PEDIATRICS Vol. 110 No. 6 December 2002, pp. 1199-1203
Hydrolyzed Protein Accelerates Feeding Advancement inVery Low Birth Weight Infants
* From the Division of Neonatology and Pediatric Critical Care, Department of Pediatrics
Department for Biometry and Medical Documentation, Ulm University, 89070 Ulm, Germany
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
Background. Feeding intolerance is common in very low birth weight (VLBW; <1500 g) infants. Hydrolyzed protein preterm infant formula (HPF) has been shown to accelerate the gastrointestinal transit of formula. The aim of this study was to investigate whether HPF improves early feeding tolerance compared with standard preterm infant formula (SPF). We hypothesized that HPF would accelerate early enteral feeding advancement.
Methods. Primary outcome was the time from initiation of milk feeds until full feeds (150 mL/kg birth weight/d) were achieved in infants who received <10% human milk (HM) to exclude HM as a confounder. Because the availability of HM was not predictable at the time of enrollment, all eligible VLBW infants (n = 129) were randomly assigned in a randomized, controlled trial to receive HPF or SPF if HM was not available. Infants who received >10% HM (n = 42) were excluded. Milk bolus feeding every 2 to 3 hours was started at the discretion of the attending physician and advanced by 16 mL/kg/d. Preprandial gastric residuals were tolerated up to 5 mL/kg; otherwise, feedings were reduced or withheld. Data are shown as median (5th and 95th percentile).
Results. Forty-six and 41 (HPF vs SPF) infants received <10% HM. There was no significant difference with regard to birth weight, gestational age, and onset of milk feeds (day 3 [1–8] vs 4 [2–6]). The time from initiation of milk feeds to full feeds was significantly shorter with HPF feeding (10 [9–27] vs 12 [9–28] days).
Conclusion. HPF improved the feeding tolerance and enabled a more rapid establishment of full enteral feeding in VLBW infants compared with SPF.
Key Words: hydrolyzed protein formula • early enteral nutrition • very low birth weight infant
Abbreviations: VLBW, very low birth weight • HM, human milk • HPF, hydrolyzed protein preterm infant formula • SPF, standard preterm infant formula • NEC, necrotizing enterocolitis
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
Feeding intolerance characterized by vomiting, large gastric residuals, abdominal distension, and constipation is common in very low birth weight (VLBW; <1500 g) infants, especially within the first weeks of life. Human milk (HM) feeding has been shown to be associated with less feeding intolerance1,2 and is believed to be the first-choice milk in VLBW infants. However, HM is not always available and formula feeding is required.
In formula-fed infants, hydrolyzed cow’s milk protein has been shown to promote gastric emptying,3,4 to accelerate the gastrointestinal transit of formula,5 and to induce frequent loose stools6 if compared with native cow’s milk protein.
The aim of the present study was to investigate whether a hydrolyzed protein preterm infant formula (HPF) enables a more rapid establishment of full enteral feeding in VLBW infants (if HM is not available) compared with a standard (native protein) preterm infant formula (SPF).
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
Patients
From July 1999 until March 2001, all VLBW infants admitted to the Division of Neonatology of the Ulm University Children’s Hospital (level 3 referral center) were eligible for the trial. The study protocol was approved by the institutional ethics committee, and informed written parental consent was obtained. Exclusion criteria were major congenital malformations and anomalies that might interfere with nourishing. Eligibility was assessed after the attending physicians decided to initiate milk feeding to exclude potential physician’s bias regarding when to start milk feedings. The infants were stratified according to birth weight (<1000 g and 1001–1500 g) and randomly allocated to the 2 different study formulas (sealed envelopes, computer generated blocked randomization lists, block size of 4). All infants received parenteral nutrition, which was gradually decreased with increasing enteral intake.
