Supplemental Material can be found at:

http://jn.nutrition.org/content/suppl/2016/02/10/jn.115.22365
1.DCSupplemental.html
The Journal of Nutrition
Nutrition and Disease

Protein Concentration in Milk Formula, Growth,

and Later Risk of Obesity: A Systematic
Review1–3
Bernadeta Patro-Go1a˛b,4* Bart1omiej M Zalewski,4 Stefanie MP Kouwenhoven,5 Jacek Karaś,4
Berthold Koletzko,6 Johannes Bernard van Goudoever,5,7 and Hania Szajewska4
4
Department of Pediatrics, The Medical University of Warsaw, Warsaw, Poland; 5Department of Pediatrics, VU University Medical
Center, Amsterdam, Netherlands; 6Ludwig Maximilians University, Dr. von Hauner ChildrenÕs Hospital, Division of Metabolic and
Nutritional Medicine, University of Munich Medical Centre, Munich, Germany; and 7Department of Pediatrics, Emma ChildrenÕs
Hospital, Amsterdam Medical Center, Amsterdam, Netherlands

Abstract
Background: Protein intake may influence important health outcomes in later life.

Downloaded from jn.nutrition.org by guest on March 8, 2017

Objective: The objective of this study was to investigate current evidence on the effects of infant formulas and follow-on
formulas with different protein concentrations on infantsÕ and childrenÕs growth, body composition, and later risk of
overweight and obesity.
Methods: In this systematic review, we searched electronic databases (including MEDLINE, Embase, and the Cochrane
Library) up until November 2014 for randomized controlled trials (RCTs). Eligible studies had to include children aged 0–3 y
who represented the general population and were fed cow milk–based infant formulas with variations in protein
concentration. Control groups received lower-protein cow milk–based formulas (as defined by the authors). The primary
outcomes were growth, overweight, obesity, and adiposity. Various time points for outcomes assessment were accepted
for inclusion. If possible, a meta-analysis was performed.
Results: Twelve RCTs met our inclusion criteria. Different formula protein concentrations did not affect linear growth other
than a transient effect on mean length at 3 mo observed in a meta-analysis of 4 studies (mean difference, – 0.27 cm; 95%
CI: 20.52, 20.02). Lower mean weight and weight z scores obtained from the infants fed lower-protein formulas were
observed only from 6 to 12 mo of age. Data from one large RCT showed that consumption of a lower-protein infant formula
may reduce body mass index at 12 mo of age and later (12 mo, 24 mo, and 6y) and the risk of obesity at 6 y. Effects on body
composition remained unclear.
Conclusions: The current evidence is insufficient for assessing the effects of reducing the protein concentration in infant
formulas on long-term outcomes, but, if confirmed, this could be a promising intervention for reducing the risk of
overweight and obesity in children. In view of the limited available evidence, more studies replicating effects on long-term
health outcomes are needed. J Nutr 2016;146:551–64.

Keywords: RCT, systematic review, infant nutrition, pediatrics, protein

Introduction
Overweight and obesity are among the most serious public
health challenges. The effectiveness of different treatment
modalities is limited, and, thus, interest in the identification of
modifiable risk and protective factors is growing. One of these
includes research on the potential role of infant feeding on body
composition and the likelihood of being overweight or obese in
later life.
Sufficient protein intake early in life is of major importance.
Deficient protein intake can lead to suboptimal growth and
impaired neurodevelopment (1). However, protein intake that is
too high may have adverse effects as well. It has been
documented that rapid weight gain in infancy is associated
ã 2016 American Society for Nutrition.
Manuscript received September 8, 2015. Initial review completed October 20, 2015. Revision accepted January 8, 2016.
First published online February 10, 2016; doi:10.3945/jn.115.223651.
with an increased chance of later obesity (2), and weight gain
from birth to 24 mo is the best overall predictor of later
overweight (3). High early protein intake in excess of metabolic
requirements may stimulate the secretion of insulin and insulin-
like growth factor I (4), thereby enhancing weight gain in
infancy and increasing the later risk of obesity. This is known as
the early protein hypothesis (5), and it is based on early
observations by Rolland-Cachera et al. (6). These investigators
related a high protein intake during early childhood to modified
endocrine responses and an increased risk of obesity at school
age. The mechanisms by which increased protein intake affects
weight gain and body composition are not yet completely clear.
However, the endocrine and metabolic responses of infants are
551
affected substantially by dietary protein intake (7). Therefore,
lowering protein intake from milk formulas may exhibit
beneficial effects on weight gain, body composition, and
metabolic diseases in later life.
There is currently no consensus regarding what should be
the appropriate amount of protein in infant formulas (8–12).
European Directive 2006/141/EC on infant and follow-on
formulas (13), as well as the FDA (14), Codex Alimentarius
(15), and European Food Safety Authority (16), define the
minimum protein-to-energy ratio for infant formulas based on
cow milk and soy protein as 1.8 g/100 kcal and 2.25 g/100 kcal,
respectively, but they do not define an optimal intake amount.
In 2014, the European Food Safety Authority carried out a
literature search and review as preparatory work for the
evaluation of the composition of infant and follow-on formulas
and growing-up milk. With regard to reduced-protein formulas,
this report concluded that neither negative health effects nor
clear benefits from the use of this type of formula could be
established (8). Because there is uncertainty regarding the effect
of dietary protein intake in childhood on growth and body
composition, and the later risk of overweight, obesity, and
metabolic syndrome, we aimed to investigate systematically the
current evidence on this proposed relation. The review protocol
was registered on the PROSPERO International Prospective
Register for Systematic Reviews (crd.york.ac.uk/prospero/
index.asp) as CRD42014015573.
Methods
Inclusion/exclusion criteria. Study inclusion and exclusion criteria are
summarized in Table 1 (CRD42014015573).
Search methods for identification of studies. The following
electronic databases were searched until November 2014 for published
studies that fulfilled our criteria: Cochrane Central Register of Controlled
Trials, PubMed (including MEDLINE), Embase, and the Cumulative
Index to Nursing and Allied Health Literature. To identify potential
systematic reviews/meta-analyses, we browsed The Cochrane Database
of Systematic Reviews and Database of Abstracts of Reviews of Effects.
1
The research leading to these results received funding from the European
Union’s Seventh Framework Programme (FP7/2007-2013), project EarlyNutrition
under grant agreement no. 289346. The study formula for the Childhood Obesity
Programme study was produced by Bledina, Steenvorde, France as part of a
contract with the European Commission.
2
B Patro-Go1a˛b and BM Zalewski participated in a conference sponsored by
Nestle. SMP Kouwenhoven is a PhD student involved in a study designed to look
at the effects of a low-protein formula on weight gain and body composition. B
Koletzko is a member of the National Breastfeeding Committee and tends to be
biased toward breastfeeding. Both B Koletzko and his employer, the
Ludwig-Maximilians-University of Munich, Germany, have received support for
scientific and educational activities by companies that market products for
infants and children, including Abbott, Baxter, B. Braun, Dairy Goat Cooperative,
Danone, Fresenius Kabi, Fonterra, Hipp, Mead Johnson, Nestlé, and Yakult,
predominantly as part of publicly funded research projects with the support of
the European Commission or the German government. JB van Goudoever holds
patents for a specific blend of essential amino acids in order to develop a new
infant formula, and is the sponsor of a study funded by the European Union that is
designed to look at the effects of a low-protein formula on weight gain and body
composition. He participated as a clinical researcher, advisory board member,
and/or speaker for Danone, Dicofarm, Ferrero, Hipp, Nestle Nutrition Institute,
Nutricia, and MJN. H Szajewska has participated as a clinical investigator,
advisory board member, and/or speaker for Arla, Biogaia, Biocodex, Danone,
Dicofarm, Hipp, Nestle, Nestle Nutrition Institute, Nutricia, Mead Johnson,
Merck, Sequoia, and Yakult. J Karaś, no conflicts of interest.
3
Supplemental Figure 1 and Supplemental Tables 1–3 are available from the
"Online Supporting Material" link in the online posting of the article and from the
same link in the online tables at http://jn.nutrition.org.
*To whom correspondence should be addressed. E-mail: abpatro@yahoo.com.
552 Patro-Go1a˛b et al.
Four reviewers (BP-G, BMZ, SMPK, and JK) were involved in the
process of the search, which was carried out independently by at least 2
reviewers. No language restrictions were applied. A detailed search
strategy (Supplemental Table 1) was prepared with the support of an
information specialist from the Centre for Reviews and Dissemination,
University of York. Initially, the title, abstract, and keywords of every
record identified with the use of our search strategy were screened.
Irrelevant articles were excluded by title or abstract. Full texts were
obtained for all potentially relevant studies. Differences between
reviewers were resolved by discussion until a consensus was reached.
The reference lists from identified studies and key review articles, as well
as selected trial registries (clinicaltrials.gov, www.clinicaltrials.gov; EU
Clinical Trials Register, www.clinicaltrialsregister.eu), were searched.
Proceedings from major scientific gastrointestinal and nutritional
meetings published in the last 3 y also were screened. Finally, an
attempt was made to obtain additional data by direct contact with
experts in the field.
Data collection and analysis. An initial screening of the title, abstract,
and keywords of every record identified was performed. The next step
was to retrieve the full text of potentially relevant studies. Three
reviewers (BP-G, BMZ, and JK) independently assessed the eligibility of
each potentially relevant study with the use of the inclusion criteria. All
cases in which different opinions or questions about the eligibility of a
particular study occurred were discussed with at least one other member
Downloaded from jn.nutrition.org by guest on March 8, 2017
of the group (usually the expert of the team).
Data extraction and management. With the use of a previously prepared
template of a table of evidence, 2 reviewers independently extracted the
data necessary for study characteristics and risk of bias assessment.
Assessment of risk of bias. The Cochrane CollaborationÕs tool for
assessing risk of bias was used while including the following criteria:
adequacy of sequence generation; allocation concealment; blinding of
participants, personnel, and outcome assessors; incomplete outcome
data; and selective reporting.
Measures of treatment effect. If feasible, for the continuous
outcomes, the results have been reported as the mean difference (MD)8
between the lower-protein formula groups and higher-protein formula
groups with a 95% CI. For the dichotomous outcome (prevalence of
obesity), the results have been reported as the RR between the lower-
protein and higher-protein formula groups with a 95% CI. Missing SDs
in 2 randomized controlled trials (RCTs) were calculated based on the
reported SEMs with the use of the method described in the Cochrane
Handbook for Systematic Reviews of Interventions (17). For outcomes
expressed with the use of different units, if possible, data were
transformed for their unification (e.g., millimeters into centimeters).
Four authors were contacted by e-mail or personally to obtain missing
data or clarify equivocal issues. This resulted in our obtaining additional
unpublished data from 2 published trials (18, 19).
Assessment of heterogeneity. Clinical and methodologic inconsistency
between the selected studies was assessed during the initial analysis, data
extraction, and quality assessment process. The decision to pool data
together into a meta-analysis was based mainly on this assessment.
Additional information about heterogeneity was obtained when the results
of the chi-square test and I2 value were analyzed for some outcomes.
Assessment of reporting biases. Although originally planned,
because of the insufficient number of studies included in the meta-
analysis (<10), we did not create a funnel plot, nor did we perform
a statistical test for its asymmetry to detect potential publication bias.
Data synthesis. A narrative synthesis was undertaken if conditions
required for performance of a meta-analysis were not fulfilled. For
8
Abbreviations used: CHOP, EU Childhood Obesity Programme; MD, mean
difference; RCT, randomized clinical trial.
TABLE 1 Study inclusion and exclusion criteria

