Meta-Analysis

Ann Nutr Metab Received: December 11, 2020

Accepted: April 18, 2021
DOI: 10.1159/000516640 Published online: July 9, 2021

Enteral Feeding Strategies in Preterm Neonates

≤32 weeks Gestational Age: A Systematic Review
and Network Meta-Analysis
Viraraghavan Vadakkencherry Ramaswamy a, f Tapas Bandyopadhyay b    

Javed Ahmed c Prathik Bandiya d Sanja Zivanovic a, g Charles Christoph Roehr a, e, h

       

aNewborn
Services, John Radcliffe Hospital, Oxford University Hospitals NHS Foundation Trust, Oxford, UK;
bDepartment
of Neonatology, Dr. Ram Manohar Lohia Hospital & Post Graduate Institute of Medical Education and
Research, New Delhi, India; cWomen’s Wellness and Research Centre, Hamad Medical Corporation, Doha, Qatar;
dDepartment of Neonatology, Indira Gandhi Institute of Child Health, Bengaluru, India; eMedical Sciences Division,

Nuffield Department of Population Health, National Perinatal Epidemiology Unit, University of Oxford, Oxford,
UK; fDepartment of Neonatology, Ankura Hospital for Women and Children, Hyderabad, India; gDepartment of
Paediatrics, Medical Sciences Division, University of Oxford, Oxford, UK; hUniversity of Bristol, Women and Children’s
Health Research Unit, The Children’s Southmead Hospital, Bristol, UK

Keywords
Nutrition · Necrotizing enterocolitis · Very preterm
Abstract
Introduction: Critical aspects of time of feed initiation, ad-
vancement, and volume of feed increment in preterm neo-
nates remain largely unanswered. Methods: Medline , Em-
base, CENTRAL and CINAHL were searched from inception
until 25th September 2020. Network meta-analysis with the
Bayesian approach was used. Randomized controlled trials
(RCTs) evaluating preterm neonates ≤32 weeks were includ-
ed. Feeding regimens were divided based on the following
categories: initiation day: early (<72 h), moderately early (72
h–7 days), and late (>7 days); advancement day: early (<72
h), moderately early (72 h–7 days), and late (>7 days); incre-
ment volume: small volume (SV) (<20 mL/kg/day), moderate
volume (MoV) (20–< 30 mL/kg/day), and large volume (≥30
mL/kg/day); and full enteral feeding from the first day. Six-
teen regimens were evaluated. Combined outcome of nec-
rotizing enterocolitis (NEC) stage ≥ II or mortality before dis-
charge was the primary outcome. Results: A total of 39 stud-
ies enrolled around 6,982 neonates. Early initiation (EI) with
moderately early or late advancement using MoV increment
enteral feeding regimens appeared to be most efficacious in
decreasing the risk of NEC or mortality when compared to EI
and early advancement with SV increment (risk ratio [95%
credible interval]: 0.39 [0.12, 0.95]; 0.34 [0.10, 0.86]) (GRADE–
very low). Conclusions: Early initiated, moderately early, or
late advanced with MoV increment feeding regimens might
be most appropriate in decreasing the risk of NEC stage ≥II
or mortality. In view of the certainty of evidence being very
low, adequately powered RCTs evaluating these 2 strategies
are warranted. © 2021 S. Karger AG, Basel
Introduction
Despite the major advances in neonatal care, there is
paucity of consensus on the most appropriate enteral
feeding strategy in very and extremely preterm neonates
[1]. Enteral feeding is a modifiable factor for necrotizing
enterocolitis (NEC) and late-onset neonatal sepsis (LOS),
130.209.6.61 - 8/12/2021 1:18:47 AM

karger@karger.com © 2021 S. Karger AG, Basel Correspondence to:

www.karger.com/anm Viraraghavan Vadakkencherry Ramaswamy, 19.vira @ gmail.com
Charles Christoph Roehr, ccroehr @ icloud.com
Glasgow Univ.Lib.
Downloaded by:
two important causes of morbidity and mortality in pre- Table 1. Classification of enteral feeding regimens
term neonates [2, 3]. While the fear of NEC can preclude
physicians from initiating and advancing enteral feeds EIEASV EI, <72 h; EA, <72 h; and SV increment,
<20 mL/kg/day
early in postnatal life, the possibility of acquiring LOS
might encourage some to do so. EIEAMoV EI, EA, and MoV (20–<30 mL/kg/day)
Broadly put, three decades of rigorous research on pre- EIEALV EI, EA, and LV (≥30 mL/kg/day)
term enteral feeding has galvanized the shift of practice EIMASV EI, MA (72 h–7 d), and SV increment
EIMAMoV EI, MA, and MoV increment
from enteral fasting to earlier initiation as well as ad-
EIMALV EI, MA, and LV increment
vancement of enteral feeds in preterm neonates [4–6]. EILASV EI, LA (>7 d), and SV increments
However, the safest window for initiation and advance- EILAMoV EI, LA, and MoV increment
ment of feeds along with the rates of feed increment is still MIMASV MI (72 h–7 d), MA, and SV increment
debated. Establishing successful enteral feeding and pre- MIMAMoV MI, MA, and MoV increment
MIMALV MI, MA, and LV increment
venting NEC in preterm neonates is influenced by a mul- MILASV MI, LA, and SV increment
titude of biological as well as potentially modifiable fac- MILAMoV MI, LA, and MoV increment
tors, including receipt of antenatal corticosteroids, gesta- LILASV LI (>7 d), LA, and SV increment
tional age, intrauterine growth restriction as well as LILAMoV LI, LA, and MoV increment
antenatal umbilical artery doppler flow status, the sick- ETEF Early total enteral feeding
ness profile, type of milk used, use of probiotics, and EI, early initiation; MI, moderately early initiation; LI, late ini-
broad-spectrum antibiotics [7, 8]. Due to such significant tiation; EA, early advancement; MA, moderately early advance-
heterogeneity in the demographic characteristics of pre- ment; LA, late advancement; SV, small volume; MoV, moderate
term neonates, alongside the interwoven practices deter- volume; LV, large volume.
mining the risk of NEC, there can be no single “silver bul-
let” feeding regimen tailored to all.
Multiple systematic reviews (SRs) have evaluated the
different aspects of preterm feeding in pair-wise meta- Inclusion Criteria
analyses [4–6, 9]. Enteral feeding is a multifaceted inter- Randomized controlled trials (RCTs) evaluating enteral feed-
vention with many aspects such as time of initiation/ad- ing strategies in neonates of ≤32 weeks’ gestational age were in-
cluded. Studies that had enrolled small for gestational age (SGA)
vancement of feeds, volume of advancement, mode of neonates with or without evidence of placental insufficiency were
feeding (bolus vs. continuous), etc. We hypothesize that eligible for inclusion. Quasi-randomized trials, observational stud-
evaluating the different enteral feeding strategies in a net- ies, and crossover RCTs were excluded.
work meta-analysis (NMA) might be the most appropri-
ate way to study their relative effectiveness. Henceforth, Interventions
Enteral feeding regimens were classified based on the day of
in this study, we systematically review the evidence re- feed initiation, day of feed advancement, and rate of feed advance-
lated to preterm enteral feeding and analyze the compar- ment. The different enteral feeding strategies are provided in Ta-
ative effectiveness of the various regimens in a NMA. ble 1.

Methods
This SR was registered with PROSPERO (CRD42020210760)
[10]. The reporting of this review is in accordance with the PRIS-
MA-NMA extension [11].
Outcomes
NEC stage ≥II (as per the modified Bell’s classification) or mor-
tality before hospital discharge was the primary outcome evaluated
[13]. The secondary outcomes were NEC stage II or more, feed
intolerance, time to establish full enteral feeding (defined as 120–
150 mL/kg/day of enteral feeds), incidence of blood culture-prov-
en sepsis, and mortality prior to hospital discharge.

