Early progressive feeding in extremely preterm infants:

a randomized trial
Ariel A Salas,1 Peng Li,2 Kelli Parks,1 Charitharth V Lal,1 Camilia R Martin,3 and Waldemar A Carlo1
1 Department of Pediatrics, School of Medicine; and 2 Department of Biostatistics, School of Public Health, University of Alabama at Birmingham, Birmingham,

AL; and 3 Department of Neonatology and Division of Translational Research, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA

ABSTRACT INTRODUCTION
Background: Due to insufficient evidence, extremely preterm in- In many hospitals that provide neonatal care for extremely
fants (≤28 wk of gestation) rarely receive early progressive feed- preterm infants (≤28 wk of gestation), the transition from par-
ing (small increments of feeding volumes between 1 and 4 d enteral to enteral nutrition often begins with minimal enteral
after birth). We hypothesized that early progressive feeding increases
feeding (MEF) or trophic feeding (≤24 mL  kg−1  d−1 ), then
the number of full enteral feeding days in the first month after
changes to progressive feeding (daily increments of feeding vol-
birth.
umes usually by 20–24 mL  kg−1  d−1 ), and concludes with full
Objective: The aim of this study was to determine the feasibility and
enteral feeding (≥120 mL  kg−1  d−1 ) (1–4). By initiating a 3-
efficacy of early progressive feeding in extremely preterm infants.
to 5-d course of MEF within the first 96 h after birth (1), most
Design: In this single-center randomized trial, extremely preterm in-
clinicians assume that prevention of gastrointestinal atrophy will
fants born between September 2016 and June 2017 were randomly
reduce the risk of feeding intolerance and necrotizing enterocol-
assigned to receive either early progressive feeding without trophic
feeding (early feeding group) or delayed progressive feeding after a
itis (NEC) in extremely preterm infants (5–7).
4-d course of trophic feeding (delayed feeding group). Treatment al- A meta-analysis of 9 randomized trials that compared early
location occurred before or on feeding day 1. The primary outcome with delayed progressive feeding in predominantly moderate-
was the number of full enteral feeding days in the first month af- preterm infants (29–32 wk of gestation) (8) and a retrospective
ter birth. Secondary outcomes were death, necrotizing enterocolitis study that compared short with extended periods of trophic feed-
(NEC), culture-proven sepsis, growth percentiles at 36 wk postmen- ing in extremely preterm infants (1) provide clinicians with ev-
strual age, use of parenteral nutrition, and need for central venous idence that early progressive feeding (small increments of feed-
access. ing volumes between 1 and 4 d after birth) reduces the time to
Results: Sixty infants were included (median gestational age: 26 wk; establish full enteral feeding without increasing the risk of NEC.
mean ± SD birth weight: 832 ± 253 g). The primary outcome dif- However, many clinicians considered this evidence insufficient to
fered between groups (median difference favoring the early feeding standardize the practice of early progressive feeding in extremely
group: +2 d; 95% CI: 0, 3 d; P = 0.02). Early progressive feed- preterm infants. Feasibility, safety, and efficacy are the main
ing reduced the use of parenteral nutrition (4 compared with 8 d; concerns (9, 10).
P ≤ 0.01) and the need for central venous access (9 compared with Because retrospective studies introduce selection bias medi-
13 d; P ≤ 0.01). The outcome of culture-proven sepsis (10% com- ated by severity of illness (1, 11) and randomized trials often ex-
pared with 27%; P = 0.18), restricted growth (weight, length, and clude extremely preterm infants (10, 12), this randomized trial
head circumference <10th percentile) at 36 wk postmenstrual age assessed the feasibility and efficacy of early progressive feeding
(25% compared with 50%; P = 0.07), and the composite outcome
of NEC or death (27% compared with 20%; P = 0.74) did not differ
between groups. Supported in part by a research award from the Gerber Foundation.
Conclusion: Early progressive feeding increases the number of full WAC and PL are also supported by the NIH (grants U10 HD34216 and
enteral feeding days in extremely preterm infants. This trial was reg- UL1TR001417, respectively).
Address correspondence to AAS (e-mail: [email protected]).
istered at www.clinicaltrials.gov as NCT02915549. Am J Clin
Abbreviations used: MEF, minimal enteral feeding; NEC, necrotizing ente-
Nutr 2018;107:365–370.
rocolitis; PN, parenteral nutrition; SGA, small for gestational age; SIP, spon-
taneous intestinal perforation.
Keywords: minimal enteral nutrition, parenteral nutrition, necrotiz- Received December 15, 2017. Accepted for publication January 11, 2018.
ing enterocolitis, central venous access, late-onset sepsis, postnatal First published online February 23, 2018; doi: https://doi.org/10.1093/
growth restriction, premature infants ajcn/nqy012.