Inclusion criteria Exclusion criteria

Types of study Randomized controlled trials and quasi-randomized trials

were included
Types of participants Infants and young children (ages 0–3 y), representing the Preterm infants
general population Children with chronic diseases
Types of intervention Cow milk–based infant formulas and/or follow-on formulas Duration of intervention ,3 mo
with variations in protein quantity
Mixed feeding groups (breastfed and concurrently formula- Formula with a protein from a different source
fed children), if formula feeding was a predominant (soy, rice, hydrolyzed, elemental, etc.)
source of milk
The differences in composition of amino acids and/or Milk formula with additional modifications other than
macronutrients that were secondary to the protein quantitative protein (i.e., addition into milk formula of
amount modification (in order to achieve an isocaloric extra components such as a-lactalbumin, lactoferrin,
formula) were not considered as an actual additional long-chain PUFAs, prebiotics, probiotics, synbiotics, fat
modification that disqualifies the study from inclusion globule membranes, tryptophan, taurine, or cholesterol).
Types of comparison Lower-protein cow milk–based infant formulas and follow-
on formulas
Types of outcome
Primary outcome measures Obesity and overweight, as defined by the authors
Anthropometric growth markers: BMI, weight, length/
height, head circumference, or changes in these markers

Downloaded from jn.nutrition.org by guest on March 8, 2017

at different time intervals (as expressed by the authors)
Fat mass and fat-free mass (measured with the use of
different methods/devices)
Secondary outcome measures1 Biomarkers of disease (i.e., concentrations of TGs; HDL,
LDL, and total cholesterol; and fasting glucose)
Metabolic syndrome criteria
Concentrations of insulin-like growth factor I and insulin
Other data reported by the investigators if relevant to the
current review
1
Studies with various time points for outcomes assessment (directly after the intervention or after a different follow-up period) were accepted for inclusion.

quantitative analysis and pooling individual study data together, we used
the Review Manager computer program, version 5.3 (The Nordic Cochrane
Centre; The Cochrane Collaboration). For the 4 multiarm studies included in
the meta-analysis (18, 20–22), in order to avoid a unit-of-analysis error, we
combined the study groups to achieve a single pairwise comparison while
following the formulas provided by the Cochrane Handbook for Systematic
Reviews of Interventions (17). Some studies, although reporting an outcome
of interest, were excluded from the meta-analysis because of a lack of
necessary data (a lack of sample size, narrative results description without
numerical data, etc.). The authors of one trial assessed the outcomes for
lower- and higher-protein formula groups at different ages of children; thus,
this trial was excluded from the analysis (23). In the case of another study, we
could not rule out the possibility of the presentation of duplicated data (21).
However, because data from this study were not pooled together in the meta-
analysis with data provided by other authors, the inclusion of this study is
unlikely to result in biased estimates. The study aimed to compare infants that
were given different weaning foods (Swedish or Mediterranean); however,
children in both groups were randomly assigned to receive formula with a
lower or higher protein concentration. Because the authors provided
sufficient data, we compared the growth of infants within each study
group (different protein concentrations), but not between the groups
themselves (different weaning food). In the case of one trial [Raiha et al.
(24)], when presenting the results, we made an assumption that the mean
values were accompanied by SD values (but not SEM values). However,
this was not clearly stated in the original paper.
Results
For a PRISMA flow diagram documenting the identification
process for eligible trials, see online Supplemental Figure 1. For
the characteristics of the excluded trials, with reasons for
exclusion, see online Supplemental Table 2. The characteristics
of 17 included publications describing 12 trials (10 RCTs and 2
quasi-randomized trials) (18–34) are summarized in Table 2. Six
of the 17 selected publications reported the results of the EU
Childhood Obesity Program (CHOP) study (19, 30–34).
All study participants were healthy, term infants. The
majority of children were from birth to 4 mo of age at
enrollment and exclusively formula-fed during the first months
of life. Except for 2 trials conducted in China (25, 26), the
included trials were undertaken in Europe.
The protein concentration in different infant formulas varied
greatly between the studies, ranging from 1.1 to 2.1 g/100 mL in the
lower-protein formula group to 1.5 to 3.2 g/100 mL in the higher-
protein formula group. In one trial (28), the protein concentration
was unreported (described as low- and high-protein formulas).
Differences in the types of protein used with respect to the whey-to-
casein ratio were observed between the studies and between the
study groups within individual studies. Some of the included studies
were multiarm studies with >2 intervention groups. Two separate
studies by Lonnerdal and Chen (25, 26) involved the same
population of infants observed during different time periods
(from birth and from 4 mo of age). In addition, in many trials, a
reference group of breastfed infants was recruited. The duration
of the intervention usually lasted between 3 and 5 mo, with the
exception of that in the CHOP study, in which infants received
study formula until 12 mo of age. Also, The CHOP study was the
only identified study with a longer (>12 mo of age) follow-up.
Protein concentration of infant formulas 553
554
Patro-Go1a˛b et al.