Literature Search Risk of Bias Assessment

Electronic databases–, Medline, Embase, CENTRAL and CI-
NAHL were searched from inception until 25th September 2020,
and no time limits were applied (see online suppl. Table 1; see
www.karger.com/doi/10.1159/000516640 for all online suppl. ma-
terial). There were no language barriers. Two authors searched the
titles and abstracts in duplicate. Full texts of relevant studies were
extracted and evaluated for inclusion. Rayyan-QCRI software was
used for the literature search [12].
The risk of bias was evaluated using Cochrane Collaboration
(London, UK) risk of bias tool 1.0 by 2 authors independently. Dis-
crepancies were resolved by consulting a third author.
Data Extraction, Data Synthesis, and Quality of Evidence
Two authors extracted the data independently using a struc-
tured pro forma. NMA was performed using a Bayesian random-
effects model with vague priors using R software [14–16]. Gener-
130.209.6.61 - 8/12/2021 1:18:47 AM

2 Ann Nutr Metab Ramaswamy/Bandyopadhyay/Ahmed/

DOI: 10.1159/000516640 Bandiya/Zivanovic/Roehr
Glasgow Univ.Lib.
Downloaded by:
alized linear models with 4 chains, burn-in of 50,000 iterations,
followed by 100,000 iterations and 10,000 adaptations were used.
Network connectivity was evaluated using network plots. Gel-
man-Rubin plots and trace and density plots were inspected to
look for model convergence. Model fit was assessed by leverage
plots, total residual deviance, and deviance information criterion.
Node splitting was performed to detect inconsistency between the
direct and indirect fraction of evidence. Final estimates were re-
ported as risk ratios (RRs) (95% credible intervals [CrIs]) and
mean difference (95% CrI/confidence interval). Network esti-
mates for various comparisons were depicted using league plots
and matrix plots. Interventions were ranked using the surface un-
der the cumulative ranking curve (SUCRA) plots [17]. SUCRA is
a ranking system which ranks each intervention from 0 to 1. The
closer is the SUCRA to 1, the better is the intervention. SUCRA
values can vary for an intervention for different outcomes and
must always be interpreted along with the certainty of the evi-
dence (CoE). The GRADE approach for NMA was used to assess
the CoE [18].
Sensitivity Analysis
Sensitivity analyses were performed based on the following cri-
teria:
• Gestational age
• Presence or absence of antenatal doppler abnormality
• Countries classified based on income level
Results
A total of 2,199 title and abstracts were screened after
removing the duplicates, and 102 full texts were retrieved
for assessing their eligibility for inclusion. Thirty-nine
studies enrolling 6,982 neonates were included in the fi-
nal synthesis [8, 19–56] (Table 2). The PRISMA flow is
given in online suppl. Figure 1. Nine authors of the in-
cluded RCTs were contacted for additional information,
and 5 of them responded to our request. The mean birth
weight of the studied neonates was 1,094 g (online suppl.
Fig. 2). Twenty-three studies had enrolled neonates of
gestational age ≤29 weeks, 6 studies had evaluated SGA
neonates with evidence of placental insufficiency, and 23
trials were from high-income countries.
Risk of Bias
While nineteen studies were classified as having a low
risk of bias [8, 19–25, 27, 28, 31, 33, 36, 37, 40, 41, 45, 47,
48], 19 studies were categorized as having a variable risk
of bias for random sequence generation and allocation
concealment [26, 29, 30, 32, 34, 35, 38, 39, 42–44, 46, 49–
55]. None of the RCTs had blinded the intervention to the
care providers due to the nature of the intervention. Ten
trials had blinded the assessment of NEC [8, 19, 20, 23,
35, 36, 40, 45, 46, 55]. Two trials had a high risk of attri-
Feeding in Preterm Neonates: Network
Meta-Analysis
tion bias [44, 52]. The risk of bias summary and risk of
bias graph is given in online suppl. Figure 3.
Primary Outcome (NEC Stage ≥II or Mortality)
A total of 37 studies enrolling 6,593 preterm neonates
with an event rate of 14.5% had reported on the primary
outcome. The network plot is given in Figure 1a. The
characteristics of the network for the primary outcome as
well as other outcomes are given in online suppl. Table 2.
Early initiation (EI) with moderately early (MA) or late
advancement (LA) using moderate volume (MoV) incre-
ment enteral feeding regimens (EIMAMoV and EI-
LAMoV) appeared to be the most efficacious in decreas-
ing the risk of NEC or mortality, when compared to EI
and early advancement (EA) with small-volume incre-
ment regimen (EIEASV) which was the least effective (RR
[95% CrI]–0.39 [0.12, 0.95]; 0.34 [0.10, 0.86]) (GRADE–
very low). The aforementioned 2 regimens, namely EI-
MAMoV and EILAMoV, resulted in lesser incidence of
the primary outcome than other feeding regimens as well.
SUCRA ranked EILAMoV (SUCRA–0.85) and EIMAM-
oV (SUCRA–0.79) as the 2 most efficacious feeding regi-
mens in decreasing the risk of NEC or mortality. The SU-
CRA plots with SUCRA values for the various feeding
strategies are given Figure 1c. Also, amongst the early ini-
tiated feeding regimens which had used small-volume in-
crements, those that had advanced to nutritive feeds
moderately early or late resulted in lesser incidence of
NEC or mortality when than those that had advanced ear-
ly in postnatal life (GRADE–very low). The forest plot
with network estimates of various regimens is depicted in
Figure 1b. The league plot illustrating the network esti-
mates for the various comparisons is given in Figure 2.
The CoE (GRADE) for the various comparisons for
the primary outcome and various secondary outcomes is
given in Table 3. The pair-wise meta-analyses/direct evi-
dence for the different enteral feeding strategies is given
in Figure 3. Evaluation for inconsistency between the di-
rect and indirect evidence is given online suppl. Figure 4.
Secondary Outcomes
NEC Stage ≥ II
EI with EA of feeds using large volume (LV) incre-
ments (EIEALV) resulted in an increased risk of NEC
when compared to multiple other feeding regimens (on-
line suppl Fig. 5–8).
Mortality
Inconsistency was detected in the network for some of
the comparisons for this outcome. None of the enteral
130.209.6.61 - 8/12/2021 1:18:47 AM
Ann Nutr Metab 3
DOI: 10.1159/000516640
Glasgow Univ.Lib.
Downloaded by:
 4
Table 2. Characteristics of included studies

Author GA/BW AGA/SGA Feed initiation day Feed increment day Increment volume, mL/kg/day Interventions
group 1 group 2 group 1 group 2 group 1 group 2

Bozkurt 2020 27.6 wk SGA+AGA <48 h <48 h <72 h 72–≤7 d 20–25 20–25 EIEAMoV
Turkey 963 g EIMAMoV
Pomar 2020 31.8 wk AGA 1d 1d 4d 4d 20 30 EIMAMoV
Columbia 1,600 g EIMALV
Dorling 2019 UK 29 wk AGA+SGA 72 h–≤7 d 72 h–≤7 d 4 4 30 18 MIMALV
1,143 g MIMASV

Ann Nutr Metab

Modi 2019 31 wk AGA+SGA 2d 2d <72 h <72 h 15–<20 30–40 EIEASV
India 1,076 EIEALV
Nangia 2019 31.5 wk AGA+SGA 1d 1d <72 h <72 h ETEF 20–<30 ETEF

DOI: 10.1159/000516640
India 1,314 g EIEAMoV
Tewari 2018 29.5 wk SGA <72 h 72 h–≤7 d 72 h–≤7 d >7 d 15 15 EIMASV
India 1,027 g Doppler abnormality MILASV
Salas 2018 26 wk AGA 3d 3d 72 h–≤7 d >7 d 24–25 24–25 EIMAMoV
USA 883 g EILAMoV
Hussain 2018 29.7 wk AGA+SGA 1d 1d 72 h–≤7 d 72 h–<7 d 10 20 EIMASV
Pakistan 1,163 g EIMAMoV
Raban 2015 29 wk AGA+SGA 1d 1d <72 h <72 h 24 36 EIEAMoV
South Africa 850 g EIEALV
Jain 2015 32.7 SGA 1d 1d <72 h <72 h 20 30 EIEAMoV
India 1,117 g Doppler abnormality EIEALV
Armanian 2013 30.5 wk AGA+SGA 72 h–≤7 d 72 h–≤7 d 72 h–≤7 d >7 d 20 20 MIMAMoV
Iran 1,199 g MILAMoV
Arnon 2013 31.5 wk SGA 1d 3d <72 h 72 h–≤7 d 20–<30 20–<30 EIEAMoV
Israel 1,483 g Doppler abnormality EIMAMoV
Sanghvi 2013 31.5 wk AGA+SGA 1d 1d <72 h <72 h ETEF 20 ETEF
India 1,330 g EIEAMoV
Hasshemi 2013 32 wk N/A <72 h <72 h <72 h <72 h 20–24 30 EIEAMoV
Iran EIEALV
Karagol 2012 28.1 wk AGA+SGA <48 h <48 h 72 h–≤7 d >72 h–≤7 20 30 EIMAMoV
Turkey 967.5 g d EIMALV
Leaf 2012 31 wk SGA <48 h 72 h–≤7 d <72 h 72 h–≤7 d 12–<20 12–<20 EIEASV