Formulas
The standard preterm formula SPF (Aptamil Prematil, Milupa GmbH, 61381 Friedrichsdorf, Germany) contained native cow’s milk protein (Table 1). The hydrolyzed protein preterm formula HPF (Aptamil Prematil HA, Milupa, Germany) contained an ultra-filtrated mixture of hydrolyzed whey and hydrolyzed casein (molecular weight: 69% <1000 Dalton, 18% 1000–2000 D, 11.9% 2000–5000 D, 0.9% 5000–10000 D, 0.2% 10000–20000 D, and < 0.1% >20000 D; 15% free amino acids; data provided by the manufacturer). The protein concentration, the whey to casein ratio (60:40), and the lipid body of the study formulas were identical. There were only minor differences with regard to mineral concentration, osmolarity (Table 1), and amino acid compositions of the proteins.7 To compensate for the increased osmolarity of the hydrolyzed protein, the predominant carbohydrate in HPF was a glucose polymer (maltodextrin, corn starch) compared with lactose in SPF. The formulas were supplied in ready-to-feed form. The study period covered the first 4 weeks of life, and the study formula was fed from the beginning of milk feeding until at least day 28.
|
Feeding Schedule
Within the first hours of life, bolus gavage 5% glucose feeding every 2 to 3 hours was started at a volume of 16 mL/kg birth weight per day. As soon as a sufficient amount of meconium had been passed, bolus milk feeding was started at the discretion of the attending physician. HM feeding was encouraged, and formula was fed only if HM was not available. Although the trial was designed and conducted in a double-blind fashion, true masking of the formulas was impossible because the taste was different. The infants were fed every 2 (<1250 g) or 3 (1251–1500 g) hours, starting at 16 mL/kg birth weight per day. Feedings were advanced every 24 hours by 16 mL/kg birth weight per day whenever >50% of the calculated amount was given during the previous 24 hours. The gastric residual volume was checked before each feeding. The scheduled feed was given at a gastric residual volume
5 mL/kg; otherwise, the difference up to the scheduled amount was given.8 A feed was skipped if the gastric residual volume was >5 mL/kg and the volume of this scheduled feed was less than the gastric residual volume. The color of the gastric residuals, the subjective impression of those in charge, and clinical conditions such as hypotension, mild abdominal distension, infection, umbilical catheters in place, or Indomethacin therapy were not allowed to change the feeding strategy. However, the feedings were temporarily discontinued and a sepsis workup or an abdominal radiograph were performed as appropriate for the following reasons: a sudden significant increase in gastric residuals, an abnormal abdominal examination, and if the infants did not pass a sufficient amount of stool on a regular basis. The feedings were resumed as soon as the abdominal investigation was normal.
An abnormal abdominal examination was defined as follows: gross abdominal distension; persistent visible bowel loops without peristalsis; abdominal tenderness; decreased abdominal bowel sounds; sudden increment in the abdominal girth (>2 cm); and blood in the stools. Five percent glucose glycerin enemas (10:1 mixture, 5 mL/kg) were administered twice a day until the infants passed at least 1 stool per day spontaneously. After extubation or intubation, feedings were withheld for 6 hours. Laxatives were not administered during the study. Oral medications were not given until full enteral feeds were attained.
Data Collection and Analysis
Clinical characteristics were recorded for all infants. With every feeding, beginning with the initiation of milk feedings and ending as soon as full enteral feedings were achieved, feeding tolerance was assessed continuously by the following characteristics: preprandial gastric residual volume, feeding volume, and the number of stools. Necrotizing enterocolitis (NEC) was diagnosed according to the modified Bell stages.9 Weight was recorded on days 21 and 28.
Statistical Analysis
The a priori determined primary outcome was the time from initiation of milk feeds until full feeds (150 mL/kg birth weight/d) were achieved in the infants who received <10% HM or formula only. If full feeds were not achieved within the first 28 days of life (study period), the 28th day was used as censored value. Secondary outcomes were the age at full feeds, stool frequency (proportion of diapers containing meconium or brownish stool irrespective whether an enema has been given before), gastric residual volume, proportion of negative gastric residuals (volume = 0), proportion of gastric residuals >5 mL/kg birth weight, incidence of proven NEC (Bell stage 
II),9 and weight at days 21 and 28. Finally, the data of all recruited infants, including those who had received >10% HM, was analyzed as secondary outcome as well.
