D F J Yumani - Thesis

VU Research Portal
NUTRITION IN RELATION TO THE ENDOCRINE REGULATION OF PRETERM GROWTH

AND BODY COMPOSITION
Yumani, Dana Foekina Johanna
2022
document version
Publisher's PDF, also known as Version of record
Link to publication in VU Research Portal
citation for published version (APA)

Yumani, D. F. J. (2022). NUTRITION IN RELATION TO THE ENDOCRINE REGULATION OF PRETERM
GROWTH AND BODY COMPOSITION. [PhD-Thesis - Research and graduation internal, Vrije Universiteit
Amsterdam]. Ridderprint.
General rights
Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners
and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights.
• Users may download and print one copy of any publication from the public portal for the purpose of private study or research.
• You may not further distribute the material or use it for any profit-making activity or commercial gain
• You may freely distribute the URL identifying the publication in the public portal
Take down policy

If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately
and investigate your claim.
E-mail address:
[email protected]
Download date: 23. May. 2024

NUTRITION IN RELATION TO THE
ENDOCRINE REGULATION OF PRETERM
GROWTH AND BODY COMPOSITION
Dana F.J. Yumani

2022
ISBN: 978-94-6458-608-4
Cover design and lay-out: Publiss | www.publiss.nl
Print: Ridderprint | www.ridderprint.nl
© Copyright 2022: Dana F.J. Yumani, Amsterdam, The Netherlands

All rights reserved. No part of this publication may be reproduced, stored in a
retrieval system, or transmitted in any form or by any means, electronic, mechanical,
by photocopying, recording, or otherwise, without the prior written permission of
the author.
VRIJE UNIVERSITEIT
NUTRITION IN RELATION TO THE ENDOCRINE

REGULATION OF PRETERM GROWTH
AND BODY COMPOSITION
ACADEMISCH PROEFSCHRIFT
ter verkrijging van de graad Doctor aan
de Vrije Universiteit Amsterdam,
op gezag van de rector magnificus
prof.dr. J.J.G. Geurts,
in het openbaar te verdedigen
ten overstaan van de promotiecommissie
van de Faculteit der Geneeskunde
op woensdag 12 oktober 2022 om 13.45 uur
in een bijeenkomst van de universiteit,
De Boelelaan 1105
door
Dana Foekina Johanna Yumani
geboren te Enschede
Promotoren: prof.dr. M.M. van Weissenbruch
prof.dr. H.N. Lafeber †
promotiecommissie: prof.dr. J.B. van Goudoever

prof.dr. R.M. van Elburg
prof.dr. A.C. Heijboer
prof.dr. A.C.S. Hokken-Koelega
dr. W. Onland
dr. J. Rotteveel
Voor papa
TABLE OF CONTENTS
INTRODUCTION 11
CHAPTER 1: General introduction 13
Background 14
The regulation of fetal and postnatal growth and body composition 14
Aims and objectives 16
Study design 17
Thesis outline 20
References 21
PART I. THE ROLE OF NUTRITION AND IGF-I ON GROWTH, BODY 25

COMPOSITION AND HEALTH OUTCOMES IN PRETERM INFANTS
IN INFANCY
CHAPTER 2: Dietary proteins and IGF I levels in preterm infants: 27
determinants of growth, body composition and neurodevelopment
Abstract 28
Introduction 29
IGF I regulation 30
The role of IGF I in growth and body composition 30
The role of IGF I in neurodevelopment 34
The role of dietary proteins in growth and body composition 35
The role of dietary proteins in neurodevelopment 37
IGF I and dietary proteins 38
Conclusions 39
References 40
CHAPTER 3: Associations between bronchopulmonary dysplasia, 47

Insulin-like growth factor I and nutrition
Abstract 48
Introduction 49
Methods 49
Results 51
Discussion 56
Conclusions 58
References 59
Supplemental material 61
CHAPTER 4: The course of IGF-1 levels and nutrient intake in 69
extremely and very preterm infants during hospitalisation
Abstract 70
Introduction 71
Methods 72
Results 75
Discussion 83
Conclusions 85
References 86
PART II. THE DETERMINANTS OF BODY COMPOSITION AND METHODS 89

TO ASSESS BODY COMPOSITION IN PRETERM INFANTS
CHAPTER 5: IGF-I, growth and body composition in preterm infants 91
up to term equivalent age
Abstract 92
Introduction 93
Methods 93
Results 96
Discussion 108
Conclusions 111
References 112
CHAPTER 6: A comparative study using Dual-energy X-ray 115

absorptiometry, air displacement plethysmography and skinfolds
to assess fat mass in preterms at term-equivalent age
Abstract 116
Introduction 117
Methods 118
Results 120
Discussion 126
References 128
CHAPTER 7: Body composition in preterm infants: a systematic 131
review on measurement methods
Abstract 132
Introduction 133
Methods 133
Results 137
Discussion 156
Conclusions 161
References 163
DISCUSSION & SUMMARY 169

