Meta-analysis

Human milk fortifier with

high versus standard protein
content for promoting
growth of preterm infants:
A meta-analysis

Journal of International Medical Research

2015, Vol. 43(3) 279289
! The Author(s) 2015
Reprints and permissions:
sagepub.co.uk/journalsPermissions.nav
DOI: 10.1177/0300060515579115
imr.sagepub.com

Tian-Tian Liu, Dan Dang, Xiao-Ming Lv,

Teng-Fei Wang, Jin-Feng Du and Hui Wu

Abstract
Objective: To compare the growth of preterm infants fed standard protein-fortified human milk
with that containing human milk fortifier (HMF) with a higher-than-standard protein content.
Methods: Published articles reporting randomized controlled trials and prospective observational
intervention studies listed on the PubMed , Embase , CINAHL and Cochrane Library databases
were searched using the keywords fortifier, human milk, breastfeeding, breast milk and human
milk fortifier. The mean difference with 95% confidence intervals was used to compare the effect
of HMF with a higher-than-standard protein content on infant growth characteristics.
Results: Five studies with 352 infants with birth weight 1750 g and a gestational age 34 weeks
who were fed human milk were included in this meta-analysis. Infants in the experimental groups
given human milk with higher-than-standard protein fortifier achieved significantly greater weight
and length at the end of the study, and greater weight gain, length gain, and head circumference gain,
compared with control groups fed human milk with the standard HMF.
Conclusions: HMF with a higher-than-standard protein content can improve preterm infant
growth compared with standard HMF.

Keywords
Human milk, preterm infants, fortifier, protein, growth, meta-analysis
Date received: 10 October 2014; accepted: 4 March 2015

Introduction
Human milk provides benets to general
health, growth and development, while signicantly decreasing the risk for a large
number of acute and chronic diseases.1 It
also has positive eects on mental and

Department of Neonatology, The First Hospital of Jilin

University, Changchun, Jilin Province, China
Corresponding author:
Dr Hui Wu, Department of Neonatology, The First
Hospital of Jilin University, 71 Xinmin Street, Changchun
130021, Jilin Province, China.
Email: wuhui97@126.com

Creative Commons CC-BY-NC: This article is distributed under the terms of the Creative Commons Attribution-NonCommercial
3.0 License (http://www.creativecommons.org/licenses/by-nc/3.0/) which permits non-commercial use, reproduction and
distribution of the work without further permission
provided
original work
is attributed
as2015
specified on the SAGE and Open Access page
Downloaded
from the
imr.sagepub.com
by guest
on May 27,
(http://www.uk.sagepub.com/aboutus/openaccess.htm).

280

Journal of International Medical Research 43(3)

motor development, intelligence quotient

and visual acuity of infants compared with
milk formula feeding.2 Therefore, human
milk is preferred for feeding neonates and
premature infants.3 However, the concentrations of protein, sodium and zinc decline in
human milk as lactation continues over time,
while the nutrient needs of very low birth
weight infants remain high.4 Therefore,
human milk fortier (HMF) may be used
when human milk provides insucient
amounts of the nutrients and energy required
at the usual feeding volumes.5 There is
evidence that the growth of infants fed
human milk fortied with standard HMF
was slower than that of infants fed preterm
formula.6 The authors of this study concluded that the protein content of human
milk with standard fortication failed to meet
the requirements of preterm infants, thereby
aecting their growth.6 Several studies have
demonstrated that HMF containing an
increased protein level is benecial to infant
growth.710 A study undertaken at eight US
hospitals demonstrated that HMFs containing both standard and higher protein levels
were well-tolerated; and that the blood
chemistries were normal and acceptable for
the infants ages.9 The aim of the current
meta-analysis was to determine whether
HMF with a higher-than-standard protein
content would benet preterm infant growth.

Materials and methods

Data sources and keywords
Eligible studies published up to 31 May 2014
were identied by a systematic search of the
electronic literature databases PubMed ,
Embase , CINAHL and the Cochrane
Library. The following keywords were used
in the search for published randomized controlled trials and prospective observational
intervention studies: fortier, human milk,
breastfeeding, breast milk and human milk
fortier.