Bandiya/Zivanovic/Roehr
UK 1,030 Doppler abnormality MIMASV
Abdelmaaboud 2012 32 wk SGA 2d 6d 72 h–≤7 d >7 d 20 20 EIMAMoV
Qatar 775 g Doppler abnormality MILAMoV
Perez 2011 30.2 wk AGA <48 h 5d <72 h 72 h–≤7 d 20 20 EIEAMoV

Ramaswamy/Bandyopadhyay/Ahmed/
Columbia 1,230 g MIMAMoV

Downloaded by:
Glasgow Univ.Lib.
130.209.6.61 - 8/12/2021 1:18:47 AM
 Table 2 (continued)

Author GA/BW AGA/SGA Feed initiation day Feed increment day Increment volume, mL/kg/day Interventions
group 1 group 2 group 1 group 2 group 1 group 2

Meta-Analysis
Krishnamurthy 2010 30.9 wk AGA+SGA 1d 1d <72 h <72 h 20 30 EIEAMoV
1,283 EIEALV
Karagianni 2009 31.6 wk SGA <72 h >7 d >7 d >7 d 15 15 EILASV
Greece 1,105 Doppler abnormality LILASV
Mosqueda 2008 26 wk AGA+SGA 2d >7 d >7 d >7 d 10 10 EILASV
USA 760 g LILASV
Weiler 2006 27.3 wk AGA+SGA <72 h 72 h–≤7 d 72 h–≤7 d >7 d <30 <30 EIMAMoV

Feeding in Preterm Neonates: Network

USA 971 g MILAMoV
Caple 2004 31.5 wk AGA+SGA <72 h <72 h 72 h–≤7 d 72 h–≤7 d 20 30 EIMAMoV
USA 1,468 g EIMALV
Salhotra 2004 32 wk AGA+SGA 1d 1d <72 h <72 h 15 30 EIEASV
India 1,036 g EIEALV
Berseth 2003 29 wk AGA >7 d >7 d >7 d >7 d 20 Not hiked LILAMoV
USA 1,032 g for 10 d LILASV
Pipaon 2003 30 wk AGA+SGA 2d >7 d >7 d >7 d 12–<20 12–<20 EILASV
Spain 1,309 g LILASV
Van Elburg 30 wk SGA <48 h >7 d >7 d >7 d 15–<20 15–<20 EILASV
2003 895 g LILASV
McClure 2000 28 wk AGA+SGA 3d >7 d >7 d >7 d 1 mL/h 1 mL/h EILAMoV
UK 1,020 every 8–12 h every 8–12 h LILAMoV

Ann Nutr Metab

Rayyis 1999 <34 wk AGA+SGA 72 h–≤7 d 72 h–≤7 d 72 h–≤7 d 72 h–≤7 d 15 35 MIMASV
USA <1,500 g MIMALV
Schanler 1999 28 wk AGA+SGA 4d >7d >7d >7d 20 20 MILAMoV

DOI: 10.1159/000516640
USA 1,055 g LILAMoV
Wilson 1997 27.2 wk AGA+SGA 1d 72 h–≤7 d 72 h–≤7 d >7 d 12–<20 12–<20 EIMASV
UK 930 g MILASV
Becerra 1996 29.7 wk AGA+SGA <72 h >7 d 72 h–≤7 d >7 d <20 <20 EIMASV
Chile 1,112 g LILASV
Troche 1995 27.3 wk AGA+SGA 24 h 72 h–≤7 d 72 h–≤7 d 72 h–≤7 d 15–<20 15–<20 EIMASV
USA 985 g MIMASV
Davey 1994 28.5 AGA+SGA <72 h 72 h–≤7 d 72 h–≤7 d >7 d <12 <12 EIMASV
USA 1,137.5 g MILASV
Meetze 1992 28.2 wk AGA+SGA 3d >7 d >7 d >7 d 25 25 EILAMoV
USA 940 g LILAMoV
Dunn 1988 27 wk AGA+SGA 48 h >7 d >7 d >7 d 15–<20 15–<20 EILASV
USA 950 g LILASV

5
Downloaded by:
Glasgow Univ.Lib.
130.209.6.61 - 8/12/2021 1:18:47 AM
 feeding strategies resulted in a statistically significant dif-

EI/MI/LI, early initiation (<72 h), moderately early initiation (72 h–7 d), and late initiation –>7 d. EA/MA/LA, early advancement (<72 h), moderately early advancement (72 h–7 d), and late
advancement –>7 d. SV/MoV/LV, small volume (<20 mL/kg/day), moderate volume–20–<30 mL/kg/day, and large volume –≥ 30 mL/kg/day. ETEF, early total enteral feeding; SGA, small for
Interventions
ference in mortality (online suppl. Fig. 9–12).

MILASV

MILASV

MILASV
EIEALV
EILASV
LILASV
Feed Intolerance
The network was not connected and network esti-
Increment volume, mL/kg/day

mates could not be derived. The direct evidence from

pair-wise comparisons revealed that regimens that initi-
group 2

<12
ated enteral feeds relatively early and advanced to nutri-
tive feeding early or moderately early resulted in lesser
15

10

feed intolerance than that in late initiation and LA regi-

mens. The direct evidence for these comparisons is given
group 1

in online suppl. Figure 13.

>30
15

10

Sepsis
SUCRA ranked the early total enteral feeding strategy
group 2

(SUCRA–0.89) as the best strategy in preventing sepsis

Feed increment day

>7 d

>7 d

>7 d

(online suppl. Fig. 14–17).

Duration to Full Enteral Feeds

group 1

<72 h

The network estimates did not show any statistical dif-

>7 d

>7 d

ference between the various enteral feeding strategies

(online suppl. Fig. 18). Node splitting was not possible
due to the sparseness of the network. The direct evidence
72 h–≤7 d
group 2

from pair-wise meta-analyses showed that strategies with

>7 d
Feed initiation day

7d

relatively earlier initiation, advancement, and higher vol-

ume increments resulted in a significantly decreased days
72 h–≤7 d

to full enteral feeds (online suppl. Fig. 19).

group 1

<24 h
1d

Sensitivity Analyses for Primary Outcome (NEC

Stage≥ II or Mortality)
Neonates Gestational Age ≤29 weeks
None of the feeding interventions showed any statisti-
cally significant differences in reducing the primary out-
come measure when compared to each other (online
AGA+SGA

AGA+SGA

AGA+SGA
AGA/SGA

suppl. Fig. 22–24).

Neonates Gestational Age >29 weeks

EI with moderately early advancement (MA) using
MoV increment (EIMAMoV) resulted in lesser incidence
<1,500 g
GA/BW

29.6 wk
<32 wk

1,258 g
27 wk
975 g

Fig. 1. a Network plot for the primary outcome-NEC stage ≥II or

mortality. b SUCRA plot with SUCRA values (%) for the primary
outcome of NEC stage ≥II or mortality. c Forest plot depicting the
network estimates (RR [95% CrI]) of the various interventions
with “EIEASV” as the common comparator for the primary out-
Table 2 (continued)

come of NEC stage ≥ II or mortality. SUCRA, surface under the

cumulative ranking curve; NEC, necrotizing enterocolitis; RR, risk
gestational age.
Ostertag 1986

ratio; CrI, credible interval; EI, early initiation; MI, moderately

Slagle 1988

Glass 1984

early initiation; LI, late initiation; EA, early advancement; MA,

Author

moderately early advancement; LA, late advancement; SV, small

USA

UK

volume; MoV, moderate volume; LV, large volume.