Because both study formulas have not been given before, there was no pilot data available. Therefore, an adaptive interim analysis10 was chosen to estimate the standard deviation of the primary outcome (1-sided Wilcoxon test) and then to calculate the final sample size (overall 
level of 5%; ß = 20%), as soon as the results of at least 40 infants in each group were available. This method considers the observed error probabilities from the subsamples before and after the interim analysis, which are combined into a global statistical test. The stopping rules in terms of critical limits for the P value in the first subsample (interim analysis, P1) were defined as follows: The trial was to be stopped early with rejection of H0 in case of P1 .0299 and with acceptance of H0 in case of P1 .3. These critical levels in the adaptive interim analysis guaranteed an 1-sided overall level of 5%.10
Two-sided Mann-Whitney U tests were used for secondary outcome analysis without correction for multiple testing (level of significance P < .05, Statistical Systems for Personal Computers Package, SAS Institute, Cary, NC). Data are presented as median (5th to 95th percentile).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
In March 2001, the data of at least 40 infants in each group who received <10% HM was available. From July 1999 to March 2001, 157 VLBW infants were admitted. Eighteen infants were excluded for the following reasons: major malformations (n = 10), severe hydrops (n = 1), meconium plug syndrome requiring laparotomy (n = 2), severe sepsis (n = 1, death on day 3), NEC on the first day of life (n = 1, clostridium difficile sepsis, death on day 3), early transfer (n = 1), and withdrawal of therapy (n = 2). Ten of these infants died within the first 2 weeks of life. Two infants were not enrolled for no apparent reason. Parents of the remaining 137 infants were approached, and informed written consent was obtained in 135 infants. No participant withdrew. Six infants had to be excluded after randomization because they were transferred within 5 days after randomization to affiliated hospitals because of limited capacity of the intensive care unit (HPF, n = 1; SPF, n = 5). Forty-two infants had to be excluded because they received >10% HM before full feeds were achieved (HPF, n = 20; SPF, n = 22). The data of the remaining 87 VLBW infants (46 HPF, 41 SPF) was included in the primary outcome analysis.
There was no significant difference between the groups with regard to the clinical characteristics and the age when milk feedings were initiated: 3 (1–8) days with HPF and 4 (2–6) days with SPF (Table 2). However, significantly less time was required to achieve full feeds after the onset of milk feedings with HPF feeding (10 [9–27] vs 12 [9–28] days, P = .0024; primary study outcome). The study was stopped early with rejection of H0 because the level of significance of the interim analysis was far below the critical limit (.0299). The observed P1 level would also have met the stopping criterion of P1 < .0045, which is connected to an overall P of <.01 using the same procedure of stochastic curtailment.
|
With regard to secondary outcome analysis, the age at 150 mL/kg birth weight per day was significantly lesser with HPF feeding. Diapers were changed at the same frequency in both groups (Table 3); however, significantly more stools were passed with HPF feeding. There was no significant difference among the groups with regard to the proportion of negative gastric residuals (volume = 0), the mean gastric residual volume, and the proportion of gastric residuals that were above 5 mL/kg birth weight. There was also no difference with regard to growth measured by weight on days 21 and 28; however, data on growth was not available in all infants, because 9 HPF infants and 6 SPF infants on full feeds have been transferred to affiliated hospitals before 28 days of age or discharged.