Am J Clin Nutr 2018;107:365–370. Printed in USA. © 2018 American Society for Nutrition. All rights reserved. 365
366 SALAS ET AL.

in extremely preterm infants. We hypothesized that in extremely feeding after the diagnosis of NEC or spontaneous intestinal per-
preterm infants receiving human milk, progressive feeding with- foration (SIP) was not regulated by the study protocol.
out MEF compared with delayed progressive feeding after a 4-d The primary efficacy endpoint of the trial was the number of
course of MEF would result in an increased number of full enteral full enteral feeding days in the first month after birth. Secondary
feeding days in the first month after birth. efficacy endpoints of the trial were time to establish full enteral
feeding, use of parenteral nutrition (PN) in days, use of central ve-
nous access in days, culture-proven sepsis, growth percentiles at
METHODS 36 wk postmenstrual age or time of hospital discharge (whichever
In this parallel-group randomized controlled trial, participants occurred first), and duration of hospital stay in days.
were randomly assigned in a 1:1 allocation ratio to receive either The primary safety endpoints of the trial were death, NEC
early progressive feeding without MEF (early feeding group) or stage 2 or 3, or SIP. Due to insufficient power to detect signifi-
delayed progressive feeding after a 4-d course of MEF (delayed cant group differences in these primary safety endpoints (∼25%),
feeding group). Extremely preterm infants with gestational ages a data safety and monitoring committee reviewed individual pa-
between 22 and 28 wk of gestation admitted to the neonatal unit tient data at 50% enrollment to exclude the possibility of a tempo-
at the University of Alabama at Birmingham Hospital were in- ral association between the intervention under investigation and
cluded. Infants born small for gestational age (SGA) with a birth the primary safety endpoints of the trial.
weight below the fifth percentile were excluded. Infants with a A sample size of 48 patients achieved 80% power to detect a
terminal illness in whom decisions to withhold or limit life sup- 5-d difference in the number of full enteral feeding days in the
port were made and infants with major congenital or chromoso- first month after birth with an SD of 6 d under the 0.05 signif-
mal anomalies were also excluded. icance level. However, anticipating that ∼20% of study partici-
This trial was approved by the University of Alabama pants would not be able to complete the intervention as assigned
at Birmingham Institutional Review Board (clinicaltrials.gov: due to acute complications (3), 6 patients were added to each
NCT02915549). Written informed consent was obtained during group and the sample size was increased to 60.
the first 48 h after birth to allow treatment allocation before or on All of the continuous endpoints were summarized as
feeding day 1, usually between 48 and 96 h after birth. Partici- means ± SDs or as medians and IQRs. The categorical endpoints
pants were randomly assigned to one of the study groups follow- were summarized as frequencies and proportions. Group differ-
ing computer-generated random-block sequences and with the ences were evaluated by using the Wilcoxon test for continuous
use of sequentially numbered, opaque, sealed envelopes, which variables and chi-square test or Fisher’s exact test for categorical
were opened in sequential order only after informed consent was variables. The effect size for the primary efficacy endpoint was
obtained. Twin infants were randomly assigned individually. The expressed as the median difference with 95% CIs. Differences in
intervention was not masked. mean values were reported for other continuous endpoints. RRs
Enteral nutrition was administered as an intermittent bolus with 95% CIs were reported for categorical endpoints.
gavage every 3 h. Infants in the early feeding group received 20– We also performed a prespecified time-to-full-enteral-feeding
24 mL enteral nutrition  kg−1  d−1 on feeding day 1. On feeding analysis by using the Kaplan-Meier method and the log-rank test.
day 2, early progressive feeding began with daily increments of For this analysis, infants who died or developed NEC before post-
24–25 mL  kg−1  d−1 and continued until full enteral feeding natal day 28 were censored.
was established (≥120 mL  kg−1  d−1 ). Infants in the delayed All of the efficacy and safety endpoints of the trial were ana-
feeding group received 20–24 mL enteral nutrition  kg−1  d−1 lyzed with the intention-to-treat principle by using SAS 9.4 (SAS
from feeding day 1 to feeding day 4. On feeding day 5, progres- Institute).
sive feeding began with daily increments of 24–25 mL  kg−1 
d−1 until full enteral feeding was established. Unfortified donor
human milk was offered as an alternative to mother’s own milk
until full enteral feeding was established. Subsequently, infant RESULTS
formula was offered as an alternative to human milk if the mother Of 114 eligible extremely preterm infants admitted between
was no longer able to supply her own milk. September 2016 and June 2017, 60 extremely preterm infants
Although a birth weight–based feeding protocol was created with gestational ages of 22 wk, 0 d, through 28 wk, 6 d, were ran-
for each study participant to standardize daily rates of progres- domly assigned to receive either early progressive feeding with-
sive feeding (24–25 mL  kg−1  d−1 ) and verify compliance, out MEF or delayed progressive feeding after a 4-d course of
deviations from the feeding protocol were allowed. Enteral feed- MEF. Four infants in the early feeding group died of respiratory
ing discontinued for <5 d due to feeding intolerance or clinical distress syndrome before initiation of progressive feeding. Three
deterioration was resumed at the clinician’s discretion with the infants in the delayed feeding group died of sepsis before estab-
feeding volume defined in the birth-weight–based feeding proto- lishment of full enteral feeding (Figure 1).
col (preferred approach), with a feeding volume previously tol- Baseline characteristics of the study participants are shown in
erated, or with feeding volumes that did not meet any of those Table 1. Mean ± SD birth weight was 832 ± 253 g, and the
criteria. If enteral feeding was discontinued for ≥5 d, infants in median gestational age was 26 wk (IQR: 24–28 wk). More than
the early feeding group received a feeding volume of 20–24 mL  one-half of infants were of non-Hispanic black race/ethnicity. All
kg−1  d−1 on day 1 of re-initiation of enteral feeding before re- of the study participants received either mother’s own milk or
ceiving progressive feeding, and infants in the delayed feeding unfortified donor human milk during the first 2 wk after birth.
group received a feeding volume of 20–24 mL  kg−1  d−1 for Subsequently, approximately one-third of the study participants
4 d before receiving progressive feeding. Re-initiation of enteral received formula.
 EARLY FEEDING IN EXTREMELY PRETERM INFANTS 367