(For figure see next page.)
130.209.6.61 - 8/12/2021 1:18:47 AM

6 Ann Nutr Metab Ramaswamy/Bandyopadhyay/Ahmed/

DOI: 10.1159/000516640 Bandiya/Zivanovic/Roehr
Glasgow Univ.Lib.
Downloaded by:
 Color version available online
Network Plot

SUCRA plot with SUCRA values

EIEALV
100

EIEAMoV

EILAMoV

EILASV Treatment
75 EIEALV
EIMALV EIEAMoV
EIEASV
Probability of ranking or better (%)

EIMAMoV EILAMoV
EILASV
EIMASV EIMALV
Treatment

EIMAMoV
ETEF 50 EIMASV
ETEF
LILAMoV LILAMoV
LILASV
LILASV MILAMoV
SUCRA (%)
MILASV
EILAMoV 85
MILAMoV EIMAMoV 79 MIMALV
EILASV 77
25 ETEF 73 MIMAMoV
EIMASV 63
MILASV LILAMoV 63 MIMASV
LILASV 62
MILAMoV 61
MIMALV MILASV 56
EIMALV 50
EIEALV 29
EIEAMoV 26
MIMAMoV MIMALV 25
MIMAMoV 16
MIMASV 15
MIMASV 0 EIEASV 15

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Risk Ratio relative to EIEASV Ranking of Treatment
b (showing posterior median with 95% CrI) c (Higher rankings associated with smaller outcome values)

1
130.209.6.61 - 8/12/2021 1:18:47 AM

Feeding in Preterm Neonates: Network Ann Nutr Metab 7

Meta-Analysis DOI: 10.1159/000516640
Glasgow Univ.Lib.
Downloaded by:
 Color version available online

Fig. 2. League plot depicting the network estimates (RR [95% CrI]) of different enteral feeding strategies for the
primary outcome of NEC stage ≥II or mortality. NEC, necrotizing enterocolitis; RR, risk ratio; CrI, credible in-
terval; EI, early initiation; MI, moderately early initiation; LI, late initiation; EA, early advancement; MA, mod-
erately early advancement; LA, late advancement; SV, small volume; MoV, moderate volume; LV, large volume.
130.209.6.61 - 8/12/2021 1:18:47 AM

8 Ann Nutr Metab Ramaswamy/Bandyopadhyay/Ahmed/

DOI: 10.1159/000516640 Bandiya/Zivanovic/Roehr
Glasgow Univ.Lib.
Downloaded by:
Table 3. Certainty of evidence/GRADE of different comparisons for all the outcomes

Primary outcome – NEC or mortality

comparisons indirect evidence direct evidence network meta-analysis^
quality of evidence quality of evidence RR (95% CrI) quality of evidence

EIEALV:EIEAMoV Very low$$,* Low$$ 0.97 (0.58, 1.42) Low

EIEALV:EIEASV Very low$$,* Low$$ 0.80 (0.50, 1.23) Low
EIEALV:EILAMoV Low$$ – 2.98 (1.00, 7.03) Low
EIEALV:EILASV Very low$$,* – 2.32 (1.01, 4.87) Very low
EIEALV:EIMALV Low$$ – 1.71 (0.41, 4.65) Low
EIEALV:EIMAMoV Low$$ – 2.58 (0.92, 5.79) Low
EIEALV:EIMASV Very low$$,* – 1.84 (0.88, 3.64) Very low
EIEALV:ETEF Low$$ – 2.69 (0.70, 7.93) Low
EIEALV:LILAMoV Very low$$,* – 1.95 (0.72, 4.36) Very low
EIEALV:LILASV Very low$$,* – 1.86 (0.86, 3.76) Very low
EIEALV:MILAMoV Very low$$,* – 1.92 (0.68, 4.47) Very low
EIEALV:MILASV Very low$$,* Very low$$,* 1.66 (0.86, 3.04) Very low
EIEALV:MIMALV Low$$ – 0.96 (0.39, 2.07) Low
EIEALV:MIMAMoV Low$$ – 0.84 (0.35, 1.63) Low
EIEALV:MIMASV Low$$ – 0.82 (0.38, 1.59) Low
EIEAMoV:EIEASV Low$$ – 0.87 (0.46, 1.62) Low
EIEAMoV:EILAMoV Very low$$,* – 3.14 (1.12, 7.44) Very low
EIEAMoV:EILASV Very low$$,* – 2.51 (1.00, 5.84) Very low
EIEAMoV:EIMALV Low$$ – 1.78 (0.47, 4.73) Low
EIEAMoV:EIMAMoV Very low$$,* Moderate$ 2.70 (1.06, 6.01) Moderate
EIEAMoV:EIMASV Very low$$,* – 1.99 (0.86, 4.39) Very low
EIEAMoV:ETEF – Low$$ 2.78 (0.82, 7.84) Low
EIEAMoV:LILAMoV Very low$$,* – 2.05 (0.80, 4.64) Very low
EIEAMoV:LILASV Very low$$,* – 2.00 (0.85, 4.45) Very low
EIEAMoV:MILAMoV Very low$$,* – 2.02 (0.77, 4.69 Very low
EIEAMoV:MILASV Very low$$,* – 1.79 (0.84, 3.72) Very low
EIEAMoV:MIMALV Very low$$,* – 1.05 (0.38, 2.55) Very low
EIEAMoV:MIMAMoV Very low$$,* Low$$ 0.87 (0.42, 1.58) Low
EIEAMoV:MIMASV Low$$ – 0.89 (0.36, 1.99) Low
EIEASV:EILAMoV Very low$$,* – 3.90 (1.16, 9.74) Very low
EIEASV:EILASV Very low$$,* – 3.03 (1.18, 6.80) Very low
EIEASV:EIMALV Low$$ – 2.23 (0.48, 6.33) Low
EIEASV:EIMAMoV Very low$$,* – 3.38 (1.05, 8.21) Very low
EIEASV:EIMASV Very low$$,* – 2.40 (1.02, 5.06) Very low
EIEASV:ETEF Low$$ – 3.53 (0.81, 10.91) Low
EIEASV:LILAMoV Very low$$,* – 2.54 (0.83, 6.06) Very low
EIEASV:LILASV Very low$$,* – 2.42 (1.00, 5.26) Very low
EIEASV:MILAMoV Very low$$,* – 2.51 (0.78, 6.28) Very low
EIEASV:MILASV Very low$$,* – 2.16 (0.98, 4.34) Very low
EIEASV:MIMALV Low$$ – 1.20 (0.56, 2.36) Low
EIEASV:MIMAMoV Very low$$,* – 1.09 (0.39, 2.36) Very low
EIEASV:MIMASV Very low$$,* Low$$ 1.02 (0.54, 1.80) Low
EILAMoV:EILASV Very low$$,* – 0.94 (0.29, 2.44) Very low
EILAMoV:EIMALV Low$$ – 0.75 (0.24, 1.87) Low
EILAMoV:EIMAMoV Very low$$,* Low$$ 0.96 (0.37, 1.99) Low
EILAMoV:EIMASV Very low$$,* – 0.75 (0.24, 1.87) Very low
EILAMoV:ETEF Low$$ – 1.12 (0.20, 3.89) Low
EILAMoV:LILAMoV Very low$$,* Moderate$ 0.70 (0.34, 1.27) Moderate
EILAMoV:LILASV Very low$$,* – 0.75 (0.25, 1.83) Very low
EILAMoV:MILAMoV Low$$ – 0.72 (0.27, 1.59) Low
EILAMoV:MILASV Very low$$,* – 0.68 (0.22, 1.68) Very low
130.209.6.61 - 8/12/2021 1:18:47 AM

Feeding in Preterm Neonates: Network Ann Nutr Metab 9

Meta-Analysis DOI: 10.1159/000516640
Glasgow Univ.Lib.
Downloaded by:
Table 3 (continued)

Primary outcome – NEC or mortality

comparisons indirect evidence direct evidence network meta-analysis^
quality of evidence quality of evidence RR (95% CrI) quality of evidence