Selection criteria
Eligible studies were identied initially by
reviewing the abstracts. If the abstract was
consistent with the inclusion criteria, then the
full-text article was obtained. The intervention
groups were feeding infants fortied human
milk that contained a higher-than-standard
protein level. The current dose of extra protein
provided by HMF is between 0.7 and 1.1 g/
100 ml depending on the brand.7 Inclusion
criteria were as follows: (i) randomized controlled trial or prospective observational intervention study; (ii) infants with birth weight
1750 g, gestational age 34 weeks and fed
with human milk; (iii) control group included
infants fed with human milk fortied with
standard HMF, while experimental group
included infants fed with human milk fortied
containing a higher-than-standard protein
level; (iv) outcome measures were growth
characteristics, including weight (g), length
(cm) and head circumference (cm).

Data extraction
Several investigators (T-T.L., D.D.,
X-M.L.) independently extracted the data
and reached a consensus on all the data to be
included in the meta-analysis. The following
data were extracted from each study: name
of rst author, year of publication, sample
size in each group, gestational age and
growth characteristics of the dierent study
groups.

Quality assessment
The quality of the included prospective
observational intervention study was
assessed using predened criteria based on
the Newcastle-Ottawa Scale (NOS) criteria;
and the randomized controlled trials were
assessed using a Modied Jadad score.11,12
This was undertaken by two investigators
(D.D., X-M.L.). The NOS criteria are based

Downloaded from imr.sagepub.com by guest on May 27, 2015

Liu et al.

281

on three study parameters: (i) subject

selection (04); (ii) comparability of subjects
(02); (iii) clinical outcome (03). Total NOS
scores range from 0 (lowest) to 9 (highest).
Studies are classied as low quality (scores
06) or high quality (scores 79). The
Modied Jadad score is based on four
study parameters: (i) randomization (02);
(ii) concealment of allocation (02); (iii)
double blinding (02); (iv) patient withdrawals and dropouts (01). The total
Jadad scores range from 0 (lowest) to 7
(highest). Studies are classied as low quality (03) or high quality (47).

Statistical analyses
The mean dierence (MD) with 95% condence intervals (CI) between the study groups
was used for analysis of continuous outcomes. Summary statistics were calculated
with either a xed-eects or random-eects
model approach, according to the heterogeneity in outcomes across the studies.
Heterogeneity was analysed using 2-test
(signicance level: P < 0.01) and quantied
using the I2 test, where I2 values 50%
indicated substantial heterogeneity. All analyses were carried out using Review Manager
software, version 5.0.13

Results
Characteristics of the included studies
The study selection procedure is presented
in Figure 1. The initial literature search
identied 421 articles based on the title and
keywords. Review of the abstracts resulted
in the exclusion of 402 articles leaving 19 for
full-text review. After screening the full-text
articles according to the inclusion criteria,
the nal meta-analysis included ve
studies.710,14 One study was of low quality
(NOS score 6)14 and four were of high
quality (Modied Jadad score 4).710
Table 1 presents the characteristics of these

ve analysed studies.710,14 A total of 352

infants were included in the ve studies, with
178 in the control groups and 174 in the
experimental groups. The experimental
groups were fed with human milk supplemented with HMF containing a higherthan-standard protein level; and the control
groups were fed with human milk supplemented with HMF containing standard
protein. The baseline growth characteristics
of the infants in the ve studies are shown in
Table 2. Table 3 shows the achieved growth
characteristics and Table 4 shows the rate of
growth of the infants included in the ve
studies.
In the study by Miller et al.,7 infants in
the higher-protein group achieved a greater
weight at the study end (P 0.03), but no
signicant dierences were observed in the
achieved lengths or achieved head circumferences. The length gain, weight gain and
head circumference gain, which were
described using percentiles, demonstrated no
signicant dierences between these two
groups.7 In the report by Moya et al.,8 the
experimental group had a signicantly higher
achieved weight (P 0.04), achieved length
(P 0.03), achieved head circumference
(P 0.043), and length gain (P 0.012) compared with the control group, but there were
no signicant dierences in weight gain and
head circumference gain between the two
groups. In the study by Porcelli et al.,9 infants
had signicantly greater achieved weight
(P 0.0153), achieved head circumference
(P 0.0043), weight gain (P 0.0436) and
head circumference gain (P 0.0354) when
given the new HMF compared with those
given the reference HMF. In the report by
Arslanoglu et al.,10 the experimental group
had signicantly greater weight gain
(P < 0.01) and head circumference gain
(P < 0.05) compared with the control group.
In the study by Alan et al.,14 infants had
signicantly higher daily growth indices for
weight (P 0.026), for length (P 0.027) and