114 Assessed for eligibility

54 Excluded
9 did not meet inclusion criteria
26 refused to participate
19 other reasons

60 Randomly assigned

30 Assigned to the Early Feeding Group 30 Assigned to the Delayed Feeding Group
25 Received intervention as assigned (standard of care group)
5 Did not receive assigned intervention 28 Received intervention as assigned
1 developed SIP before initiation of feeding 2 Did not receive assigned intervention
4 died before initiation of progressive feeding 2 had enteral trophic feeding discontinued

2 Discontinued intervention before full feeding 4 Discontinued intervention before full feeding
1 died from NEC on day 13 1 developed NEC on day 8
1 died from RDS/sepsis on day 34 3 died from sepsis
2 Lost to follow-up after full enteral feeding 2 Lost to follow-up after full enteral feeding
1 developed NEC on day 20 1 developed NEC on day 17 and died on day 21
1 died from a viral infection on day 144 1 developed NEC on day 19

30 Included in analysis 30 Included in analysis

0 Excluded from analysis 0 Excluded from analysis

FIGURE 1 Study enrollment, randomization, and outcomes. NEC, necrotizing enterocolitis; RDS, respiratory distress syndrome; SIP, spontaneous intesti-
nal perforation.

The median number of full enteral feeding days in the first 28 d 3 d; P = 0.02) (Table 2). In the time-to-event analysis, the me-
after birth was 19 d (IQR: 0–20 d) in the early feeding group dian time to establish full enteral feeding was 9 d in the early
and 17 d (IQR: 13–18 d) in the delayed feeding group (median feeding group and 12 d in the delayed feeding group (P = 0.01)
difference favoring the early feeding group: +2 d; 95% CI 0, (Figure 2).