EILAMoV:MIMALV Very low$$,* – 0.41 (0.10, 1.20) Very low

EILAMoV:MIMAMoV Low$$ – 0.34 (0.10, 0.81) Low
EILAMoV:MIMASV Very low$$,* – 0.35 (0.09, 0.96) Very low
EILASV:EIMALV Very low$$,* – 0.83 (0.16, 2.41) Very low
EILASV:EIMAMoV Very low$$,* – 1.25 (0.34, 3.08) Very low
EILASV:EIMASV Very low$$,* – 0.84 (0.45, 1.43) Very low
EILASV:ETEF Very low$$,* – 1.35 (0.25, 4.36) Very low
EILASV:LILAMoV Very low$$,* – 0.92 (0.30, 2.05) Very low
EILASV:LILASV Very low$$,* Low$$ 0.83 (0.51, 1.25) Low
EILASV:MILAMoV Very low$$,* – 0.92 (0.27, 2.21) Very low
EILASV:MILASV Very low$$,* Very low$$,* 0.76 (0.42, 1.24) Very low
EILASV:MIMALV Very low$$,* – 0.47 (0.14, 1.18) Very low
EILASV:MIMAMoV Very low$$,* – 0.41 (0.12, 0.95) Very low
EILASV:MIMASV Very low$$,* – 0.41(0.13, 0.95) Very low
EIMALV:EIMAMoV – Low$$ 1.77 (0.77, 3.67) Low
EIMALV:EIMASV Very low$$,* – 1.54 (0.35, 4.84) Very low
EIMALV:ETEF Low$$ – 2.24 (0.33, 8.40) Low
EIMALV:LILAMoV Very low$$,* – 1.50 (0.39, 4.21) Very low
EIMALV:LILASV Very low$$,* – 1.54 (0.35, 4.85) Very low
EIMALV:MILAMoV Very low$$,* – 1.51 (0.35, 4.56) Very low
EIMALV:MILASV Very low$$,* – 1.40 (0.32, 4.34) Very low
EIMALV:MIMALV Very low$$,* – 0.83 (0.15, 2.83) Very low
EIMALV:MIMAMoV Low$$ – 0.67 (0.16, 1.95) Low
EIMALV:MIMASV Very low$$,* – 0.71 (0.14, 2.29) Very low
EIMAMoV:EIMASV Very low$$,* Very low$$,* 0.86 (0.27, 2.16) Very low
EIMAMoV:ETEF Low$$ – 1.26 (0.24, 4.19) Low
EIMAMoV:LILAMoV Very low$$,* – 0.85 (0.31, 1.93) Very low
EIMAMoV:LILASV Very low$$,* – 0.86 (0.27, 2.17) Very low
EIMAMoV:MILAMoV Very low$$,* Very low$$,* 0.85 (0.27, 2.15) Very low
EIMAMoV:MILASV Very low$$,* – 0.78 (0.26, 1.93) Very low
EIMAMoV:MIMALV Very low$$,* – 0.47 (0.11, 1.34) Very low
EIMAMoV:MIMAMoV Low$$ – 0.38 (0.12, 0.92) Low
EIMAMoV:MIMASV Very low$$,* – 0.40 (0.10, 1.07) Very low
EIMASV:ETEF Very low$$,* – 1.65 (0.33, 5.20) Very low
EIMASV:LILAMoV Very low$$,* – 1.13 (0.39, 2.50) Very low
EIMASV:LILASV Very low$$,* Very low$$,* 1.03 (0.62, 1.64) Very low
EIMASV:MILAMoV Very low$$,* – 1.13 (0.35, 2.69) Very low
EIMASV:MILASV Very low$$,* Low$$ 0.94 (0.58, 1.41) Low
EIMASV:MIMALV Very low$$,* – 0.58 (0.18, 1.36) Very low
EIMASV:MIMAMoV Very low$$,* – 0.51 (0.16, 1.13) Very low
EIMASV:MIMASV Very low$$,* Very low$$,* 0.50 (0.17, 1.09) Very low
ETEF:LILAMoV Low$$ – 1.02 (0.18, 3.15) Low
ETEF:LILASV Very low$$,* – 1.00 (0.19, 3.10) Very low
ETEF:MILAMoV Very low$$,* – 1.00 (0.17, 3.10) Very low
ETEF:MILASV Very low$$,* – 0.89 (0.18, 2.67) Very low
ETEF:MIMALV Very low$$,* – 0.52 (0.09, 1.70) Very low
ETEF:MIMAMoV Low$$ – 0.43 (0.09, 1.18) Low
ETEF:MIMASV Low$$ – 0.44 (0.08, 1.37) Low
LILAMoV:LILASV Very low$$,* Very low$$,* 1.09 (0.42, 2.44) Very low
LILAMoV:MILAMoV Very low$$,* Low$$ 1.03 (0.50, 1.93) Low
LILAMoV:MILASV Very low$$,* – 1.00 (0.37, 2.28) Very low
LILAMoV:MIMALV Very low$$,* – 0.60 (0.15, 1.69) Very low
130.209.6.61 - 8/12/2021 1:18:47 AM

10 Ann Nutr Metab Ramaswamy/Bandyopadhyay/Ahmed/

DOI: 10.1159/000516640 Bandiya/Zivanovic/Roehr
Glasgow Univ.Lib.
Downloaded by:
Table 3 (continued)

Primary outcome – NEC or mortality

comparisons indirect evidence direct evidence network meta-analysis^
quality of evidence quality of evidence RR (95% CrI) quality of evidence

LILAMoV:MIMAMoV Very low$$,* – 0.49 (0.17, 1.09) Very low

LILAMoV:MIMASV Very low$$,* – 0.52 (0.14, 1.36) Very low
LILASV:MILAMoV Very low$$,* – 1.12 (0.36, 2.62) Very low
LILASV:MILASV Very low$$,* Very low$$,* 0.94 (0.54, 1.51) Very low
LILASV:MIMALV Very low$$,* – 0.58 (0.17, 1.41) Very low
LILASV:MIMAMoV Very low$$,* – 0.50 (0.16, 1.12) Very low
LILASV:MIMASV Very low$$,* – 0.50 (0.16, 1.13) Very low
MILAMoV:MILASV Very low$$,* – 1.04 (0.35, 2.51) Very low
MILAMoV:MIMALV Very low$$,* – 0.63 (0.15, 1.77) Very low
MILAMoV:MIMAMoV Very low$$,* Very low$$,* 0.50 (0.18, 1.02) Very low
MILAMoV:MIMASV Very low$$,* – 0.54 (0.14, 1.43) Very low
MILASV:MIMALV Very low$$,* – 0.63 (0.21, 1.46) Very low
MILASV:MIMAMoV Very low$$,* – 0.55 (0.18, 1.18) Very low
MILASV:MIMASV Very low$$,* – 0.54 (0.20, 1.16) Very low
MIMALV:MIMAMoV Very low$$,* – 1.04 (0.27, 2.61) Very low
MIMALV:MIMASV – Moderate$ 0.89 (0.57, 1.29) Moderate
MIMAMoV:MIMASV Very low$$,* – 1.15 (0.36, 2.99) Very low

Secondary outcome – sepsis

comparisons indirect evidence direct evidence network meta-analysis
quality of evidence quality of evidence RR (95% CrI) quality of evidence

EIEAMoV:MILASV Low*,$ – 0.24 (0.03, 0.89) Low

EIEAMoV:MIMASV Low$$ – 0.22 (0.03, 0.92) Low
EIEASV:ETEF Low$$ – 10.89 (1.27, 180.47) Low
EILAMoV:ETEF Low$$ – 17.57 (1.16, 466.33) Low
EILAMoV:MILAMoV Low*,$ – 17.43 (1.12, 390.96) Low
EIMASV:ETEF Low*,$ – 9.58 (1.18, 183.86) Low
EIMASV:MILAMoV Low*,$ – 9.46 (1, 167.85) Low
ETEF:MILASV Low$$ – 0.06 (0, 0.54) Low
ETEF:MIMALV Low$$ – 0.06 (0, 0.82) Low
ETEF:MIMASV Low$$ – 0.06 (0, 0.54) Low
MILAMoV:MILASV Low$$ – 0.07 (0, 0.67) Low
MILAMoV:MIMASV Low*,$ – 0.06 (0, 0.7) Low

Secondary outcome – feed intolerance (only direct evidence)

comparisons indirect evidence direct evidence
quality of evidence quality of evidence RR (95% CI)

ETEF:EIEAMoV – Moderate$ 0.54 (0.32, 0.91)

MIMAMoV:MILAMoV – Low*,$ 1.99 (1.06, 3.72)
EIMASV:MILASV – Moderate$ 0.45(0.27, 0.74)