CHAPTER 8: General discussion 171
Nutrition in relation to the endocrine regulation of preterm 172
growth and body composition
The developing endocrine axis in relation to comorbidities in 174
preterm infants
Determinants and assessment of body composition in preterm 175
infants
Future research directions 176
References 178
CHAPTER 9: Summary 183

Part I. IGF-I and nutrition in relation to growth, body composition 184
and health outcomes in preterm infants
Part II. Determinants and assessment of body composition in 184
preterm infants
APPENDIX 187
CHAPTER 10: PhD portfolio 189
CHAPTER 11: Acknowledgements 195
INTRODUCTION
1
CHAPTER 1
General introduction
Chapter 1
Background
Embryonic and fetal development are an exceptional phase of human life: a one-
celled zygote develops into an infant with roughly 200 different cell types, weighing
on average 3.5 kg at birth. This exponential growth and differentiation rate will
never be equaled in later life. Preterm delivery, however, abruptly interrupts this
process and results in a very precarious situation. The now premature infant finds
itself in an alien extra-uterine environment while it is yet to go through a major part
of its’ development.
As premature infants can no longer rely on the regulatory function of the placenta
for nutrient supply, immunity and endocrine control, their growth and development
depend on immature organ systems. The developing gastro-intestinal tract still has
an impaired digestive and absorptive capacity.(1) Furthermore, the intestines have
an inadequate mucosal barrier function and maternal antibodies, which are largely
transferred across the placenta in the last trimester, are lacking.(1, 2) Combined
with an overall immature immune response and a general pro-inflammatory state,
premature infants are left prone to infections. Moreover, the regulation of growth
is disrupted by insufficient levels of growth factors.(3) Altogether, preterm infants
are faced with the preposterous task of thriving in the face of comorbidities and
insufficient nutrient and hormone supplies. Obviously, this is a process prone to
error, making postnatal growth failure a common problem in preterm infants. (4, 5)
After an initial phase of impaired growth, preterm infants are likely to show
accelerated growth up to 2-3 years of age.(6) Growth patterns in infancy and
early childhood have been linked to health outcomes. Notably, postnatal growth
restriction is associated with an increased incidence of co-morbidities during
hospitalization (7) and may lead to impaired neurodevelopment in later life (8).
Furthermore, there are concerns that impaired growth triggers a thrifty phenotype
with increased adipose tissue and adverse cardiometabolic outcomes in later life.
(9, 10) Therefore, it would be of interest to gain more insight in factors determining
growth and body composition in early life, as a means to enhance health outcomes
in infants born preterm.
The regulation of fetal and postnatal growth and

body composition
The regulation of fetal growth and body composition is an intricate process.
There is an interplay between oxygen, nutrients, and hormones of fetal, placental
and maternal origin. In addition, genetic factors determine a part of the growth
potential. After preterm birth, however, a new balance needs to be found between
substrates, endocrine and genetic factors.
14
The endocrine regulation of growth and body composition

Insulin like growth factors (IGFs), insulin and growth hormone (GH) have been
established to play an important role in regulating growth in the fetal as well as
1
the postnatal phase. IGF-II mainly stimulates growth in the embryonic and early
fetal period. IGF-I, on the other hand, is the major growth stimulating factor in late
gestation and the postnatal period. (11) Hence, IGF-I is a key factor in the endocrine
regulation of growth in preterm infants.
IGF-I is a small polypeptide with a wide range of function. It is mainly synthesized

in the liver and stimulates cell division, growth, and motility, as well as glucose
uptake, and protein synthesis. In addition, IGF I inhibits apoptosis and has an anti-
inflammatory and anti-oxidative effect. (12-15)
In utero, IGF-I is secreted into the blood under control of insulin. After birth, GH
gradually takes over this role. Insulin like growth factor binding proteins (IGFBPs)
are also important factors which regulate the bioavailability of IGF-I. (11) In
addition, thyroid hormones in general stimulate cell differentiation and support
tissue accretion by regulating the IGF-I receptor and oxidative metabolism.
Glucocorticoids also stimulate cell differentiation and negatively modulate growth
in late gestation. Other factors implicated in fetal and postnatal growth are leptin
and ghrelin. Both hormones have a possible growth promoting effect, yet their
precise roles remain to be elucidated. (16)
IGF-I has also been associated with the regulation of body composition. (12)
Interestingly, the relationship between IGF-I and body composition has been
reported to vary depending on the timing of IGF-I and body composition
measurement. For example, in preterm infants higher IGF-I in the first month of
life has been associated with an increased fat free mass. Meanwhile, higher IGF-I
levels at and after term equivalent age have been associated with a decreased fat
free mass.(17) Nevertheless, studies on this are rare in the preterm population and
definitive conclusions on the relationship between IGF-I and body composition are
yet to be drawn.
Nutrition in relation to growth and body composition