TABLE 1
Baseline characteristics

Early feeding group (n = 30) Delayed feeding group (n = 30)

Demographic characteristics
Birth weight, g 873 ± 2691 793 ± 234
Gestational age, wk 26 (24–28)2 26 (24–27)
Weight-for-age z score3 −0.10 ± 0.86 −0.17 ± 0.82
Male, n (%) 13 (43) 9/30 (30)
Black race, n (%) 15 (50) 20 (67)
Exposure to a full course (2 doses) of 22 (73) 17 (57)
antenatal steroids, n (%)
Apgar score at 5 min 6 (4–7) 6 (4–7)
Initiation of enteral feeding, d 3 (1–3) 3 (1–3)
Human milk–based diet with the use of
mother’s own milk (>80%) in the first 28 d
after birth, n (%)
Week 1 13 (43) 16 (53)
Week 2 13 (43) 10 (33)
Week 3 9 (30) 11 (37)
Week 4 8 (27) 10 (33)
Formula-based diet (>80%) in the first 28 d
after birth, n (%)
Week 3 11 (37) 7 (23)
Week 4 12 (40) 12 (40)
1 Mean ± SD (all such values).
2 Median; 25th–75th percentile in parentheses (all such values).
3 z Scores were estimated by using the Fenton 2013 growth curves.
368 SALAS ET AL.
TABLE 2
Feeding and safety outcomes1

Early feeding Delayed feeding

Outcomes group (n = 30) group (n = 30) P
Full enteral feeding in the first 28 d after 19 (0–20)2 17 (13–18) 0.023
birth, d
Time to full enteral feeding, d 10 ± 34 12 ± 2 0.00035
Duration of parenteral nutrition, d 4±6 8±6 0.00055
Duration of central venous access, d 9±7 13 ± 6 0.00015
Culture-proven sepsis, n (%) 3 (10) 8 (27) 0.186
Duration of mechanical ventilation, d 8±9 10 ± 12 0.615
Supplemental oxygen at 36 wk,7 n (%) 12 (50) 13 (50) 1.006
Weight <10th percentile at 36 wk,7,8 n (%) 12 (50) 16 (62) 0.416
Length <10th percentile at 36 wk,7,8 n (%) 13 (54) 18 (69) 0.276
Head circumference <10th percentile at 9 (38) 16 (62) 0.096
36 wk,7,8 n (%)
Restricted growth at 36 wk (weight, length, 6 (25) 13 (50) 0.076
and head circumference <10th
percentile),7,8 n (%)
NEC, n (%) 2 (7) 3 (10) 1.006
Mortality before postnatal day 28, n (%) 5 (16) 3 (10) 0.706
Mortality, n (%) 7 (23) 4 (12) 0.376
NEC or death, n (%) 8 (27) 6 (20) 0.566
Age at the time of hospital discharge, d 74 (53–92) 80 (66–92) 0.205
1 NEC, necrotizing enterocolitis.
2 Median; 25th–75th percentile in parentheses (all such values).
3 Derived by using Wilcoxon’s rank-sum test. Normal distribution could not be assumed because the early feeding group had 2

distinctive peaks in the distribution of the outcome (bimodal distribution).

4 Mean ± SD (all such values).
5 Derived by using t test for independent samples assuming equal variances.
6 Derived by using chi-square test or Fisher’s exact test if some cells have an expected count of ≤5.
7 Only participants with outcome data at 36 wk (n = 50): 24 in the early feeding group and 26 in the delayed progressive group.
8 Percentiles were estimated by using the Fenton 2013 growth curves.