GRADE ranking of the certainty of evidence : High – very confident that the true effect lies close to that of the estimate of the effect.
Moderate – moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a
possibility that it is substantially different. Low – confidence in the effect estimate is limited: the true effect may be substantially different
from the estimate of the effect. Very low – very little confidence in the effect estimate: the true effect is likely to be substantially different
from the estimate of effect. EI/MI/LI, early initiation (<72 h), moderately early initiation (72 h–7 d), and late initiation –>7 d; EA/MA/
LA, early advancement (<72 h), moderately early advancement (72 h–7 d), and late advancement –>7 d; SV/MoV/LV, small volume
(<20 mL/kg/day), moderate volume–20–<30 mL/kg/day, and large volume –≥ 30 mL/kg/day; RR, risk ratio; CI, confidence interval; CrI,
credible interval. *Limitations (risk of bias). $Imprecision. $$Severe imprecision. ^Values in bold are statistically significant.
130.209.6.61 - 8/12/2021 1:18:47 AM

Feeding in Preterm Neonates: Network Ann Nutr Metab 11

Meta-Analysis DOI: 10.1159/000516640
Glasgow Univ.Lib.
Downloaded by:
Fig. 3. Direct evidence from the pair-wise
comparisons for the primary outcome
NEC stage ≥II or mortality. NEC, necrotiz-
ing enterocolitis; EI, early initiation; MI,
moderately early initiation; LI, late initia-
tion; EA, early advancement; MA, moder-
ately early advancement; LA, late advance-
ment; SV, small volume; MoV, moderate
volume; LV, large volume.
(Figure continued on next page.)
130.209.6.61 - 8/12/2021 1:18:47 AM

12 Ann Nutr Metab Ramaswamy/Bandyopadhyay/Ahmed/

DOI: 10.1159/000516640 Bandiya/Zivanovic/Roehr
Glasgow Univ.Lib.
Downloaded by:
 3
130.209.6.61 - 8/12/2021 1:18:47 AM

Feeding in Preterm Neonates: Network Ann Nutr Metab 13

Meta-Analysis DOI: 10.1159/000516640
Glasgow Univ.Lib.
Downloaded by:
of NEC or mortality than that in EI with MA using LV
increment (EIMALV) (RR [95% CrI]–0.24 [0.01, 0.89])
(GRADE–low) (online suppl. Fig. 25).
Neonates with No Antenatal Doppler Abnormalities
Similar to the primary analysis, there was a trend to-
ward EI with moderately early or LA using MoV incre-
ment enteral feeding regimens being more efficacious in
decreasing the primary outcome measure (online suppl.
Fig. 26–27).
Neonates Who Were SGA with Antenatal Doppler
Abnormalities
Five of the 6 trials had reported on the primary out-
come. Direct evidence from pair-wise meta-analysis did
not show any statistically significant differences in the
primary outcome measure between the different inter-
ventions (online suppl. Fig. 28).
Studies from High-Income Countries
There were no statistically significant differences be-
tween any of the enteral feeding strategies in decreasing
the risk of primary outcome measure when studies from
high-income countries were evaluated (online suppl. Fig.
29–30).
Studies from LMICs
The network was sparse and only direct evidence was
evaluated. Moderate CoE showed that EI, MA with MoV
increment (EIMAMoV) decreased the incidence of NEC
or mortality when compared to a similar strategy but with
earlier advancement to nutritive feeds (EIEAMoV) (RR
[95% confidence interval]: 0.25 [0.07–0.87]) (online sup-
pl. Fig. 31).
Discussion
In this SR, we investigated different enteral feeding
strategies for very preterm neonates with the primary
outcome evaluated being NEC ≥ stage II or mortality be-
fore discharge. Sixteen enteral feeding strategies were an-
alyzed, synthesizing data from 39 trials enrolling 6,982
neonates in a NMA.
While early initiated (within 72 h), moderately early
(72 h–7 days), or late advanced (>7 days) and MoV incre-
ment (20–< 30 mL/kg/day) strategies, EIMAMoV and EI-
LAMoV, were the 2 most effective strategies to decrease
the incidence of NEC or mortality, the early initiated, ear-
ly advanced (<72 h), and small-volume advancement reg-
14 Ann Nutr Metab
DOI: 10.1159/000516640
imen (<20 mL/kg/day) EIEASV was the least effective.
While our analysis revealed early initiated feeding strate-
gies to be better than the later ones, Morgan et al. [6] in
their SR and pair-wise meta-analysis of early trophic
feeds (introduced before 96 h of age and continued for at
least until 1 week after birth) versus enteral fasting did not
find any difference in the incidence of NEC or mortality.
This could be explained by the inclusion of recent studies
in our review.
In a subsequent SR by Morgan et al. [5] comparing
early introduction (day 1–day 4) versus delayed introduc-
tion of progressive feeds (>day 4, no upper limit defined),
no difference was found between the 2 strategies in pre-
venting NEC or mortality, which was contrary to our
findings. The discrepancy between our results and that of
Morgan et al. might be explained by the differences in the
way interventions were classified. While our SR classified
the progressive advancement period into 3 time periods,
Morgan et al. had done so into two. Our results suggest
that MoV advancement strategies might be superior to
both small and LV advancement ones. The Cochrane re-
view by Oddie et al. [9] revealed no differences between
small (≤24 mL/kg/day) and LV increments (>24 mL/kg/
day). Our SR classified the feeding regimens in a multi-
faceted manner with 2 other aspects of enteral feeding
namely, day of initiation and advancement of feeds be-
sides volume of increment.
Both EIMAMoV and EILAMoV decreased the rates of
NEC or mortality when compared to EIEAMoV. Also,
the other early initiated, EA regimen with LV increment,
EIEALV, resulted in higher incidence of NEC when com-
pared to multiple others. There is a possibility that faster
advancement of enteral feeds in early postnatal life might
increase the metabolic demand on the gut as well as bac-
terial overgrowth subsequent to undigested feeds, in-
creasing the risk of NEC [57, 58]. The 2 late initiated and
late advanced regimens LILAMoV and LILASV decreased
the risk of NEC without decreasing the rates of the com-
bined outcome of NEC or mortality. There is a possibility
of this being due to an increased risk of mortality with
these 2 regimens.
Our review had pragmatic inclusion criteria, evaluat-
ing neonates with differing sickness profile, treated in
NICUs with varying resources and survival rates. We
tried to address this through sensitivity analyses. The re-
sults were similar to the primary analysis when neonates
with no antenatal doppler abnormalities were evaluated
separately. However, none of the feeding regimens result-
ed in better outcomes in neonates of ≤29 weeks, those
with antenatal doppler abnormalities, and those from
130.209.6.61 - 8/12/2021 1:18:47 AM
Ramaswamy/Bandyopadhyay/Ahmed/
Bandiya/Zivanovic/Roehr
Glasgow Univ.Lib.
Downloaded by:
high-income nations. Our sensitivity analyses revealed
EIMAMoV to be better than EIMALV in neonates of >29
weeks and than EIEAMoV in neonates from LMIC.
Analysis of the secondary outcomes indicates that feed-
ing regimens with relatively earlier initiation, earlier ad-
vancement, and larger volume increments were better in
decreasing the risk of feed intolerance, the duration to full
feeds, and sepsis. One novel feeding strategy, ETEF, which
was evaluated by 2 trials in neonates of birth weight 1,000–
1,500 g, appeared most effective in preventing LOS. Such a
regimen that can lower the incidence of LOS might have a
huge impact in LMICs, where the burden of LOS with sig-
nificant antimicrobial resistance and sepsis-related mortal-
ity is a cause of major concern [59]. A large multicentric
RCT (FEED1 trial) evaluating ETEF strategy is ongoing [1].
Our SR has several limitations. To capture a wide variety
of trials and interventions, we used broad inclusion criteria.
Inconsistency was detected in some of the networks assess-
ing the secondary outcomes. The network was sparse for
some other secondary outcomes, for which our analysis and
interpretation relied on direct fraction of the evidence from
pair-wise meta-analysis. Finally, we had enrolled studies
spanning 3 decades during which certain other proposed
NEC prevention strategies, such as exclusive human milk
diet and probiotics, were not a standard of care.
Acknowledgements
We acknowledge the contributions of Prof. Sushma Nangia
and Dr. Kenny McCormick in giving intellectual inputs during the
writing of the manuscript. We also thank Abdul Kareem Pullat-
tayil S. for devising the literature search strategy for all the data-
bases and Dr. Roseanna Harrison and Dr. Debasish Nanda for as-
sisting in proofreading of the final version of the manuscript. We
thank Prof. David C. Wilson, Dr. Pamela Cairns, Prof. R.M. van
Elburg, Associate Prof. Shmuel Arnon, and Dr. M. Gharehbaghi
for providing additional information related to their studies. We
are also grateful to Dr. Roya Taheritafti for translating a study from
Persian to English.
Statement of Ethics
The authors have no ethical conflicts to disclose.
Conflict of Interest Statement
The authors have no conflicts of interest to declare.
Funding Sources
The authors did not receive any funding.