Downloaded from imr.sagepub.com by guest on May 27, 2015

282

Journal of International Medical Research 43(3)

Articles identified through electronic database

searching (n=421)

Articles excluded based on article type,

subject or abstract (n=402)

Articles retrieved for full evaluation (n=19)

Articles excluded due to:

Lack of statistical data (n=1)
Inadequate research question (n=6)
Inadequate study design (n=7)

Studies included in the meta-analysis (n=5)

Figure 1. Flow diagram of the study selection process in a meta-analysis undertaken to determine whether
human milk fortifier with an enhanced protein content would benefit preterm infant growth.

for head circumference (P 0.003) when given

the adjustable protein fortication with
human milk compared with the control group.

Meta-analysis results
In three studies,79 the authors evaluated
growth by measuring the achieved weight,
achieved length, achieved head circumference, weight gain, length gain, and head
circumference gain. In the study by Miller
et al.,7 the authors used the percentile to
describe the weight gain, length gain and
head circumference gain, which was dierent to the other two studies.8,9 In the other
two studies analysed in this meta-

analysis,10,14 the authors only used weight

gain, length gain, and head circumference
gain to compare the growth between the
two study groups. Therefore, this metaanalysis compared the achieved growth
between the two study groups using data
from three studies;79 and it compared
weight gain, length gain, and head
circumference gain using data from four
studies.810,14

Weight
The comparison of the achieved weight and
weight gain was carried out as follows. Since
no signicant heterogeneity in achieved

Downloaded from imr.sagepub.com by guest on May 27, 2015

Downloaded from imr.sagepub.com by guest on May 27, 2015

Australia

USA

USA

Macedonia

Turkey

Miller, 20127

Moya, 20128

Porcelli, 20009

Arslanoglu, 200610

Alan, 201314

Prospective
observational
intervention

RCT

Single-blind RCT

Triple-blind RCT

Double-blind RCT

Study design

RCT, randomized controlled trial; BUN, blood urea nitrogen.

a
30 weeks3 days.

Country

Reference

<32

2434

2532

30 3/7 a

31

Gestational
age, weeks

Protein dose adjustable based

on BUN levels, resulting in
higher protein intakes
Protein dose adjustable based
on BUN levels, resulting
in higher protein intakes

New HMF (1.0 g/100 ml)

HMF with enhanced protein

(1.4 g/100 ml)
New liquid HMF (3.2 g/100 ml)

Experimental group

Protein concentration in HMF

Commercial HMF
(0.8 g/100 ml)

Standard protein
(1.0 g/100 ml)
Standard powdered
HMF (2.6 g/100 ml)
Normal HMF
(0.7 g/100 ml)
Normal HMF
(0.8 g/100 ml)

Control group

29

16

35

51

43

Experimental

Population

29

16

29

55

49

Control

Table 1. Characteristics of studies included in a meta-analysis undertaken to determine whether human milk fortifier (HMF) with an enhanced protein
content would benefit preterm infant growth.

Liu et al.
283

284

Journal of International Medical Research 43(3)

Table 2. Baseline growth characteristics of infants included in five studies in a meta-analysis, undertaken to
determine whether human milk fortifier with an enhanced protein content would benefit preterm infant
growth.
Baseline head circumference, cm
Baseline length, cm

Baseline weight, g

Reference

Experimental Control
group
group

Experimental Control
group
group

Experimental
group

Miller, 20127
Moya, 20128
Porcelli, 20009
Arslanoglu, 200610
Alan, 201314

25.3  2.5
26.3  1.4
28.5  1.8
27.3  2.1
25.3  1.7

35.7  3.5
37.6  1.4
40.6  1.8
39.6  2.6
36.4  2.4

25.6  2.0
26.1  1.5
27.8  1.6
27.7  2.2
26  1.9

Control
group

35.6  3.4
1012  315
1056  289
37.5  1.5
1162  194
1124  212
39.4  2.2 1419.5  228.4 1255.2  194.9
39.6  2.1
1501  252
1526  181
37.1  2.8
1050  203
1113  231

Data presented as mean  SD.