|
The data of all recruited infants including those who had received >10% HM (66 and 63 infants; HPF vs SPF) again showed no difference between the groups with regard to birth weight (895 [515–1448] g and 880 [532–1476] g), gestational age (28 [23–32] and 28 [24–31] weeks) and the age at initiation of milk feeds (3 [1–8] days and 4 [2–8] days) but with HPF feeding full feeds were achieved significantly faster after initiation of milk feeds (10 [9–23] days vs 12 [9–28] days; P = .003) and the age at 150 mL/kg birth weight per day was significantly less (13 [10–26] days and 16 [11–31] days; P = .01). There were 3 cases of NEC (Bell stage
II) (HPF, n = 1, day 16; SPF, n = 2, days 17 and 43) and 3 cases of spontaneous intestinal perforation (HPF, n = 1, day 8; SPF, n = 2, days 7 and 17). |
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
VLBW infants frequently experience feeding difficulties when given the currently available premature infant formulas, and feeding advancement can be a difficult issue. In the face of the physiologically immature forward propulsion in preterm infants,11 the optimal preterm infant formula for early enteral nutrition should stimulate the gastrointestinal transit. Loose stools are a well-known side effect of "hypoallergenic" hydrolyzed protein formulas in infant feeding. In VLBW infants on full feeds, it has recently been shown that hydrolyzed protein improved the feeding tolerance by accelerating the gastrointestinal transit of formula substantially and increasing the stool frequency.5 Confirming these results, in the current study, HPF was associated with a higher stool frequency, and full enteral feeds were established substantially more rapidly. The median time of parenteral nutrition was reduced from 16 to 13 days. These data suggests that HPF may improve the overall outcome because faster feeding advancement saves additional days of parenteral nutrition and reduces the risk of nosocomial infections and of parenteral nutrition associated complications such as venous thromboses.12,13 In addition, costs may be reduced because oral nutrition is cheaper and requires less monitoring. However, these hypothesized advantages of hydrolyzed protein formulas still have to be proved in additional randomized, controlled trials.
There have been concerns as to whether hydrolyzed protein formulas meet the nutritional needs of VLBW infants.14 However, appropriate plasma amino acid concentrations, adequate nitrogen retention after correction for intake, and even a tendency toward a higher calcium absorption have been reported in preterm infants with modern protein hydrolysate formulas.7,15,16 The only long-term (12 weeks) study in preterm infants on growth did not find any disadvantage in comparison to standard preterm formula.15
There were minor differences between the study formulas in addition to protein hydrolysis: HPF contained more maltodextrin and less lactose than SPF (Table 1). These differences were chosen to compensate for the higher osmotic load of the hydrolyzed protein. Finally, HPF ended up with a slightly lower osmolarity. Obviously, the study design does not permit to elaborate the different effects of each of these minor differences separately. However, we speculate that protein hydrolysis increased the feeding tolerance, because in a previous study it has been shown that a formula that contained the same protein hydrolysate accelerated intestinal transit substantially compared with a standard formula with an identical lactose:maltodextrin ratio but a lower osmolarity.5 Nevertheless, the different lactose:maltodextrin ratio may also have had an effect on feeding tolerance.2,17
Although the study was a randomized, controlled trial, there are some limitations. First of all, formula was fed only if HM was not available. Therefore, the availability of HM was anticipated to be a confounder. However, analysis of the data of all infants (as randomized) and analysis of the subgroup of infants who received <10% HM came to the same result. Second, true masking of the formulas was impossible because of a major difference (bitter taste of hydrolyzed protein). Finally, velocity of feeding advancement depended on the feeding protocol itself. In VLBW infants, feedings have been advanced at velocities up to 35 mL/kg/d without increasing the incidence of NEC.18 In our clinical experience, daily increments of 16 mL/kg birth weight were appropriate, but some infants might have tolerated a faster rate as well. We speculate that the difference between the formulas would have been even more pronounced if a higher rate of feeding advancement had been used. Preprandial gastric residuals were zero approximately 50% of the time, and increased gastric residuals were found in 
3% (Table 3). We believe this is a rational basis to increase the feeding advancement further to 20 mL/kg birth weight per day as a next step. The optimal rate at which feedings should be advanced in VLBW infants still needs to be defined.