Conclusions Author Contributions

Dr. Viraraghavan Vadakkencherry Ramaswamy, Dr. Tapas

In preterm neonates of ≤32 weeks gestational age:
• Early initiated (within 72 h), moderately early (72 h–7
days), or late advanced (>7 days) with MoV increment
(20–< 30 mL/kg/day) enteral feeding regimens (EI-
MAMoV and EILAMoV) were the 2 most efficacious
regimens in reducing the risk of NEC stage ≥II or mor-
tality (CoE–very low).
• The CoE being very low for the primary outcome, we
suggest adequately powered, multicentric RCTs evalu-
ating these 2 regimens in future.
Bandyopadhyay, and Dr. Prathik Bandiya conceptualized the SR
and did the data analysis. Dr. Javed Ahmed and Dr. Tapas Bandyo-
padhyay were responsible for literature search and data extraction.
Dr. Sanja Zivanovic provided the first draft of the manuscript.
Prof. Charles Christoph Roehr gave further intellectual inputs and
revised the initial draft. All the authors approved the final version
for submission and agree to be accountable for all aspects of the
work.

References
 1 Kwok TC, Dorling J, Gale C. Early enteral  
feeding in preterm infants. Semin Perinatol.
2019;43(7):151159.
  2 Dong Y, Speer CP. Late-onset neonatal sepsis:  
recent developments. Arch Dis Child Fetal
Neonatal Ed. 2015;100(3):F257–63.
 3 Ramani M, Ambalavanan N. Feeding prac-
tices and necrotizing enterocolitis. Clin Peri-
natol. 2013;40(1):1–10.
4 Bombell S, McGuire W. Early trophic feeding  
for very low birth weight infants. Cochrane
Database Syst Rev. 2009;3:CD000504.
5 Morgan J, Young L, McGuire W. Delayed in-
troduction of progressive enteral feeds to pre-  
vent necrotising enterocolitis in very low
birth weight infants. Cochrane Database Syst
Rev. 2014;2014(12):CD001970.
6 Morgan J, Bombell S, McGuire W. Early tro-
phic feeding versus enteral fasting for very
preterm or very low birth weight infants. Co-
chrane Database Syst Rev. 2013;3:CD000504.
7 Gephart SM, Hanson C, Wetzel CM, Fleiner
M, Umberger E, Martin L, et al. NEC-zero
recommendations from scoping review of ev-
idence to prevent and foster timely recogni-
tion of necrotizing enterocolitis. Matern
Health Neonatol Perinatol. 2017;3:23.
130.209.6.61 - 8/12/2021 1:18:47 AM