Table 3. Achieved growth characteristics of infants included in three studies in a meta-analysis undertaken
to determine whether human milk fortifier with an enhanced protein content would benefit preterm infant
growth.a

Reference

Achieved head circumference, cm

Achieved length, cm

Achieved weight, g

Experimental
group

Control
group

Experimental Control
group
group

Experimental Control
group
group

33.8  1.8
29.9  1.5
30.0  1.6

46.3  2.1
42.4  2.1
42.9  2.4

33.5  1.8
Miller, 20127
Moya, 20128
30.6  2.1
Porcelli, 20009 31.0  1.8

45.5  3.0 2760  498

41.2  1.5 1829  300
41.6  2.7 1994  358

2539  494
1662  267
1711  353

Data presented as mean  SD.

a
Only three of the five studies evaluated growth by measuring achieved weight, achieved length, and achieved head
circumference.

Table 4. Growth rate of preterm infants included in four studies in a meta-analysis undertaken to determine
whether human milk fortifier with an enhanced protein content would benefit preterm infant growth.a
Head circumference, cm/week

Length, cm/week

Weight, g/kg per day

Reference

Experimental
group

Control
group

Experimental Control
group
group

Experimental Control
group
group

Moya, 20128
Porcelli, 20009
Arslanoglu, 200610
Alan, 201314

1.12  0.25
1.00  0.59
0.98  0.21
0.89  0.24

1.05  0.31
0.80  0.54
0.70  0.35
0.63  0.21

1.19  0.30
0.90  0.59
0.91  0.35
1.12  0.43

16.6  3.5
19.7  5.8
17.5  3.2
20.9  4.7

0.98  0.31
0.80  0.54
0.77  0.28
0.81  0.37

15.8  3.7
16.8  5.2
14.4  2.7
18.9  4.5

Data presented as mean  SD.

a
Only four of the five studies evaluated growth by measuring the weight gain, length gain, and head circumference gain. Miller
et al.7 used percentiles to describe the weight gain, length gain and head circumference gain, which was different
methodology to the other studies so the data were excluded.

Downloaded from imr.sagepub.com by guest on May 27, 2015

Liu et al.

285

Figure 2. Forest plots for the effect of higher versus standard protein content of human milk fortifier on
infant achieved weight.79

Figure 3. Forest plots for the effect of higher versus standard protein content of human milk fortifier on
infant weight gain.810,14

Figure 4. Forest plots for the effect of higher versus standard protein content of human milk fortifier on
infant achieved length.79

weight (2-test 1.26, I2 0%) and weight

gain (2-test 4.25, I2 29%) was indicated, a xed-eects model was used to
conduct the meta-analysis. As shown in
Figures 2 and 3, the results indicated that
the achieved weight (MD: 202.94; CI:
119.01, 286.86; P < 0.00001) and weight
gain (MD: 1.77; CI: 0.81, 2.73; P 0.0003)
of infants in the experimental groups was
greater than those of the control groups.

Length
For the comparison of the achieved length
(2-test 0.49, I2 0%) and length gain (2test 1.89, I2 0%), a xed-eects model
was used since no signicant heterogeneity
was detected. As shown in Figures 4 and 5,
the results indicated that the achieved length
(MD: 1.12; CI: 0.59, 1.64; P < 0.0001) and
length gain (MD 0.21; CI: 0.12, 0.29;

Downloaded from imr.sagepub.com by guest on May 27, 2015

286

Journal of International Medical Research 43(3)

Figure 5. Forest plots for the effect of higher versus standard protein content of human milk fortifier on
infant length gain.810,14

Figure 6. Forest plots for the effect of higher versus standard protein content of human milk fortifier on
infant achieved head circumference.79

Figure 7. Forest plots for the effect of higher versus standard protein content of human milk fortifier on
infant head circumference gain.810,14

P < 0.00001) of infants in the experimental

groups was longer than those of the control
groups.