The use of PN (4 compared with 8 d; P = 0.0005) and the This trial confirms the results of our retrospective study that
use of central venous access (9 compared with 13 d; P = 0.0001) compared short with extended periods of trophic feeding in 192
were also significantly lower in the early feeding group. The dif- extremely preterm infants (1). After adjustment for birth weight,
ferences in culture-proven sepsis and growth outcomes at 36 wk gestational age, SGA status, race, sex, type of enteral nutri-
postmenstrual age did not reach significance, but the risk of tion, and day of initiation of trophic feeding, we previously con-
culture-proven sepsis (RR: 0.38; 95% CI: 0.11, 1.28; P = 0.12) cluded that a short period of trophic feeding is associated with
and the risk of restricted growth (weight, length, and head circum- early establishment of full enteral feeding (1). Our results also
ference <10th percentile) at 36 wk of postmenstrual age (RR: corroborate the results of a meta-analysis that included >1000
0.50; 95% CI: 0.23, 1.10; P = 0.07) tended to be lower in the moderately preterm infants randomly assigned to receive early
early feeding group. The risk of NEC (RR: 0.67; 95% CI: 0.12, progressive feeding, in which early progressive feeding reduced
3.71; P = 0.68), death (RR: 1.98; 95% CI: 0.64, 6.11; P = 0.25), the time to establish full enteral feeding (8). The largest random-
and the combined outcome of NEC or death (RR: 1.33; 95% CI: ized trial included in the meta-analysis also concluded that early
0.53, 3.38; P = 0.56) did not differ between groups (Table 2). progressive feeding reduces the duration of PN and the risk of
poor growth at the time of hospital discharge (9). We could not de-
tect a significant reduction in poor growth or culture-proven sep-
DISCUSSION sis with early progressive feeding, but large observational studies
In this single-center randomized trial, we compared early pro- show that more aggressive enteral nutrition and less PN are asso-
gressive feeding without MEF with delayed progressive feeding ciated with a lower risk of sepsis (9, 13–15).
after a 4-d course of MEF. We showed that early progressive Our results contradict the results of a randomized trial
feeding increases the total number of full enteral feeding days, that favored MEF over progressive feeding in predominantly
reduces the use of PN, and reduces the need for central venous formula-fed infants with limited exposure to antenatal steroids
access in extremely preterm infants. We also found that early pro- (6, 16). The trial that compared MEF with progressive feeding
gressive feeding, compared with delayed progressive feeding af- listed several contraindications to initiate enteral feeding and de-
ter a 4-d course of MEF, reduces the use of PN without increasing layed initiation of enteral feeding for ∼10 d in all study partici-
the risk of postnatal growth restriction at 36 wk of postmenstrual pants (6). Because antenatal steroids increase survival and reduce
age. To our knowledge, this is the first trial of early progressive short-term complications in preterm infants (17), early initiation
feeding that includes only extremely preterm infants. of enteral feeding is recommended in the current era of antenatal
 EARLY FEEDING IN EXTREMELY PRETERM INFANTS 369
slow progressive feeding reduces the risk of NEC (19). Our trial
used a rapid progressive feeding rate by definition (>24 mL 
kg−1  d−1 ) (19), but our progressive feeding rates were lower
than the average rates reported in randomized trials of rapid pro-
gressive feeding.
If equipoise can be maintained to show that the unknown risk
of NEC outweighs the proven benefits of early progressive feed-
ing on full enteral feeding, PN use, and need for central access, a
larger multicenter trial powered to detect differences in the out-
come of NEC could add external validity to our results. The aver-
age time to establish full enteral feeding in this trial was shorter
than the average time reported in other trials that included ex-
tremely preterm infants (20). A larger neonatal trial might not re-
duce uncertainty as anticipated (21), but it could show that early
progressive feeding reduces the risk of postnatal growth restric-
tion and culture-proven sepsis.
Because enteral feeding is often initiated earlier and advanced
FIGURE 2 Percentage of infants with full enteral feeding established ac- more rapidly in less critically ill preterm infants (11), random-
cording to the intervention group. In this time-to-event analysis, the observa- ization of high-risk extremely preterm infants is one of the most
tion period began at birth and continued until postnatal day 28. The number
of infants eligible for or “at risk” of developing the outcome of interest (i.e., important strengths of this trial. Severe respiratory distress syn-
full enteral feeding) changed over time (numbers at the bottom of the graph). drome, sepsis, and hypotension were not listed as exclusion cri-
