Blood biochemical proﬁle of very preterm infants before and after trophic
feeding with exclusive human milk or with formula milk
, Antonio Gutiérrez
, Eli Moliner
, Gemma Ginovart
Department of Paediatrics, Hospital de la Santa Cruz y San Pablo, 90 Mas Casanovas St., 08041 Barcelona, Spain
Neonatal Unit, Department of Paediatrics, Hospital de la Santa Cruz y San Pablo, 90 Mas Casanovas St., 08041 Barcelona, Spain
Division of Hematology, Molecular Biology Unit, Hospital Son Espases, Valldemossa Road 79, 07010 Palma de Mallorca, Spain
Received 22 December 2013
Received in revised form 10 February 2014
Accepted 14 February 2014
Available online 24 February 2014
Minimal enteral nutrition
Objectives: To determine whether feeding type of trophic feeds affect haematological and biochemical
markers in the very preterm infant.
Design and methods: Fifty-six very preterm infants were enrolled in this retrospective study (30 infants
were included in the only human milk-fed group and 26 in the formula-fed group). Routine haematological
and biochemical variables were collected in both groups on days 1 and 4 of life and fourteen serum markers
Results: There were no signiﬁcant differences between the two groups before starting trophic feeds. After
startingtrophic feeds, sodium and lactatelevels were signiﬁcantly higherin the human milk-fed groupcompared
with those measured in the formula-fed group.
Conclusion: The study demonstrates that supplementation of minimal enteral feeding with human milk
does affect biochemical proﬁles in very preterm infants. Small amounts of enteral feedings of formula and/or
human milk may result in different metabolic responses; these differences are reﬂected by different serum
© 2014 The Canadian Society of Clinical Chemists. Published by Elsevier Inc. All rights reserved.
In the ﬁrst postnatal days disturbances in the ﬂuid and electrolyte
balance occur frequently in very preterm infants. This imbalance can
lead to neurological impairment, amongst others [1,2].Previousstudies
have been designed to compare the biochemical status of newborn in-
fants fed human milk (HM) with that of similar infants fed formula.
No differences have been found in mean serum concentrations of
potassium, chloride, calcium, urea nitrogen, blood glucose, lactate or
pyruvate concentrations between feeding groups [3,4]. However, the
values for serum total proteins, albumin, gamma-globulins, serum cho-
lesterol, triglyceride, alanine aminotransferase, aspartate aminotrans-
ferase, gamma-glutamyltransferase, total bilirubin, direct bilirubin,
methionine, threonine, ketonebody and sodium levels were signiﬁcant-
ly higher in the breast-fed group compared with those measured in the
formula-fed group [4–8].
The milk amount required for this metabolic effect is not known.
These studies were carried out on full term neonates at or beyond day
5 of life, involving infants on nutritive milk feeds . To date, no study
has been carried out to determine speciﬁcally the effect on very preterm
neonates of nutritionally insigniﬁcant volumes designed to stimulate
the developing gastrointestinalsystem (gut priming or trophic feeding).
We hypothesise that a very small amount of human milk in the ﬁrst
days of life is enough to change the metabolic response of very preterm
In our neonatal intensive care unit (NICU), we recently implemented
the strategy to use pasteurized human donor breast milk (DM) on
preterm infants whose mothers suffered from breastmilk shortage.
Motivated by the scarce data available in the literature, we decided to
analyse our data in order to verify that our shift to exclusive human
milk trophic feeding during the ﬁrst four days of life was associated
with changes of biochemical and haematological markers in the prema-
DM has been available in our NICU since April 2009. Thereafter, tro-
phic feeds consisted of non-fortiﬁed DM when there was an insufﬁcient
mother's breastmilk supply. Until April 1, 2009, the nutrition regimen
consisted of preterm formula when there was an insufﬁcient mother's
This is a preliminary report that compares laboratory results of in-
fants who received only-HM with a historical cohort that received any
Clinical Biochemistry 47 (2014) 584–587
Abbreviations: DM, pasteurized human donor breast milk; HM, human milk; NICU,
neonatal intensive care unit.