The mechanism by which HPF improves early feeding tolerance is not known. First, a certain degree of impaired nutrient absorption might increase the intraluminal osmotic load and accelerate the intestinal transit. Second, hydrolyzed protein formula has been suggested to induce higher motilin levels than native protein formula.19 Finally, hydrolysis of the protein may have accelerated gastrointestinal transit via a reduced activity of milk protein-derived opioid receptor agonists (ß-casomorphins). Casein-predominant formula has been shown to be associated with a longer gastrointestinal transit time than a whey-predominant formula in infants20 and rats.21 In rats, this difference has been abolished by pretreatment with the opiate-receptor antagonist naloxone.21 Mammalian milk contains highly specific and potent µ-type opioid receptors agonists (eg, ß-casomorphins).22 It is a well-known side effect of opioids to reduce bowel activity, and the opioid content of predigested formulas has been shown to be reduced.23
|
CONCLUSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
HPF improved the early feeding tolerance and enabled a more rapid establishment of full enteral feeding in VLBW infants. HPF was studied as a starter formula. SPF will still remain the standard formula as long as the long-term safety of HPF has not been established in preterm infants.
|
ACKNOWLEDGMENTS |
|---|
The study was supported by Milupa GmbH, 61381 Friedrichsdorf, Germany.
|
FOOTNOTES |
|---|
Received for publication Jan 29, 2002; Accepted Jul 3, 2002.
Reprint requests to (W.A.M.) Universitäts-Kinderklinik, 89070 Ulm, Germany. E-mail: walter.mihatsch{at}medizin.uni-ulm.de
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
- Schanler RJ, Shulman RJ, Lau C. Feeding strategies for premature infants: beneficial outcomes of feeding fortified human milk versus preterm formula [commentary].
Pediatrics.1999; 103
:1150
–1157
[Abstract/Free Full Text] - Mihatsch WA, von Schoenaich P, Fahnenstich H, et al. Randomized multi-center trial of two different formulas for very early enteral feeding advancement in extremely low birth weight infants. J Pediatr Gastroenterol Nutr.2001; 33 :155 –159[CrossRef][Web of Science][Medline]
- Tolia V, Lin CH, Kuhns LR. Gastric emptying using three different formulas in infants with gastroesophageal reflux [see comments]. J Pediatr Gastroenterol Nutr.1992; 15 :297 –301[Web of Science][Medline]
- Billeaud C, Guillet J, Sandler B. Gastric emptying in infants with or without gastro-oesophageal reflux according to the type of milk. Eur J Clin Nutr.1990; 44 :577 –583[Web of Science][Medline]
- Mihatsch WA, Högel J, Pohlandt F. Hydrolyzed protein accelerates the gastro-intestinal transport of formula in preterm infants. Acta Paediatr.2001; 90 :196 –198[CrossRef][Web of Science][Medline]
- Hyams JS, Treem WR, Etienne NL, et al. Effect of infant formula on stool characteristics of young infants.