TABLE 2 Characteristics of included studies1

Study Sample size Higher-protein Outcomes relevant

Reference type and setting Population Lower-protein group group for this review Effect size Funding Comments

Zoppi et al., Quasi-RCT n = 55, Italy Healthy, term, n = 28; commercially avail- n = 27; commercially Weight, length, and their Mean weight and mean Not reported The study was designed
1978 (29) nonbreastfed able powdered milk available powdered gain measured every 2 gain in weight slightly to assess the relation
infants (whey:casein ratio 18:82) milk (whey:casein ra- wk (data provided for lower in lower-protein between blood g-glob-
fed exclusively up to 179 d tio 18:82); protein at a days 164, 194, and 301) group at ages 164– ulin concentrations and
of life; protein concen- 15% concentration: 3.0 g 301 d, with statistically dietary protein intake
Age: from day 1 tration 15%: 2.1 g and 66 and 64.2 kcal/100 mL significant difference Higher amount of carbo-
of life kcal/100 mL (3.2 g/100 (4.7 g/100 kcal) at 5.5 mo of age only; hydrates in lower-pro-
kcal) mean gain in length the tein groups
same (NS) Children with thalassemia
and illnesses requiring
hospitalization not in-
cluded
Zoppi et al., Quasi-RCT n = 41, Italy Healthy, term n = 7; fed exclusively with n = 7; fed exclusively with Weight and length mea- Growth (weight, length, Research supported in The study was designed
1982 (23) infants cow-milk formula with a cow-milk formula with sured every 2 wk and head circumfer- part by CNR and MPI to assess immunocom-
protein content of 1.6 g/ a protein content of 3.0 g/ ence) normal and simi- grants petence
100 mL (2.2 g/kcal); 74.2 100 mL (4.0 g/kcal); lar in all infants**
Age: from birth kcal/100 mL for 4.5 mo 75.4 kcal/100 mL Serum cholesterol and Observed concentrations The aim of the study was
TG concentrations reflected the amount to compare cow-milk
assessed at 4.5 mo and quality of protein formula and soy for-
of age intake; TG concentra- mula; hospitalized in-
tions within the normal fants were excluded
range**
Mean daily gain in weight No difference between
and length, and mean groups
increase in head cir-
cumference
Zoppi et al., RCT n = 62, Italy Healthy, term Fed exclusively with low- n = 7; fed exclusively with Mean daily gains in Mean daily gains in Not reported The study was designed
1983 (28) infants who protein cow milk–based high-protein cow milk– weight, length, and weight, length, and to assess the antibody
were to be formula (n = 9) or adapted based formula head circumference head circumference: response to different
artificially fed according to ESPGHAN ``normal`` and ``similar vaccines; group 4 re-
recommendations–based to earlier findings``** ceived soy formula (n = 9);
formula (n = 10) for 5 mo 27 infants were breastfed
Age: from birth The exact amount of protein The exact amount of pro-
in the formula was not tein in the formula was
reported not reported

(Continued)

Downloaded from jn.nutrition.org by guest on March 8, 2017

 TABLE 2 Continued

Study Sample size Higher-protein Outcomes relevant

Reference type and setting Population Lower-protein group group for this review Effect size Funding Comments

R€aih€a et al., RCT n = 30, Italy Healthy, term n = 10; formula with a n = 10; formula with a Mean weight, length, and NS Study supported by Third arm of study—
1986 (24) infants protein content of 1.2 g/ protein content of 1.6 head circumference at Findus, Bjuv, Sweden; breastfed infants; 20
100 mL (whey:casein ra- g/100 mL (whey:casein 1, 2, 4, 8, and 12 wk of Nestle, Vevey, infants in formula-fed
tio 60:40), 67 kcal/100 ratio 60:40), 67 kcal/ age; Switzerland; and The groups (unclear
Age: from birth mL, given for 12 wk 100 mL Growth rate from 2 to 12 NS National Research whether the number of
wk compared only with Council, Italy subjects was the same
human milk group in each group)
Picone et al., RCT n = 43, Italy Healthy, term n = 10; experimental formula n = 10; formula with a Mean weight, length, and NS Study supported in part by Breastfed group, n = 10
1989 (22) infants with a 50:50 blend of protein concentration head circumference at Ross Laboratories,
Age: from birth bovine whey and casein of 14.8 g/L birth and 2, 4, 8, and 12 United States
at different concentra- wk of age
tions (1.12 g/L or
13.3 g/L) for 12 wk
Lonnerdal and RCT n = not reported, Healthy, term n = not reported; fed exclu- n = not reported; fed Mean weight, length, Similar for all groups** Study supported by Wei- Five formula groups and
Chen, I 1990 China infants sively formula with the exclusively formula weight:height, and Chuan Foods; Clinical one exclusively
(25) following protein and with protein and whey: head circumference Nutrition Research Unit breastfed group
whey:casein ratio: casein ratio 1.5 g/dL, and skin folds at 2, 4, 8,
21.3 g/dL, 55:45 55:45 and 12 wk of age
Age: from birth 21.4 g/dL, 55:45 Weight gain Similar for all groups**
21.4 g/dL, 60:40 Length gain
21.4 g/dL, 20:80;
67 kcal/dL for 12 wk
Protein concentration of infant formulas

Lonnerdal and RCT n = not reported, Healthy, term Follow-up formula with the Commercially available Mean weight, length, Similar for all groups** Wei-Chuan Foods; Clinical The same population
Chen, II 1990 China infants, following protein and follow-up formula weight:height, and Nutrition Research Unit studied in previous
(26) formula-fed whey:casein ratio: (Promil) with a protein head circumference (grant no. P30- Lonnerdal I study
from birth 21.5 g/dL, 55:45 with content of 2.9 g/dL and and skin folds at 5, 6, AM235747) (0–4 mo)
25 g cereal/d as supple- whey:casein ratio and 7 mo of age
Age: 4 mo mental food or without 60:40 (67 kcal/dL), with Weight gain Similar for all groups** AuthorsÕ conclusions:
cereal 25 g cereal/d as sup- Length gain higher protein intake
22.2 g/dL, 55:45 (67 plemental food appears to be exces-
kcal/dL) with 25 g cereal/d sive; lower protein in-
as supplemental food take is adequate

(Continued)
555

Downloaded from jn.nutrition.org by guest on March 8, 2017

556
TABLE 2 Continued

Study Sample size Higher-protein Outcomes relevant

Patro-Go1a˛b et al.