E-mail address: firstname.lastname@example.org (S. Verd).
0009-9120/© 2014 The Canadian Society of Clinical Chemists. Published by Elsevier Inc. All rights reserved.
Contents lists available at ScienceDirect
journal homepage: www.elsevier.com/locate/clinbiochem
amount of formula. We performed a retrospective search for laboratory
results for the ﬁrst four days of postnatal life from eligible infants
(b30 weeks of gestational age admitted to our NICU between 1st
January 2007 and 31st December 2012). Day 1 sample was taken before
starting feeds from every infant's ﬁrst available blood test result on day
1 of life. Day 4 sample was taken after starting feedsfrom every infant's
last available blood test result on day 4 of life. The following exclusion
criteria were the same for both groups: death on day 1 of life, admission
to the NICU after day 1 of life and fasting until day 5 of life. There was
one additional exclusion criterion for the historical cohort: exclusively
breastfed infants. Finally, 30 infants made up the experimental group
and 26 infants, the historical group.
Clinical features and laboratory parameters were collected from the
hospital charts. Infant growth wasclassiﬁed as small for gestational age
using the Fenton chart. Table 1 presents the variables considered to as-
sess the comparability of the two groups. The study was approved by
the Institutional Review Boards of our Hospital and informed consent
was obtained from parents of all infants.
Enteral feeding was usually initiated on the second day of life. Every
infant received 1 mL every 3 h, with daily increments b10 mL/kg/day
according to clinical condition. Whenever possible, parenteral nutrition
was initiated on the second day of life. All infants received standard
management in the NICU. None of these practices changed over the
course of the study periods.
Evaluation of biochemical and haematological parameters
Electrolyte levels were measured by indirect Ion Selective Electrode
on the Abbott Architect CI16000 laboratory analyser. Conventional lab-
oratory analyses of pH and blood gases were performed with bench-top
analysers (GEM 3500 Premier).
Biochemical parameters were also measured by the Abbott Architect
CI16000 laboratory analyser. Complete blood count and the differential
leukocyte count were carried out on blood samples by a cell counter
(Sysmex XE-500, USA).
The data were analysed using the SPSS statistical package. Categori-
cal variables were compared by means of the χ
test or, when appropri-
ate, Fisher's exact test. Continuous variables were compared by means
of the t-test. Multivariate analysis was not performed because it is not
recommended for small samples.
As shown in Table 1, only-HM and formula groups were comparable
in terms of maternal and infant characteristics, except in the case of age
at the onset of enteral feeds. Table 2 shows that there were no statisti-
cally signiﬁcant differences in laboratory results between the two
groups before starting the feeds.
After starting trophic feeds
Serum sodium levels (mean 141 versus 142 mmol/L; p = 0.026)
and serum lactate levels (mean 1.4 versus 1.6 mmol/L; p = 0.049)
were signiﬁcantly decreased in the formula group compared to the
Glycaemia levels were close to signiﬁcantly decreased in the
formula group compared to the only-HM group (mean 5.8 versus
7.6 mmol/L; p = 0.065).
There were no statistical differences in any other biochemical or
haematological markers between groups.
Multiple births: surfactant therapy
There were considerable numbers of multiple births and of infants
requiring an exogenous surfactant. Hence, we have analysed if there
are any signiﬁcant differences in baseline biochemical parameters be-
tween these groups and if they respond similarly to trophic feeds.
Single births compared to multiple births. Baseline levels of sodium (136
versus 134; p = 0.019), haemoglobin (14.6 versus 15.5; p = 0.03)
and leukocyte count (13,320 versus 8260; p = 0.008) were signiﬁcantly
different. On day 4, levels of sodium (142 versus 137; p = 0.035),
haemoglobin (13 versus 13.8; p = 0.005) and potassium (4.3 versus
5.5; p = 0.033) were signiﬁcantly different.