Pediatrics.1995; 95
:50
–54
[Abstract/Free Full Text] - Mihatsch WA, Pohlandt F. Protein hydrolysate formula maintains homeostasis of plasma amino acids in preterm infants. J Pediatr Gastroenterol Nutr.1999; 29 :406 –410[CrossRef][Web of Science][Medline]
- 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.2002; 109
:457
–459
[Abstract/Free Full Text] - Walsh MC, Kliegman RM. Necrotizing enterocolitis: treatment based on staging criteria. Pediatr Clin North Am.1986; 33 :179 –201[Web of Science][Medline]
- Bauer P, Kohne K. Evaluation of experiments with adaptive interim analyses. Biometrics.1994; 50 :1029 –1041[CrossRef][Web of Science][Medline]
- Berseth CL. Gastrointestinal motility in the neonate. Clin Perinatol.1996; 23 :179 –190[Web of Science][Medline]
- Kawagoe JY, Segre CA, Pereira CR, Cardoso MF, Silva CV, Fukushima JT. Risk factors for nosocomial infections in critically ill newborns: a 5-year prospective cohort study. Am J Infect Control.2001; 29 :109 –114[CrossRef][Web of Science][Medline]
- Mahieu LM, DeDooy JJ, Lenaerts AE, Ieven MM, DeMuynck AO. Catheter manipulations and the risk of catheter-associated bloodstream infection in neonatal intensive care unit patients. J Hosp Infect.2001; 48 :20 –26[CrossRef][Web of Science][Medline]
- Rigo J, Salle BL, Picaud JC, Putet G, Senterre J. Nutritional evaluation of protein hydrolysate formulas. Eur J Clin Nutr.1995; 49(suppl 1) :S26 –S38
- Szajewska H, Albrecht P, Stoitiska B, Prochowska A, Gawecka A, Laskowska-Klita T. Extensive and partial protein hydrolysate preterm formulas: the effect on growth rate, protein metabolism indices, and plasma amino acid concentrations. J Pediatr Gastroenterol Nutr.2001; 32 :303 –309[CrossRef][Web of Science][Medline]
- Picaud JC, Rigo J, Normand S, et al. Nutritional efficacy of preterm formula with partially hydrolyzed protein source: a randomized pilot study. J Pediatr Gastroenterol Nutr.2001; 32 :555 –561[CrossRef][Web of Science][Medline]
- Griffin MP, Hansen JW. Can the elimination of lactose from formula improve feeding tolerance in premature infants? J Pediatr.1999; 135 :587 –592[CrossRef][Web of Science][Medline]
- Rayyis SF, Ambalavanan N, Wright L, Carlo WA. Randomized trial of slow versus fast feed advancements on the incidence of necrotizing enterocolitis in very low birth weight infants. J Pediatr.1999; 134 :293 –297[CrossRef][Web of Science][Medline]
- Tormo R, Potau N, Infante D, Moran J, Martin B, Bergada A. Protein in infant formulas. Future aspects of development. Early Hum Dev.1998; 53(suppl) :S165 –S172
- Brown GA, Berger HM, Brueton MJ, Scott PH, Wharton BA. Nonlipid formula components and fat absorption in the low-birth-weight newborn.
Am J Clin Nutr.1989; 49
:55
–61
[Abstract/Free Full Text] - Daniel H, Vohwinkel M, Rehner G. Effect of casein and beta-casomorphins on gastrointestinal motility in rats. J Nutr.1990; 120 :252 –257
- Teschemacher H, Koch G, Brantl V. Milk protein-derived opioid receptor ligands. Biopolymers.1997; 43 :99 –117[CrossRef][Web of Science][Medline]
- Sturner RA, Chang JK. Opioid peptide content of predigested infant formulas [abstract]. Pediatr Res.1991; 29 :114A
- Voigt M, Schneider KT, Jährig K. [Analysis of a 1992 birth sample in Germany. 1: New percentile values of the body weight of newborn infants]. Geburtshilfe Frauenheilkd.1996; 56 :550 –558[Medline]
PEDIATRICS (ISSN 1098-4275). ©2002 by the American Academy of Pediatrics
CiteULike 
Connotea 
Del.icio.us 
Digg 
Facebook 
Reddit 
Technorati 
Twitter What's this?
This article has been cited by other articles:
|
|
 |
|
  C. L. Berseth, W. H. Johnston, S. I. Stolz, C. L. Harris, and S. H. Mitmesser Clinical Response to 2 Commonly Used Switch Formulas Occurs within 1 Day Clinical Pediatrics, January 1, 2009; 48(1): 58 - 65. [Abstract] [PDF]
|
|
|
|
 |
|
 A. R. Franz, F. Pohlandt, H. Bode, W. A. Mihatsch, S. Sander, M. Kron, and J. Steinmacher Intrauterine, Early Neonatal, and Postdischarge Growth and Neurodevelopmental Outcome at 5.4 Years in Extremely Preterm Infants After Intensive Neonatal Nutritional Support Pediatrics, January 1, 2009; 123(1): e101 - e109. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