Head circumference
The comparison of the achieved head circumference (2-test 6.14, I2 67%) and

head circumference gain (2-test 6.86,

I2 56%) were carried out using a
random-eects model due to the signicant
heterogeneity. As shown in Figures 6 and 7,
the achieved head circumference in the
experimental groups demonstrated no
signicant dierence compared with control
groups (MD: 0.45; CI: 0.31, 1.22); whereas

Downloaded from imr.sagepub.com by guest on May 27, 2015

Liu et al.

287

the results indicated that the head circumference gain of infants in the experimental
groups was larger than that of the control
groups (MD: 0.19; CI: 0.07, 0.31; P 0.002).

Discussion
With the rapid development of medical technology, the number of premature newborns
with small gestational age and low birth
weight continues to rise, and clinicians face
challenges in providing proper nourishment
for these infants. Research has conrmed
that breast milk fortiers can improve shortterm growth,15 but they result in lower rates
of weight gain and increases in length in
infants compared with preterm milk formula.16 This phenomenon may be explained
by the fact that the actual intake of protein by
preterm infants fed fortied human milk is
substantially
lower
than
previously
assumed.17 By consulting the relevant literature, the current meta-analysis attempts to
provide a comprehensive evaluation of HMF
containing a higher-than-standard level of
protein on preterm infant growth.
This current meta-analysis demonstrated
that a higher protein intake resulted in a
signicantly higher achieved weight, achieved
length, weight gain, length gain and head
circumference gain. As for the achieved head
circumference, there was no signicant difference between the two groups, but the head
circumference gain in the experimental
groups was signicantly greater than that of
the control groups. Taken together, the current results suggest that HMF containing
more protein than the standard level can
improve preterm infant growth.
Research has demonstrated that there
were no signicant dierences in blood
biochemistry data, such as mean blood
urea nitrogen, calcium levels, phosphorus
levels and alkaline phosphatase, between
infants fed HMF containing more protein
and control infants fed HMF with prtein at

a standard level.10 In addition, the higher

protein-containing HMF was well-tolerated
by the preterm infants.10 These ndings,
taken together, suggest that clinicians might
nd it favourable to use HMF containing an
increased protein content when they need to
enhance the growth of preterm infants.
There are no studies about the long-term
growth and neurodevelopmental outcomes
associated with the use of HMFs containing
higher-than-standard levels of protein, so
further research is needed in this area.
In the current meta-analysis, there was
signicant heterogeneity between the studies
when the achieved head circumference and
head circumference gain were compared.
There are a number of potential explanations for this heterogeneity: (i) dierences
in the populations at the international
research sites; (ii) variations in gestational
age; (iii) variations in the HMF product
used and its protein content (although the
protein content in the experimental groups
was higher than the control groups, the
actual dose of protein was dierent in the
dierent HMFs used across the ve studies);
and (iv) dierences in the software and
statistical methods used to analyse the data.
This present meta-analysis had several
limitations. In the ve studies analysed, not
only were protein levels in the experimental
groups higher than those in the control
groups, but the levels of other nutrients,
such as calcium and zinc, were also dierent.
Thus, the resulting growth reected the
combined eects of these nutrients, and the
exact role of the absolute protein level in
the HMF was dicult to interpret.
Although the HMF protein level was
higher in the experimental group than in
the control group of each study, the distribution in each group was not balanced (i.e.
dierences in the intake of milk when using
fortier), which could have potentially led to
bias. Although an extensive search strategy
was used to nd published studies, no

Downloaded from imr.sagepub.com by guest on May 27, 2015

288

Journal of International Medical Research 43(3)

attempt was made to identify unpublished

trials. Meta-analyses are inherently limited by
the quantity and quality of existing data,
which are likely to exhibit heterogeneity.
The statistically signicant heterogeneity
among the studies renders the results of this
current meta-analysis less meaningful, but
statistical conclusions on determinants of
heterogeneity may be awed, given the small
number of included studies. Additional
multicentre randomized controlled trials
are needed to clarify the long-term eects
of HMF containing higher-than-standard
levels of protein on preterm infant growth.
In conclusion, the ndings of this current
meta-analysis of ve studies that included
352 infants with a birth weight 1750 g and
a gestational age 34 weeks who were fed
human milk suggest that using HMF containing a higher-than-standard protein content can improve preterm infant growth.

Declaration of conflicting interest

The authors declare that there are no conicts of
interest.

Funding
This research received no specic grant from any
funding agency in the public, commercial, or notfor-prot sectors.

References
1. Breastfeeding and the use of human milk.
American Academy of Pediatrics, Work
group on breastfeeding. Breastfeed Rev 1998;
6: 3136.
2. Morales Y and Schanler RJ. Human milk and
clinical outcomes in VLBW infants: how
compelling is the evidence of benefit? Semin
Perinatol 2007; 31: 8388.
3. Kreissl A, Zwiauer V, Repa A, et al. Effect of
fortifiers and additional protein on the
osmolarity of human milk: is it still safe for
the premature infant? J Pediatr Gastroenterol
Nutr 2013; 57: 432437.

4. Ziegler EE. Breast-milk fortification. Acta

Paediatr 2001; 90: 720723.
5. Mukhopadhyay K, Narnag A and Mahajan
R. Effect of human milk fortification in
appropriate for gestation and small for
gestation preterm babies: a randomized controlled trial. Indian Pediatr 2007; 44: 286290.
6. Corvaglia L, Aceti A, Paoletti V, et al.
Standard fortification of preterm human milk
fails to meet recommended protein intake:
bedside evaluation by near-infrared-reflectance-analysis. Early Hum Dev 2010; 86:
237240.
7. Miller J, Makrides M, Gibson RA, et al.
Effect of increasing protein content of human
milk fortifier on growth in preterm infants
born at <31 wk gestation: a randomized
controlled trial. Am J Clin Nutr 2012; 95:
648655.
8. Moya F, Sisk PM, Walsh KR, et al. A new
liquid human milk fortifier and linear growth
in preterm infants. Pediatrics 2012; 130:
e928e935.
9. Porcelli P, Schanler R, Greer F, et al.
Growth in human milk-fed very low birth
weight infants receiving a new human milk
fortifier. Ann Nutr Metab 2000; 44: 210.
10. Arslanoglu S, Moro GE and Ziegler EE.
Adjustable fortification of human milk fed to
preterm infants: does it make a difference?
J Perinatol 2006; 26: 614621.
11. Stang A. Critical evaluation of the
Newcastle-Ottawa scale for the assessment
of the quality of nonrandomized studies in
meta-analyses. Eur J Epidemiol 2010; 25:
603605.
12. Jadad AR, Moore RA, Carroll D, et al.
Assessing the quality of reports of randomized clinical trials: is blinding necessary?
Control Clin Trials 1996; 17: 112.
13. Review Manager (RevMan) [Computer
program]. Version 5.0. Oxford: The
Cochrane Collaboration.
14. Alan S, Atasay B, Cakir U, et al. An
intention to achieved better postnatal
in-hospital-growth for preterm infants:
adjustable protein fortification of
human milk. Early Hum Dev 2013; 89:
10171023.
15. Lucas A, Fewtrell MS, Morley R, et al.
Randomized outcome trial of human milk

Downloaded from imr.sagepub.com by guest on May 27, 2015

Liu et al.

289

fortification and developmental outcome in

preterm infants. Am J Clin Nutr 1996; 64:
142151.
16. Schanler RJ, Shulman RJ and Lau C.
Feeding strategies for premature infants:
beneficial outcomes of feeding fortified

human milk versus preterm formula.

Pediatrics 1999; 103: 11501157.
17. Arslanoglu S, Moro GE and Ziegler EE.
Preterm infants fed fortified human milk
receive less protein than they need.
J Perinatol 2009; 29: 489492.

Downloaded from imr.sagepub.com by guest on May 27, 2015