Effect of exogenous surfactant. Baseline aspartate aminotransferase
levels were signiﬁcantly decreased in the ‘without exogenous surfac-
tant’group (25.5 vs 35.5; p = 0.017). On day 4, platelet count was
signiﬁcantly increased in the ‘without exogenous surfactant’group
(272,500 vs 161,000; p = 0.016).
There were no statistical differences in any other biochemical or
haematological markers between these groups.
Comparison with other studies.
A recent international survey observed that most of the units with
access to DM start enteral feeding on the ﬁrst day of life and advance
more rapidly than units without access to DM . These data are in
line with our ﬁndings, where DM recipients exhibit an earlier onset of
Slight differences in sodium intake do not change sodium levels in
adults and supplemented diets (4–5 mmol sodium/day versus 1–1.5
Premature newborn characteristics an d postnatal outcom e in the ﬁrst six days of life
according to type of feeding.
Baseline characteristics Formula
Gender (M/F), n 11/15 13/17 1.000
Primiparous, n (%) 10 (42%) 19 (63%) 0.11
Gestational age (wk), mean ± SD 27.36 ± 1.52 26.88 ± 1.54 0.67
Multiple births, n (%) 8 (32%) 8 (27%) 0.66
NI from another hospital, n (%) 4 (16%) 2 (7%) 0.27
Birth weight (g), mean ± SD 1000 ± 221 899 ± 237 0.55
Small for gestational age,n (%) 3 (12%) 5 (17%) 0.62
Prenatal steroids, n (%) 20 (83%) 29 (97%) 0.093
Mode of delivery (V/CS), n 11/13 18/12 0.30
Apgar 5th minute, mean ± SD 8.13 ± 1.55 8.48 ± 1.05 0.10
Exogenous surfactant, n (%) 18 (72%) 20 (67%) 0.67
Age at onset of enteral feeds (days),
mean ± SD
2.24 ± 0.83 1.97 ± 0.32 0.002⁎
Early onset sepsis, n (%) 6 (24%) 3 (10%) 0.16
IVH grades 3–4, n (%) 2 (8%) 2 (7%) 0.42
Abbreviations: CPAP —continuous positive airway pressure; CS —Caesarean section;
F—female; g —grams; IPPV —intermittent positive pressure ventilation; IVH —
intraventricular haemorrhage; M —male; n —number; NI —newborn infant; SD —
standard deviation; V —vaginal delivery; wk —weeks.
Given as a supplement to mother's own breast milk.
⁎Signiﬁcant at the 0.05 level.
585S. Verd et al. / Clinical Biochemistry 47 (2014) 584–587
mmol sodium/day) have been seen not to change sodium levels in very
pretermneonates from the 5th to the 11th postnatal day . Therefore,
elevated breast milk sodium concentration is not regarded as responsi-
ble for elevated plasma sodium levels. Conversely, recent evidence sug-
gests that artiﬁcialfeedings appear to exerta separate detrimental effect
when they replace breast milk as initial postnatal nutrition .
Our data are in agreement with those of Anctil et al. . They have
unequivocally shown that mean blood sodium concentration was sig-
niﬁcantly higher in exclusively breastfed infants, when compared to
formula-fed infants. It is worthy to note that as the volume of breast
milk supplied increases, the sodium blood level gap between exclusive-
ly breastfed infants and non-exclusively breastfed infants widens (from
1 mmol/L in our study to 6 mmol/L in Anctil's study).
Mean lactate levels in our sample are close to hyperlactataemia
(≥2 mmol/L) on day 1, whereas on day 4 they are well below this ﬁg-
ure. This is at odds with previous analyses which have noted that lactic
acid levels decline rapidly in the ﬁrst postnatal week .Thisdecline
can be explained in part by the metabolic effects of feeding practices
in the preterm infant. Lucas et al.  measured blood levels of several
metabolites and hormones before and after the ﬁrst feed of breastmilk
in a group of neonates. They found that the mean fasting lactate blood
level was 2.0 mmol/L and the mean lactate blood level after an infusion
of 2.5 mL of breastmilk/kg/h was 1.67 mmol/L.