Reference type and setting Population Lower-protein group group for this review Effect size Funding Comments

Szajewska et al., RCT n = 60, Poland Healthy, term, n = 14; experimental formula n = 14; formula with a Mean weight, length, and NS Study supported in part by Breastfed group, n = 15
1993 (18) exclusively with 1.5 g/100 mL pro- protein content of 2.2 head circumference at Milupa
formula-fed tein; whey:casein ratio g/100 mL; whey:casein day 10; also at weeks
infants 60:40 for 12 wk ratio 60:40; 4, 8, and 12
Age: from birth n = 15; formula with a
Akeson et al., RCT n = 80, Sweden; Healthy, term Infant formula containing 13 protein content of 2.2 Gains in weight and NS Study financially sup- Infants that withdrew
1998 (20) number of in- infants, g protein/L (n = 25 at 3 g/100 mL; whey:casein length at 4–8 and 8–12 ported by the Swedish from the study were
fants in the previously mo) or 15 g protein/L (n = ratio 18:82; mo of age Nutrition Foundation replaced by next eligi-
different exclusively BF 26 at 3 mo) gradually infant formula contain- ble subject; at 6 mo,
groups Age: 3 mo introduced to infants ing 18 g protein/L (n = z scores of the absolute NS infants fed formula
changed over along with breastfeeding. 23 at 3 mo) weight and length at 3, ,125 mL/d formed the
time (see At the age of 6 mo, 6, and 12 mo of age breastfed group, and
Comments row) exclusively formula-fed Crown–heel length, lower NS those fed .125 mL/d
groups, a mixed feeding leg length, and head formed the mixed-
group, and a breastfed and arm circumfer- feeding group; then got
group were formed ences back to previous groups
when breastfeeding
was reduced or stop-
ped; These infants
were replaced by the
subjects exclusively
formula-fed
Karlsland Akeson RCT n = 80, Sweden; Healthy, term Gradually introduced (when Infant formula with pro- Weight and length at 3, 4, Normal growth; no differ- Study supported by Ross An exclusion criterion was
et al., 2000 n = 59, Italy infants, breastfeeding termi- tein concentrations of 6, and 12 mo of age ences between corre- Laboratories, United breastfeeding at $6
(21) breastfed nated) infant formula 18 g/L (Swedish in- sponding Swedish and States; Nestlé Sweden mo of age (47 Swedish
for $2–3 mo with different protein fants; n = 8) or 20 g/L Italian groups during AB; the Albert Pahlsson and 18 Italian infants);
concentrations: (Italian infants; n = 9) the study period Foundation, Sweden; between 10 and 12 mo
Age: 3 mo 213 g/L; n = 10 (Swedish Gains in weight and Data not compared within the F€orenade Liv of age, formula ex-
infants) and n = 12 (Italian length in Swedish and one group (low- vs. Mutual Group Life changed for cow-milk
infants) Italian infants between high-protein formulas) Insurance Company, products in most Italian
215 g/L; n = 9 (Swedish 3 and 6 mo and 6 and ** Sweden; and the infants; at 12 mo of
infants) and n = 9 (Italian 12 mo of age Swedish Nutrition age, all Swedish and
infants) exclusively fed Foundation 21% (n = 6 of 28) of
for $4 mo (6–10 mo of Italian infants still re-
age; see Comments col- ceived formula
umn);
in addition, Swedish or
Mediterranean weaning
foods given

(Continued)

Downloaded from jn.nutrition.org by guest on March 8, 2017

 TABLE 2 Continued

Study Sample size Higher-protein Outcomes relevant

Reference type and setting Population Lower-protein group group for this review Effect size Funding Comments

Turck et al., 2006 RCT n = 162, France Healthy term n = 51; fed exclusively iso- n = 50; fed exclusively a Primary outcome: daily Mean daily weight gain in The isocaloric formulas Breastfed, reference group
(27) infants caloric whey-predomi- conventional casein- weight gain between the formula-fed groups were supplied by Nestle (n = 55)
nant (70:30 whey:casein predominant (70:30 days 0 and 120 (non- differed by 0.38 g/d France
ratio) study formula (pro- casein:whey ratio) for- inferiority criterion: dif- (95% CI: 22.59, 1.83),
tein:energy ratio: 1.8 g/ mula (protein:energy ference in daily weight signifying the noninfer-
100 kcal; 1.2 g/100 mL) ratio: 2.6 g/100 kcal; gain #4 g); iority of the study for-
for 120 d 1.7 g/100 mL) mula**
Age: ,7 d Secondary outcomes: NS
daily gain in weight,
length, head circumfer-
ence, and BMI at
monthly intervals
CHOP study (19, RCT n = 1138, multi- Healthy, term n = 540; 1) infant formula n = 550; 1) infant formula Mean weight and z scores Significantly higher in Supported by EU grants Breastfed group created;
30, 31, 32, *) center infants with 1.25 g protein/100 with a protein content at 3, 6, 12, 24, and 72 higher protein formula (5th and 6th FP); the additional analyses in
European mL (1.77 g/100 kcal) and of 2.05 g/100 mL (2.9 g/ mo of age group only at 6 and 12 International Danone subsample of infants
study energy 69.9 g/100 mL; 100 kcal); energy 69.8 mo of age Institute; Germany: the done (weight gain,
(Poland, Italy, 2) follow-on formula with g/100 mL; Child Health Foundation; body composition,
Germany, a protein content of 1.6 g/ 2) follow-on formula LMU innovative research echocardiography, and
Spain, 100 mL (2.2 g/100 kcal) with a protein content priority project; and kidney volume)
and Belgium) Age: from birth and energy content 72.7 g/ of 3.2 g/100 mL (4.4 g/ Mean length and z scores NS Federal Ministry of All formulas, whey:cas-
to 8 wk 100 mL; after introduction 100 kcal) and energy at 3, 6, 12, 24, and 72 Education and Research sein ratio 1:4
of complementary feeding content 72.7 g/100 mL; mo of age
(not before start of month after introduction of Mean head circumference Not reported
5); up to 12 mo of age complementary feeding and z scores at 6, 12,
Protein concentration of infant formulas

(not before start of and 24 mo of age

month 5); up to 12 mo Mean BMI and BMI z Significantly higher in
of age scores at 3, 6, 12, 24, higher-protein formula
and 72 mo of age group from 12 mo of age
Prevalence of obesity at 6 Significantly greater in
y of age higher-protein formula
group
Insulin and IGF-I axis; Higher protein intake
other biochemical stimulates the IGF-I
markers at 6 mo of age axis and insulin release
1
For NS, difference between the groups, P . 0.05. *Unpublished data; **Results reported by the authors without providing numerical data and statistical significance. BF, breastfed; CHOP, EU Childhood Obesity Programme; ESPGHAN, European
Society for Pediatric Gastroenterology Hepatology and Nutrition; EU, European Union; FP, framework programme; IGF-I, insulin-like growth factor I; LMU, Ludwig Maximilians University; RCT, randomized controlled trial.
557

Downloaded from jn.nutrition.org by guest on March 8, 2017

 TABLE 3 Results of individual studies (not included in a meta-analysis) evaluating the effects of varying
protein concentrations in infant formulas on growth markers and BMI, assessed at different ages of
children1

Lower Higher Statistical Effect size

Outcome, reference, and age of children protein, n protein, n method (95% CI)

Obesity
Weber et al., 2014 (30), 6 y 227 221 RD 20.06 (20.10, 20.01)
Mean BMI, kg/m2
CHOP study (*), 3 mo 454 451 MD 20.07 (20.25, 0.10)
Turck et al., 2006 (27), 4 mo 51 50 MD 0.24 (20.24, 0.72)
CHOP study (*), 6 mo 419 417 MD 20.29 (20.49, 20.09)
CHOP study (*), 12 mo 374 374 MD 20.33 (20.55, 20.11)
Koletzko et al., 2009 (19), 24 mo 313 322 MD 20.30 (20.49, 20.11)
Weber et al., 2014 (30), 6 y 227 221 MD 20.50 (20.89, 20.11)
BMI z score
CHOP study (*), 3 mo 454 451 MD 20.04 (20.16, 0.08)
CHOP study (*), 6 mo 419 417 MD 0.11 (20.02, 0.24)
CHOP study (*), 12 mo 374 374 MD 20.23 (20.38, 20.08)
Koletzko et al., 2009 (19), 24 mo 313 322 MD 20.21 (20.35, 20.07)
Weber et al., 2014 (30), 6 y 227 221 MD 20.30 (20.52, 20.08)
Mean weight, kg
Zoppi et al., 1978 (29), 5.5 mo 6 7 MD 20.96 (21.71, 20.22)