Infants removed from the study as a consequence of serious adverse events teria for this trial. Progressive feeding was initiated according
(i.e., necrotizing enterocolitis, spontaneous intestinal perforation, or death) to treatment allocation and following an individualized birth-
were censored (not counted in the denominator used to report the percentage
of infants with full enteral feeding established). Infants unable to achieve full weight–based feeding protocol in all study participants. Vari-
enteral feeding by postnatal day 28 were also censored. ability in progressive feeding practices was minimized through
daily measurements of compliance to avoid differential noncom-
steroid use. Moreover, the presence of umbilical catheters or the pliance, which often introduces bias; however, the 4-d delay in
infusion of pressor agents are no longer listed as contraindications the progression of feeding was not reflected in the final differ-
to initiate enteral feeding, particularly if human milk is available ence between groups, likely because the progression of feeding
(18). This important difference between trials suggests that, when after feeding day 1 in the early feeding group had the compliance
human milk is not available, early progressive feeding may not be challenges of any new intervention.
suitable for extremely preterm infants who develop intestinal at- In summary, this trial shows that early progressive feeding is
rophy due to delayed initiation of enteral feeding. The results of not only feasible in critically ill, extremely preterm infants but
a subgroup analysis of growth-restricted extremely preterm in- also effective in increasing the number of full enteral feeding
fants randomly assigned to receive early progressive feeding also days, reducing the use of PN, and reducing the use of central ve-
suggest that early progressive feeding might be less effective in nous access. It remains uncertain whether early progressive feed-
extremely preterm infants born SGA (10). ing increases the risk of NEC in extremely preterm infants born
Although we adequately powered this trial to test the effect at the limits of viability, but larger studies of feeding practices
of early progressive feeding on the outcome “full enteral feeding aimed at promoting early progression of feeding could reduce the
days in the first 28 d after birth,” a surrogate outcome measur- risk of postnatal growth restriction and culture-proven sepsis in
able in all study participants that quantifies the negative effects all extremely preterm infants.
of SIP, NEC, or death on enteral feeding and combines efficacy
and safety endpoints of the trial in a continuous scale, the power The authors’ responsibilities were as follows—AAS: conceptualized and
designed the study, carried out the initial analysis, and drafted the initial
of this trial was insufficient to determine the effect of early pro-
manuscript; PL: performed the randomization and carried out the statistical
gressive feeding on the outcome of NEC. Unlike the interim anal- analysis; KP: designed the data collection instruments, monitored patient en-
ysis of a large randomized trial of trophic feeding compared with rollment and compliance, and collected data; CVL, CRM, and WAC: helped
delayed progressive feeding (6, 16), neither the interim nor the fi- design the study and critically reviewed the manuscript; and all authors: read
nal analysis of this trial identified a temporal association between and approved the final manuscript. The authors had no conflicts of interest
progressive feeding and NEC. All but one of the NEC cases were relevant to this article to disclose.
attributed to severity of critical illness, and the overall risk of NEC
reported in this trial was comparable to the baseline risk of NEC
observed in extremely preterm infants admitted to our unit (8% REFERENCES
compared with 10%, respectively). 1. Salas AA, Kabani N, Travers CP, Phillips V, Ambalavanan N, Carlo
WA. Short versus extended duration of trophic feeding to reduce
Our findings are consistent with the results of other random- time to achieve full enteral feeding in extremely preterm infants: an
ized trials that compared early with delayed progressive feed- observational study. Neonatology 2017;112(3):211–6.
ing in moderate-preterm infants (8) and growth-restricted ex- 2. Klingenberg C, Embleton ND, Jacobs SE, O’Connell LA, Kuschel
tremely preterm infants (10). In addition, a recently updated CA. Enteral feeding practices in very preterm infants: an international
survey. Arch Dis Child Fetal Neonatal Ed 2012;97(1):F56–61.
meta-analysis in >500 extremely preterm infants randomly 3. Salas AA, Cuna A, Bhat R, McGwin G Jr., Carlo WA, Ambalavanan
assigned to either rapid (30 and 35 mL  kg−1  d−1 ) or slow pro- N. A randomised trial of re-feeding gastric residuals in preterm infants.
gressive (15–20 mL  kg−1  d−1 ) feeding did not suggest that Arch Dis Child Fetal Neonatal Ed 2015;100(3):F224–8.
370 SALAS ET AL.