Our study has several limitations owing to its retrospective design,
including the risk of conﬁrmation bias. Hence, the comparison of groups
according to feeding type could not be controlled. To minimise these
limitations, we have adjusted for prenatal and postnatal characteristics
to reduce confounding, but there is a potential confounding impact of
age at start of enteral feeds. It is unlikely that by day 4 the babies re-
ceived different cumulative volumes of milk because of cautious ad-
vances of b10 ml/kg/day. Our data are particularly robust thanks to
the lack of variability of objective standardised laboratory results. The
descriptive design of this study yields enlightening conclusions,
avoiding unethical experimental trials.
This study demonstrates that trophic feeding with human milk is
reﬂected by different serum biochemistries. It appears that a very
small amount of human milk in the ﬁrst days of life affect metabolic re-
sponse of very preterm infants.
We thank the families of patients who agreed to participate in the
study. The authors would like to acknowledge nurses and paediatricians
working at the nursery section and, Cecília Martínez-Brú, M.D. and the
technical staff who have performed the analyses. We thank Berta
Verd, M.Sc., for editing the manuscript and Prof. Micah Leshem for
proof reading and thoughtful contributions.
 BaratonL, Ancel PY, FlamantC, Orsonneau JL, Darmaun D, Rozé JC. Impact of changes
in serum sodium levels on 2-year neurologic outcomes for very preterm neonates.
 Mandel D, Littner Y, Mimouni FB, Stavarovsky Z, Dollberg S. Increased serum potas-
sium and intraventricular hemorrhage revisited. Isr Med Assoc J 2004;6:91–4.
 Gross SJ. Growth and biochemical response of preterm infants fed human milk or
modiﬁed infant formula. N Engl J Med 1983;308:237–41.
 Calvert SA, Soltesz G, Jenkins PA, Harris D, Newman C, Adrian TE, et al. Feeding pre-
mature infants with h uman milk or preterm milk formula. Effects on postnatal
growth and on circulating concentrations of intermediary metabolites, amino
acids, and regulatory peptides. Biol Neonate 1985;47:189–98.
Comparison of serum biochemical values and hematologic characteristics in preterm infants fed with supplements of donor milk or formula.
Median (interquartile range)
Number of patients = 26
Human donor milk
Number of patients = 30
DOL 1, before initiation of enteral feeds
/L 252 (207–288) 214 (170–284) 0.24
Haemoglobin, g/dL 15.2 (13.8–16.2) 15 (13.2–17.05) 0.83
/L 10,000 (6225–15,750) 11,590 (6020–19,1 00) 0.82
Sodium, mmol/L 135 (133.5–137) 136 (134–137) 0.66
Potassium, mmol/L 4.1 (3.7–4.4) 4.2 (4–4.5) 0.35
Chloride, mmol/L 108 (104.7–110.2) 108 (104.7–112.5) 0.69
Ionised calcium, mmol/L 1.28 (1.18–1.34) 1.27 (1.19–1.41) 0.76
Glycemia, mmol/L 3.1 (1.8–4.3) 3.3 (2.4–5.3) 0.40
AST, IU/L 31 (19.2–41.7) 30 (24–44.5) 0.89
Creatine kinase, IU/L 268 (143–410) 217 (131.5–391) 0.64
pH 7.27 (7.23–7.33) 7.26 (7.20–7.33) 0.46
Base excess, mEq/L −3.6 (−5.75; −0.95) −3.7 (−6.32; −2.07) 0.40
Bicarbonate, mEq/L 19.9 (18.3–22.55) 19.7 (19.07–22.1) 0.6 0