Feeding in Preterm Neonates: Network Ann Nutr Metab 15

Meta-Analysis DOI: 10.1159/000516640
Glasgow Univ.Lib.
Downloaded by:
  8 Leaf A, Dorling J, Kempley S, McCormick K,
Mannix P, Linsell L, et al. Early or delayed en-
teral feeding for preterm growth-restricted
infants: a randomized trial. Pediatrics. 2012;
129(5):e1260–8.
 9 Oddie SJ, Young L, McGuire W. Slow ad-
vancement of enteral feed volumes to prevent
necrotising enterocolitis in very low birth
weight infants. Cochrane Database Syst Rev.
2017;8(8):CD001241.
10 Ramaswamy VV, Bandyopadhyay T, Bandiya
P, Nanda D, Ahmed J, Pandita A, et al. Com-
parative effectiveness of different enteral
feeding strategies in preterm neonates </= 32
weeks’ gestational age: a systematic review
and network meta-analysis. PROSPERO 2020
CRD42020210760. Available from: https: //
www.crd.york.ac.uk/prospero/display_re-
cord.php?ID=CRD42020210760 Accessed
2020 Oct 23.
11 Hutton B, Salanti G, Caldwell DM, Chaimani
A, Schmid CH, Cameron C, et al. The PRIS-
MA extension statement for reporting of sys-
tematic reviews incorporating network meta-
analyses of health care interventions: check-
list and explanations. Ann Intern Med. 2015;
162(11):777–84.
12 Ouzzani M, Hammady H, Fedorowicz Z, El-
magarmid A. Rayyan-a web and mobile app
for systematic reviews. Syst Rev. 2016;5:210.
13 Walsh MC, Kliegman RM. Necrotizing en-
terocolitis: treatment based on staging crite-
ria. Pediatr Clin North Am. 1986;33:179–201.
14 R Core Team. R: a language and environment
for statistical computing. Vienna, Austria: R
Foundation for Statistical Computing; 2017.
Available from: https: //www.R-project.org/
Accessed 2020 Oct 23.
15 van Valkenhoef G, Lu G, de Brock B, Hillege
H, Ades AE, Welton NJ. Automating network
meta-analysis. Res Synth Methods. 2012;3(4):
285–99.2012
16 Béliveau A, Boyne DJ, Slater J, Brenner D,
Arora P. BUGSnet: an R package to facilitate
the conduct and reporting of Bayesian net-
work meta-analyses. BMC Med Res Method-
ol. 2019;19(1):196.
17 Salanti G, Ades AE, Ioannidis JP. Graphical
methods and numerical summaries for pre-
senting results from multiple-treatment me-
ta-analysis: an overview and tutorial. J Clin
Epidemiol. 2011;64(2):163–71.
18 Puhan MA, Schünemann HJ, Murad MH, Li
T, Brignardello-Petersen R, Singh JA, et al. A
GRADE Working Group approach for rating
the quality of treatment effect estimates from
network meta-analysis. BMJ. 2014;349:g5630.
19 Bozkurt O, Alyamac Dizdar E, Bidev D, Sari
FN, Uras N, Oguz SS. Prolonged minimal en-
teral nutrition versus early feeding advance-
ments in preterm infants with birth weight
≤1250 g: a prospective randomized trial. J Ma-
tern Fetal Neonatal Med. 2020:1–7.
16 Ann Nutr Metab
DOI: 10.1159/000516640
20 Montealegre-Pomar ADP, Bertolotto-Cepeda
AM, Romero-Marquez Y, Muñoz-Ramírez
KJ. Effectiveness and safety of fast enteral ad-
vancement in preterm infants between 1000
and 2000 g of birth weight. J Parenter Enteral
Nutr. 2020.
21 Dorling J, Abbott J, Berrington J, Bosiak B,
Bowler U, Boyle E, et al. Controlled trial of
two incremental milk-feeding rates in pre-
term infants. N Engl J Med. 2019; 381(15):
1434–43.
22 Modi M, Ramji S, Jain A, Kumar P, Gupta N.
Early aggressive enteral feeding in neonates
weighing 750–1250 grams: a randomized
controlled trial. Indian Pediatr. 2019; 56(4):
294–8.
23 Nangia S, Vadivel V, Thukral A, Saili A. Early
total enteral feeding versus conventional en-
teral feeding in stable very-low-birth-weight
infants: a randomised controlled trial. Neona-
tology. 2019;115(3):256–62.
24 Tewari VV, Dubey SK, Kumar R, Vardhan S,
Sreedhar CM, Gupta G. Early versus late en-
teral feeding in preterm intrauterine growth
restricted neonates with antenatal doppler ab-
normalities: an open-label randomized trial. J
Trop Pediatr. 2018;64(1):4–14.
25 Salas AA, Li P, Parks K, Lal CV, Martin CR,
Carlo WA. Early progressive feeding in ex-
tremely preterm infants: a randomized trial.
Am J Clin Nutr. 2018;107(3):365–70.
26 Hussain A, Rehman A, Fatima N. Compari-
son of volume and frequency advancement
feeding protocols in very low birth weight ne-
onates. Pak J Med Sci. 2018;34(1):78–81.
27 Raban S, Santhakumaran S, Keraan Q, Joolay
Y, Uthaya S, Horn A, et al. A randomised con-
trolled trial of high vs low volume initiation
and rapid vs slow advancement of milk feeds
in infants with birthweights ≤1,000 g in a re-
source-limited setting. Paediatr Int Child
Health. 2016;36(4):288–95.
28 Jain S, Mukhopadhyay K, Jain V, Kumar P.
Slow versus rapid enteral feed in preterm neo-
nates with antenatal absent end diastolic flow.
J Matern Fetal Neonatal Med. 2016; 29(17):
2828–33.
29 Arnon S, Sulam D, Konikoff F, Regev RH, Lit-
manovitz I, Naftali T. Very early feeding in
stable small for gestational age preterm in-
fants: a randomized clinical trial. J Pediatr.
2013;89(4):388–93.
30 Armanian A-M, Kazemipour S, Mirbod S-M,
Hassanzade A. Comparison of prolonged low
volume milk and routine volume milk on in-
cidence of necrotizing enterocolitis in very
low birth weight neonates. Pak J Med Sci.
2013;29(1 Suppl):312–6.
31 Sanghvi KP, Joshi P, Nabi F, Kabra N. Feasi-
bility of exclusive enteral feeds from birth in
VLBW infants >1200 g: an RCT. Acta Paedi-
atr. 2013;102(7):e299–304.
32 Hasshemi B, Gharehbaghi M, Ghojazadeh M,
Sanaie G. Comparing outcomes of different
feeding volumes in preterm infants. J Urmia
Univ Med Sci. 2013;24(1):58–64.
33 Karagol BS, Zenciroglu A, Okumus N, Polin
RA. Randomized controlled trial of slow vs.
rapid enteral feeding advancements on the
clinical outcomes of preterm infants with
birth weight 750–1250 g. JPEN J Parenter En-
teral Nutr. 2013;37(2):223–8.
34 Abdelmaaboud M, Mohammed A. Early ver-
sus late minimal enteral feeding in weeks pre-
term growth-restricted neonates with abnor-
mal antenatal doppler studies. J Matern Fetal
Neonatal Med. 2012.
35 Pérez LA, Pradilla GL, Díaz G, Bayter SM.
[Necrotizing enterocolitis among preterm
newborns with early feeding]. Biomedica.
2011;31(4):485–91.
36 Krishnamurthy S, Gupta P, Debnath S,
Gomber S. Slow versus rapid enteral feeding
advancement in preterm newborn infants
1,000–1,499 g: a randomized controlled trial.
Acta Paediatr. 2010;99(1):42–6.
37 Karagianni P, Briana DD, Mitsiakos G, Elias
A, Theodoridis T, Chatziioannidis E, et al.
Early versus delayed minimal enteral feeding
and risk for necrotizing enterocolitis in pre-
term growth-restricted infants with abnormal
antenatal Doppler results. Am J Perinatol.
2010;27(5):367–73.
38 Mosqueda E, Sapiegiene L, Glynn L, Wilson-
Costello D, Weiss M. The early use of minimal
enteral nutrition in extremely low birth
weight newborns. J Perinatol. 2008; 28(4):
264–9.
39 Weiler HA, Fitzpatrick-Wong SC, Schellen-
berg JM, Fair DE, McCloy UR, Veitch RR, et
al. Minimal enteral feeding within 3 d of birth
in prematurely born infants with birth weight.
Am J Clin Nutr. 2006;83(1):155–62.
40 Caple J, Armentrout D, Huseby V, Halbardier
B, Garcia J, Sparks JW, et al. Randomized,
controlled trial of slow versus rapid feeding
volume advancement in preterm infants. Pe-
diatrics. 2004;114(6):1597–600.
41 Salhotra A, Ramji S. Slow versus fast enteral
feed advancement in very low birth weight in-
fants: a randomized control trial. Indian Pe-
diatr. 2004;41(5):435–41.
42 Berseth CL, Bisquera JA, Paje VU. Prolonging
small feeding volumes early in life decreases
the incidence of necrotizing enterocolitis in
very low birth weight infants. Pediatrics.
2003;111(3):529–34.
43 Sáenz de Pipaón M, VanBeek RH, Quero J,
Pérez J, Wattimena DJ, Sauer PJ. Effect of
minimal enteral feeding on splanchnic uptake
of leucine in the postabsorptive state in pre-
term infants. Pediatr Res. 2003;53(2):281–7.
44 van Elburg RM, van den Berg A, Bunkers CM,
van Lingen RA, Smink EW, van Eyck J, et al.
Minimal enteral feeding, fetal blood flow pul-
satility, and postnatal intestinal permeability
in preterm infants with intrauterine growth
retardation. Arch Dis Child Fetal Neonatal
Ed. 2004;89:F293–6.
130.209.6.61 - 8/12/2021 1:18:47 AM
Ramaswamy/Bandyopadhyay/Ahmed/
Bandiya/Zivanovic/Roehr
Glasgow Univ.Lib.
Downloaded by:
45 McClure RJ, Newell SJ. Randomised con-
trolled study of clinical outcome following
trophic feeding. Arch Dis Child Fetal Neona-
tal Ed. 2000;82(1):F29–33.
46 Rayyis SF, Ambalavanan N, Wright L, Carlo
WA. Randomized trial of “slow” versus “fast”
feed advancements on the incidence of necro-
tizing enterocolitis in very low birth weight
infants. J Pediatr. 1999;134(3):293–7.
47 Schanler RJ, Shulman RJ, Lau C, Smith EO,
Heitkemper MM. Feeding strategies for pre-
mature infants: randomized trial of gastroin-
testinal priming and tube-feeding method.
Pediatrics. 1999;103(2):434–9.
48 Wilson DC, Cairns P, Halliday HL, Reid M,
McClure G, Dodge JA. Randomised con-
trolled trial of an aggressive nutritional regi-
men in sick very low birthweight infants.
Arch Dis Child Fetal Neonatal Ed. 1997;77(1):
F4–11.
Feeding in Preterm Neonates: Network
Meta-Analysis
49 Becerra M, Ambiado S, Kuntsman G, Figueroa
A, Balboa P, Fernandez P, et al. Feeding
VLBW infants: effect of early enteral stimula-
tion (EES) [abstract]. Pediatric Research.
1996;39:304A.
50 Troche B, Harvey-Wilkes K, Engle WD,
Nielsen HC, Frantz ID, Mitchell ML, et al.
Early minimal feedings promote growth in
critically ill premature infants. Biol Neonate.
1995;67(3):172–81.
51 Davey AM, Wagner CL, Cox C, Kendig JW.
Feeding premature infants while low umbili-
cal artery catheters are in place: a prospective,
randomized trial. J Pediatr. 1994;124(5 Pt 1):
795–9.
52 Meetze WH, Valentine C, McGuigan JE, Con-
lon M, Sacks N, Neu J. Gastrointestinal prim-
ing prior to full enteral nutrition in very low
birth weight infants. J Pediatr Gastroenterol
Nutr. 1992;15(2):163–70.
53 Dunn L, Hulman S, Weiner J, Kliegman R.
Beneficial effects of early hypocaloric enteral
feeding on neonatal gastrointestinal function:
preliminary report of a randomized trial. J Pe-
diatr. 1988;112(4):622–9.
Ann Nutr Metab
DOI: 10.1159/000516640
54 Slagle TA, Gross SJ. Effect of early low-vol-
ume enteral substrate on subsequent feeding
tolerance in very low birth weight infants. J
Pediatr. 1988;113(3):526–31.
55 Ostertag SG, LaGamma EF, Reisen CE, Fer-
rentino FL. Early enteral feeding does not af-
fect the incidence of necrotizing enterocolitis.
Pediatrics. 1986;77(3):275–80.
56 Glass EJ, Hume R, Lang MA, Forfar JO. Par-
enteral nutrition compared with transpyloric
feeding. Arch Dis Child. 1984;59(2):131–5.
57 Berman L, Moss RL. Necrotizing enterocoli-
tis: an update. Semin Fetal Neonatal Med.
2011;16(3):145–50.
58 Morgan JA, Young L, McGuire W. Pathogen-
esis and prevention of necrotizing enterocoli-
tis. Curr Opin Infect Dis. 2011;24(3):183–9.
59 Kandasamy Y. Infection control during ad-
ministration of parenteral nutrition in pre-
term babies. Arch Dis Child Fetal Neonatal
Ed. 2009;94(1):F78.
130.209.6.61 - 8/12/2021 1:18:47 AM
17
Glasgow Univ.Lib.
Downloaded by: