ARTICLE IN PRESS

THE JOURNAL OF PEDIATRICS • www.jpeds.com ORIGINAL

ARTICLES
Improved Outcomes in Preterm Infants Fed a Nonacidified Liquid Human
Milk Fortifier: A Prospective Randomized Clinical Trial
Richard J. Schanler, MD1, Sharon L. Groh-Wargo, PhD, RDN2, Bridget Barrett-Reis, PhD, RDN3, Robert D. White, MD4,
Kaashif A. Ahmad, MD5, Jeffery Oliver, MS3, Geraldine Baggs, PhD3, Larry Williams, MD3, and David Adamkin, MD6

Objective To compare growth, feeding tolerance, and clinical and biochemical evaluations in human milk-fed preterm
infants randomized to receive either an acidified or a nonacidified liquid human milk fortifier.
Study design This prospective, controlled, parallel, multicenter growth and tolerance study included 164
preterm infants (≤32 weeks of gestation, birth weight 700-1500 g) who were randomized to acidified or
nonacidified liquid human milk fortifier from study day 1, the first day of fortification, through study day 29 or until
hospital discharge.
Results There was no difference in the primary outcome of weight gain from study days 1 to 29 (acidified liquid
human milk fortifier, 16.4 ± 0.4 g/kg/day; nonacidified liquid human milk fortifier, 16.9 ± 0.4 g/kg/day). However, in
both the intention-to-treat and the protocol evaluable analyses, infants fed nonacidified liquid human milk fortifier
had significantly greater weight gain from study days 1 to 15 (17.9 g/kg/day vs 15.2 g/kg/day; P = .001). Infants
fed with acidified liquid human milk fortifier received more protein (4.26 vs g/kg/day 4.11 g/kg/day, P = .0099) yet
had lower blood urea nitrogen values (P = .010). The group fed acidified liquid human milk fortifier had more vom-
iting (10.3% vs 2.4%; P = .018), gastric residuals (12.8% vs 3.7%; P = .022), and metabolic acidosis (27% vs 5%;
P < .001) in the intention-to-treat analysis and more abdominal distension (14.0% vs 1.7%; P = .015) in the proto-
col evaluable analysis.
Conclusions Infants fed an acidified liquid human milk fortifier had higher rates of metabolic acidosis and poor
feeding tolerance compared with infants fed a nonacidified liquid human milk fortifier. Initial weight gain was poorer
with the acidified liquid human milk fortifier. (J Pediatr 2018;■■:■■-■■).
Trial registration ClinicalTrials.gov: NCT02307760.

egardless of gestational age, preterm infants often grow poorly before their discharge home.1-3 Moreover, poor growth

R in the neonatal intensive care unit increases the risk of long-term growth deficits and developmental delay.4-9 Al-
though the benefits of human milk for preterm infants are well-established, fortification is necessary to provide suf-
ficient nutrients to support growth.10,11 Although single nutrient additions occasionally are used, human milk fortifiers have
greatly simplified the work of the neonatal team. The first human milk fortifiers were powdered formulations that allowed in-
creasing nutrient density without diluting the beneficial properties of human milk. However, the powder manufacturing process
does not allow for a terminally sterile product. In the early 2000s, several cases of infant infection and death from Cronobacter
sakazakii sepsis led the US Food and Drug Administration to ban powdered formula from the neonatal intensive care unit unless
no alternative was available.12-14
The first concentrated liquid human milk fortifier in the US was acidified to achieve sterility.15 This advance was followed
by a nonacidified, aseptically filled product.16 Aseptic filling involves sterilizing
the liquid with heat, then filling the final containers under aseptic conditions.
The difference in these 2 sterilization processes has raised questions about the
bioactive properties of human milk,17 because the addition of the acidified
liquid human milk fortifier to human milk reduces the pH from 7.4 to 4.7,18 as From the 1Neonatal-Perinatal Medicine, Cohen Children’s
Medical Center, New Hyde Park, NY; 2MetroHealth
well as on the physiological effects in infants consuming human milk fortified Medical Center, Case Western Reserve University,
Cleveland; 3Abbott Nutrition, Columbus, OH; 4Pediatrix
with acidified liquid human milk fortifier. Several reports have indicated that Medical Group, Memorial Hospital of South Bend, South
Bend, IN; 5Baylor College of Medicine, Pediatrix Medical
preterm infants fed acidified liquid human milk fortifier have a more acidic Group, North Central Baptist Hospital, San Antonio, TX;
19-21
physiology and issues with tolerance and growth. and 6Department of Pediatrics, University of Louisville,
Louisville, KY
The objective of this multicenter trial was to compare the growth, tolerance, Funded by Abbott Nutrition. R.S., S.W., R.W., K.A. and
and biochemical outcomes in preterm infants fed human milk fortified with an D.A. received research funding from the study sponsor,
Abbott Nutrition, to conduct the study. B.R., J.O., G.B.,
acidified or nonacidified liquid human milk fortifier. We hypothesized that and L.W. are employees of Abbott.
growth, feeding tolerance, and biochemical measures would be adversely Portions of this study were presented at the Advanced
Practice Neonatal Nurses Conference, Sept 14, 2017,
Las Vegas, NV and at the Pediatric Academic Societies
annual meeting, May 5-8, Toronto, Canada.

0022-3476/$ - see front matter. © 2018 Elsevier Inc. All rights

ITT Intention-to-treat reserved.
https://doi.org10.1016/j.jpeds.2018.07.005

FLA 5.5.0 DTD ■ YMPD10104_proof ■ September 5, 2018

THE JOURNAL OF PEDIATRICS • www.jpeds.com
affected in preterm infants receiving an acidified compared
with a nonacidified human milk fortifier.
Methods
This prospective, randomized, multicenter (14 sites), con-
trolled, parallel study (Clinical Trial Registration:
NCT02307760) was conducted from November 2014 to March
2017. It was not possible to blind the study at the research sites
because study products were packaged differently; however, the
central study team remained blinded during the study and was
unblinded only after the study was completed. Human milk-
fed preterm infants (n = 164) were assigned randomly, using
an algorithm implemented in Medidata Rave Balance to receive
either acidified liquid human milk fortifier (Enfamil Human
Milk Fortifier Acidified Liquid, Mead Johnson, Evansville,
Indiana) or nonacidified liquid human milk fortifier (Similac
Human Milk Fortifier Hydrolyzed Protein Concentrated Liquid,
Abbott Nutrition, Columbus, Ohio). Fortification began within
72 hours after the infant reached a human milk intake of 80 mL/
kg/day. The first day of human milk fortification was desig-
nated as study day 1. The suggested enteral intake goal for all
infants was ≥150 mL/kg/day (≥120 kcal/kg/day). Parenteral nu-
trition was permitted during the study if the infant main-
tained an intake of ≥100 mL/kg/day of fortified human milk.
The same nutrition protocol was described for all sites.
The assigned study product added to maternal human milk
was fed as the sole source of nutrition throughout the study
period (study days 1-29) or to hospital discharge, whichever
came first. The nutritional contents of the human milk for-
tifiers and fortified human milks are provided in Table I (avail-
able at www.jpeds.com). Infants were followed for growth
(weight, length, and head circumference), feeding tolerance,
morbidity, biochemical measurements, duration of paren-
teral nutrition, medication and dietary supplement use, and
adverse events. These items were documented to study day 29;
documentation continued if the infant remained on the study
human milk fortifier. After study day 29, the infants were fol-
lowed until hospital discharge, a weight of 3600 g was reached,
or until study human milk fortifier was discontinued.
The primary outcome was rate of weight gain (in grams
per kilogram per day) from study days 1 to 29 or hospital
discharge. Secondary variables included growth during the
interval from study days 1 to 15, rates of weekly gains in length
and head circumference, feeding tolerance (stool characteris-
tics, feedings withheld due to abdominal distention, gastric
residuals, vomiting, and other causes). Supportive variables
included biochemical measures, morbidity and safety out-
comes. Infants who received >1 study feeding were included
in the intention-to-treat (ITT) analyses. A predefined, proto-
col evaluable analysis for a study outcome included data col-
lected from infants who followed the study feeding protocol
up to the time of observation or assessment of the outcome.
For the protocol evaluable data analysis, infants were in-
cluded if they remained in the assigned group and received
<25% of total enteral energy from sources other than the study
fortified milk for >14 days or received protein supplements
2 Schanler et al
FLA 5.5.0 DTD ■ YMPD10104_proof ■ September 5, 2018
Volume ■■ • ■■ 2018
for <7 days during study days 1-29. Because the use of donor
human milk was not proscribed, we used consumption of >50%
donor milk and/or nonstudy fortifier/formula/supplements
for >3 days as further exclusions in a post hoc analysis of the
protocol evaluable group.
The protocol was approved by the human subjects review
board at the respective institutions. Preterm infants expected
to survive with minimal morbidity were eligible for inclu-
sion if the birth weight was 700-1500 g, gestational age was ≤32
weeks, the birth weight was appropriate for gestational age, the
mother agreed to provide human milk, enteral feeding of
human milk was initiated by 21 days after birth, the infant was
a singleton or twin, and parents agreed to allow feeding of
human milk and human milk fortifier and had voluntarily
signed and dated an approved informed consent form. Infants
were not enrolled or were excluded from study if they re-
ceived preterm infant formula for >7 days, were expected to
be transferred to another facility and not able to be followed
for ≥15 days, had serious congenital abnormalities or under-
lying disease that might affect growth and development, had
a 5-minute Apgar score of ≤4, and at the time of randomiza-
tion were receiving glucocorticoids, had grade III or IV intra-
ventricular hemorrhage, were dependent on a mechanical
ventilation, experienced asphyxia or seizures, had a history of
major surgery, had confirmed necrotizing enterocolitis (Bell’s
stage II or III) or confirmed sepsis (positive culture requir-
ing antibiotic treatment), had maternal HIV positivity, were
exposed to cocaine or alcohol in utero, or had any condition
that, in the opinion of the investigator, precluded participa-
tion in the study.
The study sample size was estimated using the standard de-
viation of the primary variable, rate of weight gain (grams per
kilogram per day) from similar study populations.16 A sample
size of 106 infants (53 per group) was needed to detect a dif-
ference in mean weight gain of 1.6 g/kg/day assuming a stan-
dard deviation of 2.89 g/kg/day, using a 0.05 level 2-sided t-test
with 80% power. Based on an estimated 35% attrition rate, ran-
domization of 164 infants was planned (82 per group). The
nQuery Advisor 5.0 software (Statsols, Cork, Ireland) was used
in these determinations.
Weight gain (grams per kilogram per day) for each infant
was calculated by an exponential model that involved a re-
gression line fit on loge (wt), where wt is weight (in grams)
on each day. Owing to excessive variability in the length
and head circumference measurements, the Gompertz
distribution22,23 was used to fit curves for length and head cir-
cumference values collected at birth and weekly through study
day 29, and the resulting fitted values were used to derive length
and head circumference gains for study days 1-29 and for study
days 1-15. Metabolic acidosis was reported during the study,
or met an a priori definition of a base excess of <-6 mmol/L.
If the distribution of a continuous variable was normal,
analysis of variance with factors for feeding group and study
center was used. For growth outcomes, an a priori specified
confirmatory analysis of covariance was performed with factors
for birth weight, sex, feeding group, and study center. If the
distribution was non-normal, nonparametric analyses (such
■■ 2018 ORIGINAL ARTICLES

N = 1354
Screened
561 failed BW or GA criteria
69 failed breast feeding criterion
22 multiple births
218 refused consent
320 met other exclusion criteria.

N = 164
Randomized subjects

n = 81 n = 83
ALHMF NALHMF

n=3 n=1
Never received study fortifier Never received study fortifier

n = 82
n = 78 n = 50 ITT n = 59
ITT PE Analysis Analysis PE analysis
Analysis

n = 37 n = 40
Post Post
n = 67 hoc n = 74 hoc
Completed study Analysis Completed study Analysis

Figure 1. Disposition of study participants. BW, birth weight; GA, gestational age; ITT, intention-to-treat; PE, protocol evaluable;
Post hoc, excluded PE infants consuming >50% donor human milk or nonstudy fortifier/formula/supplements for >3 days.

as Wilcoxon rank-sum tests) were used. Cochran-Mantel- (Table II). In the protocol evaluable analysis, the nonacidified
Haenszel and Fisher exact tests were used for categorical data. liquid human milk fortifier group had lower Apgar scores than
Data for continuous variables are reported as mean ± stan- the acidified liquid human milk fortifier group (P = .039).
dard error unless otherwise noted. Effect sizes are expressed There were no differences in the primary outcome of weight
as the difference of least squares means (acidified minus gain from study days 1-29 between the acidified or nonacidified
nonacidified) ± standard error or as relative risk (95% CI). An liquid human milk fortifier groups in either the ITT analysis
independent data monitoring committee evaluated safety data
at regular intervals. The investigators or sponsors did not par-
ticipate in the closed sessions of the committee.
Table II. Clinical characteristics of study infants
Results Acidified liquid
human milk
Nonacidified
liquid human
Characteristics fortifier milk fortifier
The complete disposition of study subjects in the ITT, proto-
n 78 82
col evaluable, and post hoc analyses is described in Figure 1. Birth weight, g* 1211 ± 22 1184 ± 23
In the acidified liquid human milk fortifier group, 12 infants Birth head circumference, cm 27.0 ± 0.3 26.0 ± 0.2
exited the study and 9 did so in the nonacidified liquid human Birth length, cm 38.0 ± 0.3 38.0 ± 0.3
Gestational age, weeks 29.0 ± 0.2 29.0 ± 0.2
milk fortifier group. Reasons for study exit in the acidified vs Gender, n (% female) 42 (54) 45 (55)
nonacidified liquid human milk fortifier groups, respec- Race, n (%)
tively, were insufficient maternal milk (2 vs 0), parent concern White 53 (68) 51 (62)
Black 20 (26) 26 (32)
(3 vs 0), changed clinical condition of infant (3 vs 3), feeding Asian 1 (1) 4 (5)
intolerance (2 vs 3), and other reasons (2 vs 3). In the ITT analy- Other 4 (5) 1 (1)
sis, there were no significant differences between groups for Apgar score of <7 at 5 min, n (%) 5 (6) 13 (16)
Intraventricular hemorrhage, n (%)† 8 (10) 19 (23)
birth weight, length, head circumference, sex, race, or Apgar Patent ductus arteriosus, n (%) 3 (4) 7 (9)
scores (Table II). The nonacidified liquid human milk forti- Receipt of any donor human milk, n (%) 24 (31) 31 (38)
fier group had more intraventricular hemorrhage than the acidi- *Mean ± standard error of the mean.
fied liquid human milk fortifier group before study day 1 †P = .028 nonacidified human milk fortifier > acidified human milk fortifier.

Improved Outcomes in Preterm Infants Fed a Nonacidified Liquid Human Milk Fortifier: A Prospective 3
Randomized Clinical Trial
FLA 5.5.0 DTD ■ YMPD10104_proof ■ September 5, 2018
THE JOURNAL OF PEDIATRICS • www.jpeds.com Volume ■■

ITT analysis PE analysis

ALHMF ALHMF
25 NALHMF NALHMF 25

20 20
†

% of Infants
% of Infants

*
‡
15 15
**
10 10

5 5

0
0 0

Abdominal Gastric Vomiting Abdominal Gastric Vomiting

distention residuals distention residuals

Figure 4. Percent of infants who had feedings withheld owing to either abdominal distention, gastric residuals, or vomiting. ALHMF,
acidified human milk fortifier; ITT, intention-to-treat; NALHMF, nonacidified human milk fortifier; PE, protocol evaluable. Values
are mean ± standard error of the mean. *Effect size: 2.6 (95% CI, 0.9–7.1), P = .022, **effect size: 3.4 (95% CI, 1.02-11.3),
P = .018, †effect size: 2.7 (95% CI, 0.8-8.7), P = .015, ‡effect size: 4.6 (95% CI, 0.9-23.9), P = .017.

(effect size, -0.6 ± 0.6 g/kg/day) or protocol evaluable analy-
sis (effect size, -0.8 ± 0.5 g/kg/day; Figure 2; available at
www.jpeds.com). However, in the post hoc analysis that ex-
cluded infants from the protocol evaluable analysis who, for
>3 days, received >50% donor milk or nonstudy feedings,
weight gain from study days 1-29 differed significantly between
groups (effect size, -1.5 ± 0.6 g/kg/day; P = .009). Weight gain
from study days 1-15 in the acidified liquid human milk for-
tifier group was significantly lower than the nonacidified liquid
human milk fortifier group (ITT analysis effect size, -2.2 ± 0.8 g/
kg/day [P = .004] and protocol evaluable analysis effect size,
-2.7 ± 0.8 g/kg/day [P = .001]). Variability in effect sizes for
weight gain from study days 1-29 in the ITT analysis are il-
lustrated by study site (Figure 3; available at www.jpeds.com).
There were no differences in either length or head circum-
ference gains from study days 1-29 for the ITT analysis (length
0.97 and 1.00 cm/week [effect size, -0.04 ± 0.07]; head cir-
cumference, 0.85 and 0.84 cm/week [effect size, 0.01 ± 0.04]
in the acidified vs nonacidified liquid human milk fortifier
groups, respectively). Length and head circumference gains were
also similar between groups in the protocol evaluable analysis.
There were differences in tolerance that resulted in feedings
being withheld during the study. Infants fed acidified liquid
human milk fortifier had significantly more vomiting and
gastric residuals leading to withholding of feedings than infants
fed nonacidified liquid human milk fortifier in the ITT analy-
sis and more abdominal distention and vomiting in the pro-
tocol evaluable analysis (Figure 4). Also, the acidified liquid
human milk fortifier group had significantly more diaper
4 Schanler et al
dermatitis (9% vs 1%; effect size, 7.4; 0.9-58.4), than
nonacidified liquid human milk fortifier group (P = .031).
The infants in both groups had similar mean ages at ini-
tiation of enteral feedings (2 days), achievement of an enteral
intake of 80 mL/kg/day (9 days), regaining birth weight (10
vs 11 days in the acidified liquid human milk fortifier and
nonacidified liquid human milk fortifier groups, respec-
tively), and age at randomization (10 vs 9 days in the acidi-
fied liquid human milk fortifier and nonacidified liquid human
milk fortifier groups, respectively).
There were no differences in energy intake between the
groups in the protocol evaluable analysis (Table III; available
at www.jpeds.com). There were differences in the protein
content of the human milk fortifiers (10% more in acidified
liquid human milk fortifier than in the nonacidified liquid
human milk fortifier) that led to significant greater protein
intakes in the acidified compared with the nonacidified liquid
human milk fortifier group for fortified human milk and total
protein intake from all sources (including nonstudy fortifier/
formula/supplements; Table III). Blood urea nitrogen values
did not reflect this difference in protein intake; by study day
29, the blood urea nitrogen values in infants fed acidified liquid
human milk fortifier were significantly lower than in infants
fed nonacidified liquid human milk fortifier (Table IV).
Significantly different biochemical assessments are pro-
vided in Table IV. The group receiving acidified liquid human
milk fortifier had significantly lower values of bicarbonate
content on study days 15 and 29 than the group receiving
nonacidified liquid human milk fortifier in both the ITT and

FLA 5.5.0 DTD ■ YMPD10104_proof ■ September 5, 2018

■■ 2018 ORIGINAL ARTICLES

Table IV. Biochemical assessments

Variable Acidified liquid human milk fortifier Nonacidified liquid human milk fortifier P value
ITT analysis
Bicarbonate (HCO3−)*, mEq/L
Study day 15 23.3 ± 0.5 (65) 26.5 ± 0.4 (76) <.001
Study day 29 25.4 ± 0.4 (57) 27.2 ± 0.4 (61) .039
Alkaline phosphatase,* U/L
Study day 15 334 ± 17 (69) 339 ± 16 (72) .019
Study day 29 296 ± 12 (61) 340 ± 15 (62) <.001
Blood urea nitrogen, mg/dL
Study day 15 13.2 ± 1.0 (69) 14.5 ± 0.7 (75) >.05
Study day 29 9.2 ± 0.6 (61) 11.3 ± 0.5 (61) >.05
Magnesium,† mg/dL
Study day 15 1.99 ± 0.07 (66) 2.30 ± 0.02 (72) <.001
Study day 29 1.91 ± 0.04 (57) 2.28 ± 0.03 (59) <.001
Potassium,† mEq/L
Study day 15 4.79 ± 0.08 (69) 5.47 ± 0.07 (77) <.001
Study day 29 4.57 ± 0.11 (59) 5.25 ± 0.08 (63) <.001
Sodium, mEq/L
Study day 15 137.7 ± 0.5 (69) 138.0 ± 0.4 (77) >.05
Study day 29 139.6 ± 0.4 (60) 139.2 ± 0.4 (63) >.05

Protocol evaluable analysis

Bicarbonate,* mEq/L
Study day 15 22.9 ± 0.6 (47) 26.7 ± 0.4 (59) <.001
Study day 29 24.9 ± 0.6 (35) 27.2 ± 0.5 (45) .006
Alkaline phosphatase,* U/L
Study day 15 321 ± 21 (49) 317 ± 15 (57) >.05
Study day 29 300 ± 14 (35) 328 ± 17 (47) .001
Blood urea nitrogen,* mg/dL
Study day 15 14.4 ± 1.4 (49) 14.5 ± 0.7 (59) >.05
Study day 29 9.1 ± 0.5 (36) 12.5 ± 0.6 (46) <.001
Magnesium,† mg/dL
Study day 15 1.89 ± 0.04 (46) 2.32 ± 0.02 (56) <.001
Study day 29 1.83 ± 0.05 (33) 2.31 ± 0.03 (45) <.001
Potassium,† mEq/L
Study day 15 4.80 ± 0.10 (49) 5.59 ± 0.07 (59) <.001
Study day 29 4.53 ± 0.13 (35) 5.40 ± 0.08 (47) <.001
Sodium, mEq/L
Study day 15 137.8 ± 0.6 (49) 138.4 ± 0.4 (59) >.05
Study day 29 139.6 ± 0.4 (36) 139.6 ± 0.5 (47) >.05

Data are presented as the mean ± standard error of the mean (n).
*Repeated measures analysis of covariance with significant feeding group by study day interaction; difference in group adjusted for study day.
†Repeated measures analysis of covariance, significant feeding group as main effect.

protocol evaluable analyses. Correspondingly, in the ITT analy-
sis, infants who received acidified liquid human milk forti-
fier had significantly more metabolic acidosis than infants fed
nonacidified liquid human milk fortifier (27% vs 5%; P < .001).
Eight infants (10%) in the group receiving acidified liquid
human milk fortifier and no infants in the group receiving
nonacidified liquid human milk fortifier were treated for meta-
bolic acidosis before study day 15 and through the study
(P < .001). Weight gain from study days 1-29 in the protocol
evaluable analysis correlated positively with base excess (r = 0.23;
P = .04) and weight gain for study days 1-15 correlated posi-
tively with base excess (r = 0.26; P = .01). The presence of meta-
bolic acidosis was associated with significantly longer duration
of stay in the neonatal intensive care unit (66 ± 4.1 days vs
57 ± 1.8 days) when comparing infants with metabolic aci-
dosis vs no acidosis, respectively (post hoc analysis: P = .028).
There were no differences in morbidity outcomes between
groups. The incidence of necrotizing enterocolitis (1.3% vs
1.2%) and confirmed sepsis (2.6% vs 3.7%) were similar in
Improved Outcomes in Preterm Infants Fed a Nonacidified Liquid Human Milk Fortifier: A Prospective
Randomized Clinical Trial
FLA 5.5.0 DTD ■ YMPD10104_proof ■ September 5, 2018
the acidified and nonacidified liquid human milk fortifier
groups, respectively. Bronchopulmonary dysplasia was noted
in 11% and 10% of infants in the acidified and nonacidified
liquid human milk fortifier groups, respectively. More infants
had adverse events in the acidified than the nonacidified liquid
human milk fortifier group (30% vs 11%; P = .001). Five sub-
jects (4 acidified liquid human milk fortifier, 1 nonacidified
liquid human milk fortifier) experienced serious adverse events.
Of the 4 infants in the acidified liquid human milk fortifier
group, 1 died from sepsis and 1 infant each had an infection,
metabolic acidosis, and necrotizing enterocolitis. The infant
in the nonacidified liquid human milk fortifier group devel-
oped renal failure.
Discussion
This prospective, multicenter, randomized trial of 2 liquid
human milk fortifiers available in the US identified short-
term differences in weight gain, feeding tolerance, and metabolic
5
THE JOURNAL OF PEDIATRICS • www.jpeds.com
acidosis in human milk-fed preterm infants receiving an acidi-
fied liquid human milk fortifier compared with similar infants
fed a nonacidified liquid human milk fortifier. Although the
overall 28-day weight gain did not differ between groups, aci-
dosis and its treatment were more common in infants receiv-
ing acidified liquid human milk fortifier likely enabling a catch-
up in weight gain.
Metabolic acidosis and growth faltering in premature infants
are serious morbidities. It is not unexpected that study site cli-
nicians would adjust the nutritional management (with treat-
ments for acidosis, discontinuance of human milk fortifier, or
additional nutritional supplements) to attempt to correct these
aberrations in this vulnerable population. Although these in-
terventions are clinically justified, they introduce substantial
variability making it more difficult to detect group differ-
ences in growth outcomes. Thus, the result of the post hoc
analysis of weight gain where much of the nutritional vari-
ability is excluded shows significantly better 28-day weight gain
with a nonacidified compared with an acidified liquid human
milk fortifier (Figure 2).
When compared with published data from other studies that
included infants fed acidified liquid human milk fortifier, the
data from our study are consistent and demonstrate the clini-
cal effects associated with feeding of an acidified human milk
fortifier.19-21,24 The findings in this trial can be explained by the
physiological controls that maintain blood pH within a narrow
range: bicarbonate buffering, urea excretion, and bone me-
tabolism. The bicarbonate buffering system drives a series of
reactions that decrease hydrogen ion concentration [H+].25 This
response leads to the hyperventilation typically seen with meta-
bolic acidosis and likely contributes to the significant de-
creases in [HCO3−] and [CO2] seen in acidified compared with
nonacidified liquid human milk fortifier in this and other
studies.19-21,24
Renal urea excretion also is a buffering system that elimi-
nates H+ ions.25 When induced experimentally, metabolic aci-
dosis is associated with decreased fractional protein synthesis,26,27
increased amino acid oxidation,27 decreased albumin synthesis,28
and an overall negative nitrogen balance.28 The decrease in blood
urea nitrogen values despite greater protein intake in acidi-
fied liquid human milk fortifier presumably indicates a dis-
ruption in protein metabolism owing to metabolic acidosis
(Table IV).
Bone serves as a reservoir of base [OH−] that is used to buffer
acidosis through the degradation of hydroxyapatite.29 H+ ions
directly stimulate osteoclast activity with a simultaneous re-
duction in alkaline phosphatase.30 The significant decrease in
serum alkaline phosphatase in acidified liquid human milk for-
tifier observed in this study is consistent with this response.
As illustrated in this and other studies, acidotic conditions
associated with the feeding of acidified liquid human milk for-
tifier lead to tolerance-related issues, including increases in
gastric residuals, abdominal distension, vomiting, and diaper
dermatitis. Our findings are consistent with published asso-
ciations between acidified vs nonacidified feedings: greater ces-
sation of feedings (66% vs 31%; P = .005), discontinuance
of human milk fortifier (41.0% vs 7.4%; P < .001), and
6 Schanler et al
FLA 5.5.0 DTD ■ YMPD10104_proof ■ September 5, 2018
Volume ■■
withholding of feedings for abdominal distension (25% vs 0%;
P = .04).20,21
Thus, the effects of acidified human milk fortifier in human
milk-fed preterm infants are strikingly consistent among studies.
In the current and previously published studies, the acidified
liquid human milk fortifier group received more protein,19-21,24
had no improved weight gain20,24 or significantly lower weight
gain,19,21 and significantly more metabolic acidosis.19-21,24 The
excess energy that may be required for increased CO2 produc-
tion and protein catabolism (to combat acidosis), could con-
tribute, at least in part, to the lack of expected growth with
the greater protein intake.
As reported in this study, infants with metabolic acidosis had
longer durations of hospital stay, which are associated with an
increased economic burden and are an indication of the ad-
ditional time required to achieve an appropriate discharge
weight. Therefore, given that acidified liquid human milk for-
tifier also is associated with poor feeding tolerance and that
nonacidified liquid human milk fortifiers are available, these
data support a greater use of nonacidified human milk