Downloaded from jn.nutrition.org by guest on March 8, 2017

CHOP study (*), 6 mo 419 417 MD 20.16 (20.27, 20.04)
Zoppi et al., 1978 (29), 10 mo 9 7 MD 20.28 (21.59, 1.03)
CHOP study (*), 12 mo 374 374 MD 20.24 (20.40, 20.09)
Koletzko et al., 2009 (19), 24 mo 313 322 MD 20.18 (20.40, 0.04)
Weber et al., 2014 (30), 6 y 227 221 MD 20.59 (21.31, 0.13)
Mean weight z score
CHOP study (*), 3 mo 454 451 MD 20.09 (20.19, 0.02)
CHOP study (*), 6 mo 419 417 MD 20.16 (20.28, 20.04)
CHOP study (*), 12 mo 374 374 MD 20.20 (20.33, 20.07)
Koletzko et al., 2009 (19), 24 mo 313 322 MD 20.10 (20.24, 0.04)
Weber et al., 2014 (30), 6 y 227 221 MD 20.15 (20.36, 0.06)
Mean weight gain, g/d
R€aih€a et al., 1986 (24), 2 wk to 3 mo 10 10 MD 21.57 (27.98, 4.84)
Turck et al., 2006 (27), 0–4 mo 36 38 MD 0.40 (21.72, 2.52)
Zoppi et al., 1978 (29), 0–5.5 mo 6 7 MD 22.40 (26.07, 1.27)
Zoppi et al., 1978 (29), 0–10 mo 9 7 MD 21.90 (26.41, 2.61)
Akeson et al., 1998 (20), 4–8 mo 51 23 MD 0.92 (20.99, 2.83)
Akeson et al., 1998 (20), 8–12 mo 51 23 MD 20.79 (22.18, 0.60)
Mean weight gain, g/(kg
d)
Akeson et al., 1998 (20), 4–8 mo 51 23 MD 0.17 (20.18, 0.52)
Akeson et al., 1998 (20), 8–12 mo 51 23 MD 20.09 (20.25, 0.07)
Mean length:height, cm
CHOP study (*), 6 mo 419 417 MD 20.10 (20.40, 0.20)
CHOP study (*), 12 mo 374 374 MD 20.20 (20.58, 0.18)
Koletzko et al., 2009 (19), 24 mo 313 322 MD 0.20 (20.29, 0.69)
Weber et al., 2014 (30), 6 y 227 221 MD 0.39 (20.47, 1.25)
z score for absolute length:height
CHOP study (*), 3 mo 454 451 MD 20.09 (20.22, 0.04)
CHOP study (*), 6 mo 419 417 MD 20.03 (20.16, 0.10)
CHOP study (*), 12 mo 374 374 MD 20.08 (20.23, 0.07)
Koletzko et al., 2009 (19), 24 mo 313 322 MD 0.07 (20.08, 0.22)
Weber et al., 2014 (30), 6 y 227 221 MD 0.07 (20.10, 0.24)
Mean length gain, cm/mo
Zoppi et al., 1978 (29), 0–5.5 mo 6 7 MD 20.10 (20.48, 0.28)
Zoppi et al., 1978 (29), 0–10 mo 9 7 MD 0.10 (20.34, 0.54)
Mean length gain, mm/(m
d)
Akeson et al., 1998 (20), 4–8 mo 51 23 MD 0.00 (20.07, 0.07)
Akeson et al., 1998 (20), 8–12 mo 51 23 MD 20.02 (20.08, 0.04)

(Continued)

558 Patro-Go1a˛b et al.

 TABLE 3 Continued

Lower Higher Statistical Effect size

Outcome, reference, and age of children protein, n protein, n method (95% CI)

Mean length gain, mm/d

Akeson et al., 1998 (20), 4–8 mo 51 23 MD 20.01 (-0.05, 0.03)
Akeson et al., 1998 (20), 8–12 mo 51 23 MD 20.03 (20.07, 0.01)
Mean crown–rump length gain, cm/wk
R€aih€a et al., 1986 (24), 2 wk to 3 mo 10 10 MD 20.09 (20.19, 0.01)
Mean crown–heel length gain, cm/wk
R€aih€a et al., 1986 (24), 2 wk to 3 mo 10 10 MD 20.06 (20.17, 0.05)
Weight-for-length z score
CHOP study (*), 3 mo 454 451 MD 20.02 (20.15, 0.11)
CHOP study (*), 6 mo 419 417 MD 20.19 (20.32, 20.06)
CHOP study (*), 12 mo 374 374 MD 20.23 (20.38, 20.08)
Koletzko et al., 2009 (19), 24 mo 313 322 MD 20.19 (20.33, 20.05)
1
*Unpublished data. CHOP, EU Childhood Obesity Programme; MD, mean difference; RD, risk difference.

The risk of bias in included studies results for all study groups. However, no numeral data were
The included studies are described with respect to their risk of available.
bias in Supplemental Table 3. Methodologic limitations
included a very small sample size, participantsÕ replacement, Anthropometric growth markers.

Downloaded from jn.nutrition.org by guest on March 8, 2017

pseudo-random allocation, a high loss to follow-up, per-
protocol analysis, and selective outcomes reporting. In many
studies, because of poor reporting, unambiguous evaluation of
the risk of bias was not possible, which often resulted in the
description of some elements as unclear.
Effects of interventions—primary outcomes
Overweight and obesity. Only one RCT (30) (the CHOP
study; n = 448) reported this outcome (Table 3). The risk of
becoming obese at the age of 6 y was significantly lower in the
lower-protein formula group than in the higher-protein formula
group (RR 0.44; 95% CI: 0.21, 0.91).
Fat mass and fat-free mass. The analysis of body composition
by isotope dilution was performed in a subsample (n = 52) of 6-
mo-old infants who were enrolled in the CHOP study (31).
Although correlations between the fat-mass z score (but not the
fat-free mass z score) and both growth velocity and BMI were
observed, fat-free mass and fat mass z scores did not differ
between the higher- and lower-protein formula groups. In the
study by Lonnerdal and Chen (26), skin fold measurements were
obtained at 5, 6, and 7 mo of age. The authors reported similar
BMI. This outcome was reported at different time points in 4
publications (19, 27, 30, 32) (Table 3). At 3 or 4 mo of age, there
were no significant differences between the study groups.
Starting at 6 mo of age, only one RCT (CHOP Study) reported
data on BMI. BMI was significantly lower in the lower-protein
formula group than in the higher-protein formula group at all
time points (i.e., at 6, 12, and 24 mo and at 6 y).
BMI z scores. Only the CHOP study (unpublished data)
reported BMI z scores (Table 3). No significant differences were
found at 3 and 6 mo of age between the lower-protein formula
group and the higher-protein formula group. Starting at 12 mo
of age, BMI z scores were significantly lower in the lower-protein
formula group than in the higher-protein formula group at all
time points (i.e., at 12 and 24 mo and at 6 y).
Mean weight. Twelve publications (18–20, 22–26, 29–32)
reported mean weight at different time points (Figure 1 and
Table 3). The pooled results of 4 RCTs (18, 19, 22, 24) showed
no significant difference between the groups consuming lower-
protein and higher-protein formula at 3 mo of age (n = 998; MD,
–0.07 kg; 95% CI: 20.15, 0.01). In one small (n = 13) trial by
Zoppi et al. (29), mean weight was significantly lower in infants
fed formula with a lower-protein concentration than in those fed

FIGURE 1 Forest plot of the effects of varying protein concentrations in infant formulas on the mean weight of infants at 3 mo of age.
*Unpublished data; ^combined results from 2 study arms. CHOP, EU Childhood Obesity Programme; F2.2, study group with formula
protein concentration of 2.2 g/100 mL; F11, study group with formula protein concentration of 11.2 g/L; F13, study group with formula protein
concentration of 13.3 g/L; IV, inverse variance.