Lactic, acid mmol/L 1.85 (1.35–4.45) 2 (1.42–3.45) 0.50
DOL 4, after initiation of enteral feeds
/L 201 (129–244) 161 (115–311) 0.72
Haemoglobin, g/dL 13 (11.8–14.6) 13.5 (11.6–16.3) 0.65
/L 10,100 (7200–15,500) 8010 (4500–26,370) 0.21
Sodium, mmol/L 141 (136–142.5) 142 (140.7–151) 0.026⁎
Potassium, mmol/L 4.7 (4.2–5.5) 4.8 (4.2–5) 0.77
Chloride, mmol/L 115 (111–118) 121 (113–135) 0.21
Ionised calcium, mmol/L 1.27 (1.16–1.31) 1.23 (1.12–1.29) 0.35
Glycemia, mmol/L 5.8 (3.0–7.0) 7.6 (5.8–11.1) 0.065
AST, IU/L 20 (17.2–27) 33 (17–51) 0.24
Creatine kinase, IU/L 69 (57.5–155.5) 238 (30–246) 0.31
pH 7.28 (7.25–7.35) 7.31 (7.21–7.39) 0.27
Base excess, mEq/L −4.5 (−6.2; −2.7) −5(−7; −3.95) 0.26
Bicarbonate, mEq/L 20 (18.6–20.7) 20 (19–20.8) 0.50
Lactic acid, mmol/L 1.4 (1.1–1.8) 1.6 (1.4–2.7) 0.049⁎
Abbreviations: AST —aspartate aminotransferase; DOL —day of life; IU —international units; mEq —milliequivalents; mmol —millimols; L —litre; WBCs —white blood cells.
Given as a supplement to mother's own breast milk.
⁎Signiﬁcant at the 0.05 level.
586 S. Verd et al. / Clinical Biochemistry 47 (2014) 584–587
 El-Hawary MFS, Soliman AA, Nosseir SA, Kholief TS. Biochemical studies on
the effect of breast and artiﬁcial feeding in newborn Egyptian infants I. Serum
proteins and immunoglobulins in 1–4-day-old newborn. Z Ernährungswiss
 Schanler RJ, Lau C, Hurst NM, Smith EO. Randomized trial of donor human milkver-
sus pretermformula as substitutesfor mothers' own milk in the feeding of extreme-
ly premature infants. Pediatrics 2005;116:400–6.
 Wu T-Z, Huang I-F, Chen Y-C, Chen P-H, Yang L-Y. Differences in serum bio-
chemistry between breast-fed and formula-fed infants. J Chin Med Assoc
 de Rooy L, Hawdon J. Nutritional factors that affect the postnatal metabolic adapta-
tion of full-term small- and large-for-gestational-age infants. Pediatrics 2002;109:
 McClureRJ, Newell SJ. Randomisedcontrolled trial of clinical outcome following tro-
phic feeding. Arch Dis Child Fetal Neonatal Ed 2000;82:F29–33.
 Klingenberg C, Embleton ND, Jacobs SE, O'Connell LA, Kuschel CA. Enteral feeding
practices in very preterm infants: an international survey. Arch Dis Child Fetal
Neonatal Ed 2012;97:F56–61.
 Al-Dahhan J, Jannoun L, Haycock GB. Effect of salt supplementation of newborn pre-
mature infants on neurodevelopmental outcome at 10–13 years of age. Arch Dis
Child Fetal Neonatal Ed 2002;86 [F120–F123].
 Meier PP, Engstrom JL, Patel AL, Jegier BJ, Bruns NE. Improving the use of human
milk during and after the NICU stay. Clin Perinatol 2010;37:217–45.
 Anctil S, Carceller A, Chevalier I, Gonthier M, Bernard-Bonnin AC, Cousineau J, et al.
Natremia of healthy term newborns at 48 h of life: inﬂuence of feeding patterns.
Clin Biochem 2006;39:992–6.
 Nadeem M, Clarke A, Dempsey EM. Day 1 serum lactate values in preterm infants
less than 32 weeks gestation. Eur J Pediatr 2010;169:667–70.
 Lucas A, Bloom SR, Aynsley-Green A. Metabolic and endocrine events at the time of the
ﬁrst feed of human milk in preterm and term infants. Arch Dis Child 1978;53:731–6.
587S. Verd et al. / Clinical Biochemistry 47 (2014) 584–587