysis and role of gastric residuals in very low birth weight infants. J Pediatr Gastroenterol Nutr
24. Cobb BA, Carlo WA, Ambalavanan N. Gastric residuals and their relationship to necrotizing enterocolitis in very low birth weight infants. Pediatrics
2004; 113 (1 pt 1):50–53.
25. Horbar JD, Badger GJ, Carpenter JH, et al. Trends in mortality and morbidity for very low birth weight infants, 1991–1999. Pediatrics
2002; 110 (1 Pt 1):143–151.
26. Stoll BJ, Hansen NI, Bell EF, et al. Neonatal outcomes of extremely preterm infants from the NICHD Neonatal Research Network. Pediatrics
Keywords:© 2015 by European Society for Pediatric Gastroenterology, Hepatology, and Nutrition and North American Society for Pediatric Gastroenterology,
enteral feeding; low-birth-weight infants; nutrition; tolerance; weight gain