Protein concentration of infant formulas 559

FIGURE 2 Forest plot of the effects of varying protein concentrations in infant formulas on the mean weight of infants from 3 to 6 mo and from
6 to 12 mo of age. ^Combined results from 2 study arms. F13, study group with formula protein concentration of 13 g/L; F15, study group with
formula protein concentration of 15 g/L; IV, inverse variance.

formula with a higher-protein concentration at 5.5 mo of age; meta-analysis, no statistically significant difference between the
however, the difference between the groups was no longer groups consuming lower-protein and higher-protein formulas
significant when these infants were 10 mo old. The CHOP study was found in infants up to 12 mo of age.

Downloaded from jn.nutrition.org by guest on March 8, 2017

reported a significantly lower mean weight in the group consum-
ing lower-protein formula than in the group consuming higher-
protein formula at 6 and 12 mo of age, but there was no significant
difference between the groups at 24 mo and at 6 y.
In one study (23), this outcome was assessed at different child
ages in the groups consuming lower-protein and higher-protein
formula. Therefore, we found it improper to compare the results.
Mean weight z scores. Only the CHOP study (unpub-
lished data) reported mean weight z scores (Table 3). Mean
weight z scores were significantly lower at 6 and 12 mo of age
in the group consuming the lower-protein formula than in the
group consuming the higher-protein formula; however, there
were no significant differences between the groups at 3 mo,
24 mo, and 6 y of age.
Mean weight gain. The authors of almost all of the studies
[except for the studies by Szajewska et al. (18), Picone et al. (22),
and the CHOP study, unpublished data] reported gains in weight
in a variety of ways (e.g., different units or different time
intervals) (Figure 2 and Table 3). However, in almost all of the
comparisons (regardless of the time intervals assessed), as well as
when the results of some studies were pooled together into a
Mean length/height. Mean length/height (absolute mean
length/height or the z score for absolute length/height) was
reported in 7 RCTs (18–20, 22, 24–26) (Figure 3 and Table 3).
The pooled results of 4 RCTs (18, 19, 22, 24) revealed a
significantly shorter length at 3 mo of age in the group consuming
a lower-protein formula than in the group consuming the higher-
protein formula (n = 998; MD, –0.27 cm; 95% CI: 20.52,
20.02). No significant differences between groups were found at
any other time points (i.e., at 6 mo, 12 mo, and 24 mo, and 6 y
of age). Three RCTs (20, 25, 26) provided only descriptive
results (concluding that length was similar in all groups).
Length gain. Length gain up to 12 mo of age was assessed
in 10 RCTs (20–29) (Figure 4 and Table 3). Different units and
different approaches to length measurements (i.e., length,
crown–heel length, and crown–rump length) were used to
express this outcome. The majority of available results showed a
similar length gain in infants regardless of the protein concen-
tration in the consumed formula.
In 2 studies performed by Lonnerdal and Chen (25, 26), it
was not clearly stated whether the authors described a similar
mean weight and length or weight and length gain.

FIGURE 3 Forest plot of the effects of varying protein concentrations in infant formulas on the mean length/height of infants at 3 mo of age.
*Unpublished data; ^combined results from 2 study arms. CHOP, EU Childhood Obesity Programme; F2.2, study group with formula
protein concentration of 2.2 g/100 mL; F11, study group with formula protein concentration of 11.2 g/L; F13, study group with formula protein
concentration of 13.3 g/L; IV, inverse variance.

560 Patro-Go1a˛b et al.

FIGURE 4 Forest plot of the effects of varying protein concentrations in infant formulas on the mean length gain of infants from 3 to 6 mo and
from 6 to 12 mo of age. ^Combined results from 2 study arms. F13, study group with formula protein concentration of 13 g/L; F15, study group
with formula protein concentration of 15 g/L; IV, inverse variance.

Head circumference. In 5 RCTs (18, 22, 24–26), data gain at monthly intervals; etc.). However, we found the results of
regarding head circumference were available (Figure 5). The these trials to be irrelevant to our review.
pooled results of 3 RCTs (18, 22, 24) showed no significant

Downloaded from jn.nutrition.org by guest on March 8, 2017

difference in head circumference between the groups consuming
lower-protein and higher-protein formula at 12 wk of age (MD,
20.31 cm; 95% CI: 20.63, 0.02). However, significant heteroge-
neity was found (I2 = 81%). The observed heterogeneity was not
explained by the subgroup analysis based on the amount of protein
in milk formulas, which was performed as originally planned in
the protocol. We also considered different measurement techniques
as a potential source of the identified heterogeneity. However, the
authors did not report how the measurements were obtained. In
2 studies by Lonnerdal and Chen (25, 26), the authors reported a
similar mean head circumference in all study groups (numerical
data were not shown).
Head circumference gain. In the study by Turck et al. (27),
the daily gain in head circumference (millimeters per day) at
monthly intervals was described, showing a similar growth rate in
both study groups. In the study by Zoppi et al. (23), the mean
increase in head circumference was evaluated in the group consum-
ing higher-protein formula only. In another report (28), the mean
daily gain in head circumference was described as normal.
Weight-for-length z scores. Weight-for-length z scores
were analyzed in the CHOP study [(unpublished data) Table 3].
The weight-for-length z scores were significantly higher at 6, 12,
and 24 mo of age in the group consuming higher-protein formula
than in the group consuming lower-protein formula.
Other. Other anthropometric assessments were reported in
some of the included trials (particularly an early assessment of
growth at 2, 4, and 8 wk of age; evaluation of weight and length
Secondary outcomes
Our secondary outcomes were only partially addressed in 2
studies. A summary of these findings is presented in Table 4.
Discussion
Summary of the evidence. This review revealed that the vast
majority of identified studies evaluated only short-term out-
comes related to infant growth. The gathered evidence suggests
that during the first year of life, different protein concentrations
in infant formulas do not seem to affect infantsÕ linear growth
significantly apart from a transient effect on mean length during
the first months of life. Data with respect to weight and weight
gain suggest similar growth in the first months of life, with a
lower mean weight (and weight z score) obtained in the infants
fed lower-protein formulas from ;6 mo until 12 mo of age only.
A significantly lower BMI associated with the intake of lower-
protein formula was firmly observed in 12-mo-old infants and in
children up until early school age. No conclusions with regard to
the effects of formula protein concentration on body composi-
tion can be formulated. We have identified only one RCT (the
CHOP study) that assessed the outcomes of interest after
infancy. The intake of lower-protein formula during infancy was
observed to reduce the risk of obesity at the age of 6 y in this
study. However, these findings need to be interpreted with
caution, because this is a single report, and some methodologic
limitations of the study were identified.

FIGURE 5 Forest plot of the effects of varying protein concentrations in infant formulas on the mean head circumference of infants at 3 mo of
age. ^Combined results from 2 study arms. F2.2, study group with formula protein concentration of 2.2 g/100 mL; F11, study group with formula
protein concentration of 11.2 g/L; F13, study group with formula protein concentration of 13.3 g/L; IV, inverse variance.