fortifier in the contemporary management of the preterm
infant. ■
We acknowledge the guidance provided by the data monitoring board.
We thank Marc Masor, PhD, for assistance in preparing the manu-
script. Dr Masor is an editing consultant supported by Abbott and has
no other conflicts of interest. We thank the following individuals for their
hard work and dedication: Debra Potak, RN, Mashelle Monhaut, MSN,
NNP-BC, Sue Zhang, MS, MAS, Maggie Hroncich, BS, Susan Toth,
MPH, Deborah Shye, Kelly Vance-Jude, Mary C. Sommerville, Melanie
Drummond, RN, BSN, Sonya Verill, MS, Kelly Samaie, BSN, RN, and
Amy Devitt-Maicher, PhD. A list of study sites is available at
www.jpeds.com (Appendix).
Submitted for publication Feb 13, 2018; last revision received Jun 26, 2018;
accepted Jul 2, 2018
Reprint requests: Richard Schanler, MD, Neonatal-Perinatal Medicine, Cohen
Children’s Medical Center, 269-01 76th Ave, New Hyde Park, NY 11040.
E-mail: schanler@northwell.edu
References
1. Ehrenkranz RA, Younes N, Lemons JA, Fanaroff AA, Donovan EF, Wright
LL et al. Longitudinal growth of hospitalized very low birth weight infants.
Pediatrics 1999;104:280-9.
2. Clark RH, Wagner CL, Merritt RJ, Bloom BT, Neu J, Young TE, et al. Nu-
trition in the neonatal intensive care unit: how do we reduce the inci-
dence of extrauterine growth restriction? J Perinatol 2003;23:337-44.
3. Clark RH, Thomas P, Peabody J. Extrauterine growth restriction remains
a serious problem in prematurely born neonates. Pediatrics 2003;111:986-
90.
4. Guellec I, Lapillonne A, Marret S, Picaud JC, Mitanchez D, Charkaluk ML,
et al. Effect of intra- and extrauterine growth on long-term neurologic
outcomes of very preterm infants. J Pediatr 2016;175:93-9.
5. Latal-Hajnal B, von Sebenthal K, Kovari H, Bucher HU, Largo RH. Post-
natal growth in VLBW infants: significant association with
neurodevelopmental outcome. J Pediatr 2003;143:163-70.
6. Ehrenkranz RA, Dusick AM, Vohr BR, Wright LL, Wrage LA, Poole WK.
Growth in the neonatal intensive care unit influences neurodevelopmental
and growth outcomes of extremely low birth weight infants. Pediatrics
2006;117:1253-61.
7. Franz AR, Pohlandt F, Bode H, Mihatsch WA, Sander S, Kron M, et al.
Intrauterine, early neonatal, and post discharge growth and
■■ 2018
neurodevelopmental outcome at 5.4 years in extremely preterm infants
after intensive neonatal nutritional support. Pediatrics 2009;123:e101-9.
8. Ramel SE, Demerath EW, Gray HL, Younge N, Boys C, Georgieff MK.
The relationship of poor linear growth velocity with neonatal illness and
two-year neurodevelopment in preterm infants. Neonatology 2012;102:19-
24.
9. Stoll BJ, Hansen NI, Bell EF, Shankaran S, Laptook AR, Walsh MC, et al.
Neonatal outcomes of extremely preterm infants from the NICHD Neo-
natal Research Network. Pediatrics 2010;126:443-56.
10. Schanler RJ, Shulman RJ, Lau C. Feeding strategies for premature infants:
beneficial outcomes of feeding fortified human milk versus preterm
formula. Pediatrics 1999;103:1150-7.
11. Schanler RJ. Mother’s own milk, donor human milk, and preterm for-
mulas in the feeding of extremely premature infants. J Pediatr Gastroenterol
Nutr 2007;45(Suppl 3):S175-7.
12. Lai KK. Enterobacter sakazakii infections among neonates, infants, chil-
dren, and adults. Case reports and a review of the literature. Medicine
(Baltimore) 2001;80:113-22.
13. MMWR. Enterobacter sakazakii infections associated with the use of pow-
dered infant formula—Tennessee, 2001. MMWR Morb Mortal Wkly Rep
2002;51:297-300.
14. Taylor CT. Health Professionals Letter on Enterobacter sakazakii infec-
tions associated with use of powdered (dry) infant formulas in neonatal
intensive care units. College Park, MD: Office of Nutritional Products,
Labeling and Dietary Supplements Center for Food Safety and Applied
Nutrition, USFDA; Safety Alerts and Advisories, October 10, 2002.
15. Euber J, Solorio H, Batema R, Walsh K. (inventors). Acidified liquid human
milk supplement. US Patent 8147894 B2; Original Assignee: Mead Johnson
Nutrition Company, Apr. 3, 2012.
16. Kim JH, Chan G, Schanler R, Groh-Wargo S, Bloom B, Dimmit R,
et al. Growth and tolerance of preterm infants fed a new extensively
hydrolyzed liquid human milk fortifier. J Pediatr Gastroenterol Nutr
2015;61:665-71.
17. Erickson TA, Gill G, Chan GM. The effects of acidification on human milk’s
cellular and nutritional content. J Perinatol 2013;33:371-3.
18. Mead Johnson Nutrition. Enfamil® human milk fortifier acidified liquid
http://www.meadjohnson.com/pediatrics/us-en/product-information/
Improved Outcomes in Preterm Infants Fed a Nonacidified Liquid Human Milk Fortifier: A Prospective
Randomized Clinical Trial
FLA 5.5.0 DTD ■ YMPD10104_proof ■ September 5, 2018
ORIGINAL ARTICLES
products/newborns/enfamil-human-milk-fortifier-acidified-liquid#product
-features. March 10, 2015.
19. Thoene M, Hanson C, Lyden E, Dugick L, Ruybal L, Anderson-Berry A.
Comparison of the effect of two human milk fortifiers on clinical out-
comes in premature infants. Nutrients 2014;6:261-75.
20. Cibulskis CC, Armbrecht ES. Association of metabolic acidosis with bovine
milk-based human milk fortifiers. J Perinatol 2015;35:115-9.
21. Kumar N, Monga R, Sampath V, Ehrhart B. Prospective comparison of
Enfamil and Similac liquid human milk fortifier on clinical outcomes in
premature infants. Am J Perinatol 2017;34:1411-6. ISSN 0735-1631.
22. Winsor CP. The Gompertz curve as a growth curve. Proc Natl Acad Sci
U S A 1932;18:1-8.
23. Song CQ, Kuznetsova OM. Fitting Gompertz nonlinear mixed models to
infancy growth data with SAS® version 8 procedure NLMIXED. SAS Con-
ference Proceedings: PharmaSUG 2001, paper sp03. http://
www.lexjansen.com/pharmasug/2001/proceed/stats/sp03_song.pdf.
Accessed May 16, 2017.
24. Moya F, Sisk PM, Walsh KR, Berseth CL. A new liquid human milk for-
tifier and linear growth in preterm infants. Pediatrics 2012;130:e928-35.
25. Kraut JA, Madias NE. Metabolic acidosis: pathophysiology, diagnosis and
management. Nat Rev Nephrol 2010;6:274-85.
26. Caso G, Garlick BA, Casella GA, Sasvary D, Garlick PJ. Acute metabolic
acidosis inhibits muscle protein synthesis in rats. Am J Physiol Endocrinol
Metab 2004;287:E90-6.
27. Reaich D, Channon SM, Scrimgeour CM, Goodship THJ. Ammonium
chloride-induced acidosis increases protein breakdown and amino
acid oxidation in humans. Am J Physiol Endocrinol Metab 1992;263:E735-
9.
28. Ballmer PE, McNurlan MA, Hulter HN, Anderson SE, Garlick PJ, Krapf
R. Chronic metabolic acidosis decreases albumin synthesis and induces
negative nitrogen balance in humans. J Clin Invest 1995;95:39-45.
29. Arnett TR. Extracellular pH regulates bone cell function. J Nutr
2008;138:415S-418S.
30. Rochow N, Jochum F, Redlich A, Korinekova Z, Linnemann K, Weitmann
K, et al. Fortification of breast milk in VLBW infants: metabolic acido-
sis is linked to the composition of fortifiers and alters weight gain and
bone mineralization. Clin Nutr 2011;30:99-105.
7
THE JOURNAL OF PEDIATRICS • www.jpeds.com Volume ■■