4. Hans DM, Pylipow M, Long JD, Thureen PJ, Georgieff MK. Nutritional infants: a randomized clinical trial. J Pediatr (Rio J) 2013;89(4):
practices in the neonatal intensive care unit: analysis of a 2006 neonatal 388–93.
nutrition survey. Pediatrics 2009;123(1):51–7. 13. Stoll BJ, Hansen NI, Adams-Chapman I, Fanaroff AA, Hintz SR,
5. Henderson G, Craig S, Brocklehurst P, McGuire W. Enteral feeding Vohr B, Higgins RD, National Institute of Child Health; Human
regimens and necrotising enterocolitis in preterm infants: a multicentre Development Neonatal Research Network. Neurodevelopmental and
case-control study. Arch Dis Child Fetal Neonatal Ed 2009;94(2): growth impairment among extremely low-birth-weight infants with
F120–3. neonatal infection. JAMA 2004;292(19):2357–65.
6. Berseth CL, Bisquera JA, Paje VU. Prolonging small feeding volumes 14. Flidel-Rimon O, Friedman S, Lev E, Juster-Reicher A, Amitay
early in life decreases the incidence of necrotizing enterocolitis in very M, Shinwell ES. Early enteral feeding and nosocomial sepsis in
low birth weight infants. Pediatrics 2003;111(3):529–34. very low birthweight infants. Arch Dis Child Fetal Neonatal Ed
7. Jasani B, Patole S. Standardized feeding regimen for reducing 2004;89(4):F289–92.
necrotizing enterocolitis in preterm infants: an updated systematic 15. Hay WW. Optimizing nutrition of the preterm infant. Zhongguo Dang
review. J Perinatol 2017;37(7):827–33. Dai Er Ke Za Zhi 2017;19(1):1–21.
8. Morgan J, Young L, McGuire W. Delayed introduction of progressive 16. Engle WD, Lair CS. Early feeding of premature infants questioned.
enteral feeds to prevent necrotising enterocolitis in very low birth weight Pediatrics 2004;113(4):931–2.
infants. Cochrane Database Syst Rev 2014;12:CD001970. 17. Stoll BJ, Hansen NI, Bell EF, Walsh MC, Carlo WA, Shankaran S,
9. Leaf A, Dorling J, Kempley S, McCormick K, Mannix P, Linsell L, Laptook AR, Sanchez PJ, Van Meurs KP, Wyckoff M, et al. Trends in
Juszczak E, Brocklehurst P; Abnormal Doppler Enteral Prescription care practices, morbidity, and mortality of extremely preterm neonates,
Trial Collaborative Group. Early or delayed enteral feeding for 1993–2012. JAMA 2015;314(10):1039–51.
preterm growth-restricted infants: a randomized trial. Pediatrics 18. Quigley M, McGuire W. Formula versus donor breast milk for feeding
2012;129(5):e1260–8. preterm or low birth weight infants. Cochrane Database Syst Rev
10. Kempley S, Gupta N, Linsell L, Dorling J, McCormick K, Mannix P, 2014;4:CD002971.
Juszczak E, Brocklehurst P, Leaf A. Feeding infants below 29 weeks’ 19. Oddie SJ, Young L, McGuire W. Slow advancement of enteral feed
gestation with abnormal antenatal Doppler: analysis from a randomised volumes to prevent necrotising enterocolitis in very low birth weight
trial. Arch Dis Child Fetal Neonatal Ed 2014;99(1):F6–F11. infants. Cochrane Database Syst Rev 2017;8:CD001241.
11. Ehrenkranz RA, Das A, Wrage LA, Poindexter BB, Higgins RD, Stoll 20. Shah SD, Dereddy N, Jones TL, Dhanireddy R, Talati AJ. Early versus
BJ, Oh W; Eunice Kennedy Shriver National Institute of Child Health; delayed human milk fortification in very low birth weight infants—a
Human Development Neonatal Research Network. Early nutrition randomized controlled trial. J Pediatr 2016;174:126–31, e1.
mediates the influence of severity of illness on extremely LBW infants. 21. Hay SC, Kirpalani H, Viner C, Soll R, Dukhovny D, Mao WY, Profit
Pediatr Res 2011;69(6):522–9. J, DeMauro SB, Zupancic JAF. Do trials reduce uncertainty? Assessing
12. Arnon S, Sulam D, Konikoff F, Regev RH, Litmanovitz I, Naftali impact through cumulative meta-analysis of neonatal RCTs. J Perinatol
T. Very early feeding in stable small for gestational age preterm 2017;37(11):1215–9.