Protein concentration of infant formulas 561

Strengths and limitations. The major strength of our review is
that it collates the largest number of studies, to our knowledge,
available on the effects of varying protein concentrations in
infant formulas on growth, body composition, and the later risk
of obesity. We used a rigorous systematic review methodology
proposed by the Cochrane Collaboration. We used several
methods to reduce bias from the initial preparatory stage
until the final report (i.e., protocol registration, comprehen-
sive literature search, duplicate data abstraction, and prespecified
criteria for methodologic assessment and analysis). We did not
impose restrictions by language or year of publication. Attempts
were made to identify unpublished trials. However, this review has
some limitations. The methodologic quality of the included trials
varied. Potential limitations include unclear sequence generation,
unclear allocation concealment and blinding, small sample
sizes, lack of sample size calculations, and high dropout rates in
some trials, including the CHOP trial. Despite the fact that a
number of trials met the inclusion criteria, our conclusions after
the infancy period are based solely on the results of one RCT (the
only experimental study with a large sample size and longer
follow-up). It is worth mentioning that the sample sizes of the
included studies varied, and that sample size in the CHOP study
was greater than the total number of participants of all other
included studies. We observed a great deal of heterogeneity in
the protein concentrations in the formulas used in different
studies, with the protein amount being the same in groups
consuming lower- and higher-protein formula, depending on the
study in some cases. Therefore, our review cannot provide an
answer to the question of what the particular and optimal
amount of protein to be included in infant formulas should be.
Another important aspect raised by our review is that not
only protein quantity, but also protein quality, must be taken
into account to secure a sufficient amount of essential amino
acids to support adequate growth and safety in infants. Stud-
ies that use the Indicator Amino Acid Oxidation method to
determine the actual requirement of each essential amino acid in
neonates have shown that infant formulas that are currently
used may not contain the optimal amino acid composition and
may provide an amount of essential amino acids that is too high
(35–37). However, a detailed analysis of the amino acid content
of formula was outside the scope of our review. Also, we have
excluded from our analysis a number of studies in which
additional differences between the groups were present, such as
enriched milk formula (with a-lactalbumin, for example) or
formula with extra components added (see exclusion criteria),
apart from quantitative protein modifications, differences in
the whey-to-casein ratio, and other differences in composition
TABLE 4 Effects of different protein concentrations in infant formulas on lipid profile, IGF-I axis, and kidney size assessed during
infancy1
Reference Outcome
Zoppi et al. (23) Serum cholesterol and TG concentrations at 5 mo of age
CHOP (32) IGF-I axis
CHOP (33, 34) Kidney size (at 6 mo of age)
1
CHOP, EU Childhood Obesity Programme; IGF-I insulin-like growth factor I.
562 Patro-Go1a˛b et al.
secondary to the protein amount changes (in order to achieve an
isocaloric formula). These restrictive criteria were developed to
reduce the great heterogeneity between the included studies.
However, we realize that these data may be a substantial
addition to this complex issue. Finally, we analyzed only one
aspect of protein intake provided by milk formula consumption,
whereas other sources of protein in the diet also play an
important role after the introduction of complementary feeding.
Agreement and disagreement with other studies or
reviews. The findings of this review are in accordance with 2
recently published reviews. The first review by Hornell et al.
(38), published in 2013, focused on protein intake in the diet
(not only in milk formula) of children of different ages (0–18 y)
and its relation to health based on both experimental and
observational studies. The authors concluded that higher
protein intake during infancy is associated with increased
growth, a higher BMI during childhood, and an increased risk
of being overweight later in life. The second review by Abrams
et al. (39), published when this manuscript was in the final stages
of preparation, addressed the effect of the intake of low-protein
and low-energy formula by full-term infants on growth. Six
publications were included in that review. Based on the findings
Downloaded from jn.nutrition.org by guest on March 8, 2017
from 4 RCTs, the authors concluded that the use of infant
formulas with concentrations of protein and energy slightly
below historical standards in the United States results in
adequate growth during infancy and early childhood. The
authors also stated that ‘‘Further long-term research is needed to
assess the impact of the use of lower-protein and/or lower-energy
products, especially for nutritionally at-risk populations such as
preterm infants and infants who are born small for
gestational age’’ (39). Although the authors asked a similar
question to the one that is the subject of our review, their
approach differed. The search was limited to 3 electronic
databases only (i.e., Medline, Cochrane, and the Cumulative
Index to Nursing and Allied Health Literature). Furthermore,
the set of key words used for searching differed, as well as the
inclusion and exclusion criteria, resulting in the differences
observed in the number of identified trials (4 studies in the
review by Abrams et al. (39) compared with 12 trials identified
in this review). This was not unexpected, because it is well
known that small differences in search strategies may result in
substantial differences in the set of trials identified. Nevertheless,
taken together, all 3 reviews clearly prove how complex the issue
of protein intake can be and leave room for more evidence and
other systematic analyses to answer the questions that remain
unanswered.
Results
Observed values largely within normal range
Concentrations reflected amount of protein consumed
No statistically significant difference between the study
groups observed
Higher protein intake stimulates IGF-I axis and insulin release
Significantly increased kidney size in infants who consumed
a higher-protein formula
IGF-I in part mediates protein-induced kidney growth
Effect of higher protein intake during early infancy on long-term
kidney function requires further evaluation
 The evidence is insufficient for assessing the effects of
reducing protein concentration in infant formulas on long-
term outcomes, but, if confirmed, this could represent a
promising intervention for reducing the risk of overweight and
obesity in children. In view of the limited available evidence,
more studies replicating the effects on long-term health out-
comes are needed. Careful analysis of observational studies that
have assessed long-term outcomes of interest may provide a
valuable addition to the limited experimental evidence.
Acknowledgments
BK, JBvG, and HS initially conceptualized this study; BP-G,
BMZ, SMPK, and JK were responsible for data collection, data
analysis, data interpretation, and preparation of the report; BP-G
assumed primary responsibility for the writing of the first draft of
the manuscript; and all authors contributed to the initial protocol
of the study. All authors read and approved the final manuscript.
References
1. World Health Organization. Protein and amino acid requirements in
human nutrition. [Technical Report Series, 2002 935:265.]
2. Inostroza J, Haschke F, Steenhout P, Grathwohl D, Nelson S, Ziegler E.
Low-protein formula slows weight gain in infants of overweight
mothers. J Pediatr Gastroenterol Nutr 2014;59:70–7.
3. Toschke AM, Grote V, Koletzko B, von Kries R. Identifying children at
high risk for overweight at school entry by weight gain during the first 2
years. Arch Pediatr Adolesc Med 2004;158:449–52.
4. Brands B, Demmelmair H, Koletzko B. EarlyNutrition Project. How
growth due to infant nutrition influences obesity and later disease risk.
Acta Paediatr 2014;103:578–85.
5. Koletzko B. Early nutrition and its later consequences: new opportu-
nities. In: Koletzko B, Akerblom H, Dodds PF, Ashwell M (eds). Early
nutrition and its later consequences: new opportunities. Adv Exp Med
Biol 2005;569:1–12.