Appendix
Research Study Sites

Locations Site PIs and Co-Is

University of Louisville School of Medicine PI = David H. Adamkin MD
Louisville, KY
North Central Baptist Hospital PI = Kaashif A. Ahmad, MD
San Antonio, TX Co-Is: Mary Elizabeth Wearden, MD, Jacklyn Marie Levan, MD
Duke University Medical Center PI = Margarita Bidegain, MD
Durham, NC
Lehigh Valley Health Network PI = Nachammai Chinnakaruppan, MD
Allentown, PA
Wheaton Franciscan Healthcare, Inc.—St. Joseph, Milwaukee, WI PI = Jeffery Garland, MD
MetroHealth Medical Center PI = Sharon L. Groh-Wargo, PhD
Cleveland, OH
Doctors Hospital at Renaissance/Women's Hospital at Renaissance, Edinburg, TX PI = Dynio Honrubia, MD
Banner Good Samaritain Medical Center PI = Suganya Kathiravan, MD
Phoenix, AZ
Texas Health and Education Institute PI = Russell Lawrence, MD
Fort Worth, TX
Cook Children's Medical Center PI = David Riley, MD
Fort Worth, TX
Medstar Georgetown University Hospital PI = Suna C. Seo, MD
Washington DC Co-I = Nitin Mehta, MD
Cohen Children's Medical Center at North Shore, Manhasset, NY PI = Richard Schanler, MD
Co-Is: Dalibor Kurepa, MD, Olena Predtechenska, MD, Alla Zaytseva, MD
Cohen Children's Medical Center PI = Richard Schanler, MD
New Hyde Park, NY Co-Is: Dalibor Kurepa, MD, Olena Predtechenska, MD, Alla Zaytseva, MD
All Children's Hospital / Johns Hopkins Medicine, St Petersburg, FL PI = Fauzia Shakeel, MD
Memorial Hospital of South Bend PI = Robert D. White, MD
South Bend, IN

Co-Is, Co-investigator; PI, principal investigator.

* ** † ‡

Figure 2. Weight gain in study groups grouped by analysis: ITT, PE, and Post Hoc. ALHMF, acidified human milk fortifier; ITT,
intention-to-treat; NALHMF, nonacidified human milk fortifier; PE, protocol evaluable; SD, study day. Values are mean ± stan-
dard error of the mean. *P = .004, **P = .001, †P = .009, ‡P < .001.

7.e1 Schanler et al

FLA 5.5.0 DTD ■ YMPD10104_proof ■ September 5, 2018

■■ 2018 ORIGINAL ARTICLES

Figure 3. Differences in weight gain between groups from study days (SDs) 1-29 among 12 of the 14 study sites. Two sites
are not shown because each site enrolled 1 infant in one or both arms of study.

Table I. Key nutrient approximate levels of study forti-

fiers and as fed per 100 kcal
Acidified liquid Nonacidified liquid
human milk human milk
fortifier fortifier
Added to Added to
human milk human milk
Per 20 mL as fed per Per 20 mL as fed per Table III. Protein and energy intakes for study days 1-29
Nutrients (4 vials) 100 kcal (4 packets) 100 kcal
Acidified liquid human Nonacidified liquid
Energy, kcal 30 100 28 100 Intakes milk fortifier human milk fortifier
Volume, mL 20 124 20 125
Protein, g 2.2 4.0 2 3.6 Protein, g/kg/d
Calcium, mg 116 145 120 150 Fortified human milk† 4.17 ± 0.06 4.03 ± 0.04
Phosphorus, mg 63 80 68 83 Total intake*‡ 4.26 ± 0.06 4.11 ± 0.04
Magnesium, mg 1.8 5.3 8.4 12 Energy intake, kcal/kg/d
Iron, mg 1.76 1.91 0.44 0.58 Fortified human milk 114.0 ± 1.4 116.0 ± 1.1
Zinc, mg 0.96 1.37 1.24 1.63 Total intake* 118.0 ± 1.3 119.0 ± 1.0
Sodium, mg 27 57 20 50
Potassium, mg 45 98 84 146 Protocol evaluable analysis. Data are presented as the mean ± standard error of the mean.
Chloride, mg 28 89 52 113 *Other sources of intake other than fortified human milk were human milk alone, nonstudy fortifier/
formula/supplements. Percent calories or protein from other sources was <5% for both study
Vitamin C, mg 15.2 21 30.8 43.3
groups.
Lutein, mcg — — 14.0 23.0 †P = .029.
‡P = .010.

Improved Outcomes in Preterm Infants Fed a Nonacidified Liquid Human Milk Fortifier: A Prospective 7.e2
Randomized Clinical Trial
FLA 5.5.0 DTD ■ YMPD10104_proof ■ September 5, 2018