6. Rolland-Cachera MF, Deheeger M, Akrout M, Bellisle F. Influence of
macronutrients on adiposity development: a follow up study of
nutrition and growth from 10 months to 8 years of age. Int J Obes
Relat Metab Disord 1995;19:573–8.
7. Kirchberg FF, Harder U, Weber M, Grote V, Demmelmair H, Peissner
W, Rzehak P, Xhonneux A, Carlier C, Ferre N, et al.; European
Childhood Obesity Trial Study Group. Dietary protein intake affects
amino acid and acylcarnitine metabolism in infants aged 6 months. J
Clin Endocrinol Metab 2015;100:149–58.
8. Tijhuis MJ, Doets EL, Noordegraaf-Schouten V. Extensive literature
search and review as preparatory work for the evaluation of the
essential composition of infant and follow-on formulae and growing-up
milk [Internet]. [cited 2015 Apr 20.] Available from: www.efsa.europa.
eu/en/supporting/doc/551e.pdf.
9. Koletzko B. Standards for infant formula milk. Commercial inter-
ests may be the strongest driver of what goes into formula milk. BMJ
2006;332:621–2.
10. Koletzko B, Baker S, Cleghorn G, Neto UF, Gopalan S, Hernell O, Hock
QS, Jirapinyo P, Lonnerdal B, Pencharz P, et al. Global standard for the
composition of infant formula: recommendations of an ESPGHAN
coordinated international expert group. J Pediatr Gastroenterol Nutr
2005;41:584–99.
11. European Union Commission Directive 2006/141/EC. Foods for infants
& young children - Infant formulae and follow-on formulae. [Internet].
[cited 2015 Mar 20.] Available from: ec.europa.eu/food/food/label-
lingnutrition/children/formulae_en.htm.
12. US Food and Drug Administration. Current manufacturing regulations
and guidelines. 21 CFR \107.100 Nutrient requirements for infant
formulas [Internet.] [cited 2015 Mar 20.] Available from: www.fda.gov/
Food/GuidanceRegulation/GuidanceDocumentsRegulatoryInformation/
InfantFormula/default.htm.
13. European Union Commission Directive 2006/141/EC. Infant formulae
and follow-on formulae. Official Journal of the European Communi-
ties [Internet]. [cited 2015 Mar 20.] Available from: eur-lex.europa.eu/
LexUriServ/LexUriServ.do?uri=OJ%3AL%3A2006%3A401%3A0001%
3A0033%3AEN%3APDF.
14. US Food and Drug Administration. Food and Drug Administration rules
and regulations. Nutrient requirements for infant formulas (21 CFR,
Part 107). Code of Federal Regulations. Title 21, Volume 2. Revised as
of April 1, 2014. 21CFR107.100.
15. Codex Alimentarius Commission. Standard for infant formula
and formulas for special medical purposes intended for infants
(Codex Stan 72–1981) [Internet]. [cited 2015 Mar 20.] Available
from: www.codexalimentarius.org/input/download/standards/288/
CXS_072e.pdf.
16. European Food Safety Agency, Panel on Dietetic Products, Nutrition
and Allergies. Scientific opinion on the essential composition of infant
and follow-on formulae. EFSA Journal. 2014;12:3760.
17. Higgins JPT, Green S. (eds)Cochrane Handbook for systematic reviews of
interventions version 5.1.0 [Internet]. The Cochrane Collaboration,
2011. [cited 2014 Jul 21.] Available from: www.cochrane-handbook.org.
18. Szajewska H, Albrecht P, Krawczyk A. The effect of various intakes on
growth and protein metabolizm of term infants during the first three
months of life (part I). Pediatr Pol 1993;LXVIII:11.
19. Koletzko B, von Kries R, Closa R, Escribano J, Scaglioni S, Giovannini
M, Beyer J, Demmelmair H, Gruszfeld D, Dobrzanska A, et al.;
European Childhood Obesity Trial Study Group. Lower protein in
infant formula is associated with lower weight up to age 2 y: a
randomized clinical trial. Am J Clin Nutr 2009;89:1836–45.
20. Akeson PM, Axelsson IE, Räihä NC. Growth and nutrient intake in
three- to twelve-month-old infants fed human milk or formulas with
varying protein concentrations. J Pediatr Gastroenterol Nutr
Downloaded from jn.nutrition.org by guest on March 8, 2017
1998;26:1–8.
21. Karlsland Akeson PK, Axelsson IE, Räihä NC, Warm A, Minoli I, Moro
G. Protein intake and metabolism in formula-fed infants given Swedish
or Italian weaning foods. Acta Paediatr 2000;89:158–64.
22. Picone TA, Benson JD, Moro G, Minoli I, Fulconis F, Rassin DK, Raiha
NC. Growth, serum biochemistries, and amino acids of term infants fed
formulas with amino acid and protein concentrations similar to human
milk. J Pediatr Gastroenterol Nutr 1989;9:351–60.
23. Zoppi G, Gerosa F, Pezzini A, Bassani N, Rizzotti P, Bellini P, Todeschini
G, Zamboni G, Vazzoler G, Tridente G. Immunocompetence and
dietary protein intake in early infancy. J Pediatr Gastroenterol Nutr
1982;1:175–82.
24. Räihä N, Minoli I, Moro G. Milk protein intake in the term infant. I.
Metabolic responses and effects on growth. Acta Paediatr Scand
1986;75:881–6.
25. Lönnerdal B, Chen CL. Effects of formula protein level and ratio on
infant growth, plasma amino acids and serum trace elements. I. CowÕs
milk formula. Acta Paediatr Scand 1990;79:257–65.
26. Lönnerdal B, Chen CL. Effects of formula protein level and ratio on
infant growth, plasma amino acids and serum trace elements. II. Follow-
up formula. Acta Paediatr Scand 1990;79:266–73.
27. Turck D, Grillon C, Lachambre E, Robiliard P, Beck L, Maurin JL,
Kempf C, Bernet JP, Marx J, Lebrun F, et al. Adequacy and safety of an
infant formula with a protein/energy ratio of 1.8 g/100 kcal and
enhanced protein efficiency for term infants during the first 4 months of
life. J Pediatr Gastroenterol Nutr 2006;43:364–71.
28. Zoppi G, Gasparini R, Mantovanelli F, Gobio-Casali L, Astolfi R,
Crovari P. Diet and antibody response to vaccinations in healthy infants.
Lancet 1983;2:11–4.
29. Zoppi G, Zamboni G, Siviero M, Bellini P, Cancellieri ML. Gamma-
Globulin level and dietary protein intake during the first year of life.
Pediatrics 1978;62:1010–8.
30. Weber M, Grote V, Closa-Monasterolo R, Escribano J, Langhendries JP,
Dain E, Giovannini M, Verduci E, Gruszfeld D, Socha P., Koletzko for
The European Childhood Obesity Trial Study Group. Lower protein in
infant formula reduces BMI and obesity risk at school age: follow-up of
a randomized trial. Am J Clin Nutr 2014;99:1041–51.
31. Escribano J, Luque V, Ferre N, Mendez-Riera G, Koletzko B, Grote V,
Demmelmair H, Bluck L, Wright A, Closa-Monasterolo R.; European
Childhood Obesity Trial Study Group. Effect of protein intake and
weight gain velocity on body fat mass at 6 months of age: the EU
Childhood Obesity Programme. Int J Obes (Lond) 2012;36:548–53.
32. Socha P, Grote V, Gruszfeld D, Janas R, Demmelmair H, Closa-
Monasterolo R, Subı́as JE, Scaglioni S, Verduci E, Dain E, et al.;
European Childhood Obesity Trial Study Group. Milk protein intake, the
metabolic-endocrine response, and growth in infancy: data from a
randomized clinical trial. Am J Clin Nutr 2011; 94(6, Suppl)1776S–84S.
Protein concentration of infant formulas 563
33. Luque V, Escribano J, Grote V, Ferre N, Koletzko B, Gruszfeld D,
Socha P, Langhendries JP, Goyens P, Closa-Monasterolo R.; Euro-
pean Childhood Obesity Project. Does insulin-like growth factor-
1 mediate protein-induced kidney growth in infants? A secondary
analysis from a randomized controlled trial. Pediatr Res 2013;74:
223–9.
34. Escribano J, Luque V, Ferre N, Zaragoza-Jordana M, Grote V, Koletzko
B, Gruszfeld D, Socha P, Dain E, Van Hees JN, et al. Increased protein
intake augments kidney volume and function in healthy infants. Kidney
Int 2011;79:783–90.
35. de Groof F, Huang L, van Vliet I, Voortman GJ, Schierbeek H,
Roksnoer LC, Vermes A, Chen C, Huang Y, van Goudoever JB.
Branched-chain amino acid requirements for enterally fed term neonates
in the first month of life. Am J Clin Nutr 2014;99:62–70.
564 Patro-Go1a˛b et al.
36. Huang L, Hogewind-Schoonenboom JE, de Groof F, Twisk JW,
Voortman GJ, Dorst K, Schierbeek H, Boehm G, Huang Y, Chen C,
et al. Lysine requirement of the enterally fed term infant in the first
month of life. Am J Clin Nutr 2011;94:1496–503.
37. Huang L, Hogewind-Schoonenboom JE, van Dongen MJ, de Groof F,
Voortman GJ, Schierbeek H, Twisk JW, Vermes A, Chen C, Huang Y, et al.
Methionine requirement of the enterally fed term infant in the first month
of life in the presence of cysteine. Am J Clin Nutr 2012;95:1048–54.
38. Hörnell A, Lagström H, Lande B, Thorsdottir I. Protein intake from 0 to
18 years of age and its relation to health: a systematic literature review for
the 5th Nordic Nutrition Recommendations. Food Nutr Res 2013;57:.
39. Abrams SA, Hawthorne KM, Pammi M. A systematic review of
controlled trials of lower-protein or energy-containing infant formulas
for use by healthy full-term infants. Adv Nutr 2015;6:178–88.
Downloaded from jn.nutrition.org by guest on March 8, 2017