Nutrients are a substrate for growth and are crucial for the development of the
endocrine axes. In particular protein availability has significant impact on hormone
levels and hormone sensitivity in fetal as well as neonatal life.(18-20) Protein
malnutrition in infants depresses hepatic IGF-I, IGFBP-3 and IGFBP-4 synthesis
and enhances hepatic IGFBP-1 and IGFBP-2 synthesis, leading to reduced IGF-I
bioavailability and impaired growth.
Preterm infants require 3 to 4 grams of protein per kg per day depending on their
gestational age and co-morbidities. To optimize nitrogen accretion, the protein
15
Chapter 1
intake needs to be combined with sufficient fatty acids and carbohydrates,

resulting in a high energy diet of 95-140 kcal per kg per day. In an attempt to
prevent complications from the high energy and high protein diet, nutritional
intake has to be build up in the first week of life causing relative malnutrition in this
period. Moreover, parenteral nutrition is required until the preterm infant is able
to tolerate full enteral feeds and studies have shown that parenteral nutrition, in
contrast to enteral feeding, is associated with lower IGF-I levels. (21, 22) In addition,
the type of enteral nutrition, i.e. own mother’s milk, donor human milk or formula,
influences IGF-I levels. (23) Furthermore, macronutrient intake has been linked
to body composition. A high protein intake has been associated with decreased
fat mass and increased fat free mass in infancy. (24, 25) Therefore, the route of
administration and the type of nutrition play a vital role in optimizing postnatal
growth and body composition.
Aims and objectives

The research outlined in this thesis aims to explore the postnatal modulation of
growth and body composition in very preterm and extremely preterm infants
(gestational age at birth 24 – 32 weeks). Furthermore, the influence of the
developing endocrine axis on health outcomes in infancy is investigated.
It is hypothesized that:
• IGF-I has to reach a threshold concentration before it can effectively influence
growth. Once IGF-I passes this threshold concentration, the maximum growth
rate is expected to be potentiated by IGF-I;
• In states with low IGF I levels to ensure an easily accessible energy store, but
resulting in an increased fat mass percentage at term age;
• Before IGF-I reaches the threshold con, such as critical illness, nutrient
restriction and extreme prematurity, the IGF system would stimulate
mesenchymal stem cell differentiation towards adipogenesis as a mechanism
centration, a high-energy and high-nutrient diet is required to potentiate
growth;
• Once IGF-I reaches the threshold concentration, a continued high-energy
and high-nutrient diet could potentially lead to increased fat deposition;
• The protective effect on inflammation and the anti-oxidative effects of IGF-I
may be important in decreasing the risk of developing comorbidities.
To investigate these hypotheses a longitudinal cohort study was designed and

literature reviews were conducted. In this thesis all papers, except the literature
reviews, are based on the results of the Nutrition in relation to the endocrine
regulation of preterm growth study (NUTRIE study).
16
Study design
The NUTRIE study, a longitudinal cohort study, was conducted between August
2015 and August 2018. The primary objective of this study was to study the endocrine
1
regulation of preterm growth and body composition. Secondary objectives were
to study the influence of the endocrine regulation and early nutritional intake on
neurodevelopmental outcome, bone mineralization, lipid profile and blood pressure
in preterm infants. To detect a medium size effect (r = 0.35) of IGF-I concentration
on fat mass percentage a sample size of at least 62 infants was required (power
80%, significance 5%). With an expected dropout rate of 10% drop out rate, the aim
was to include 70 patients.
Infants admitted to the neonatal intensive care unit (NICU) of the Amsterdam UMC
- location VU University Medical Center were assessed for eligibility if they were
born at a gestational age of 24 weeks + 0/7 days up to and including 31 weeks
+ 6/7 days. Infants who had substantial congenital abnormalities were excluded.
Informed consent was obtained within the first week of life. Ninety patients were
enrolled in the study. (Figure 1)
Study participants were followed-up until 2 years corrected age. Anthropometric

measures and endocrine parameters were registered during hospital stay as
well as at follow-up. Body composition and bone mineral density were assessed
at term age and 3, 6, 12, and 24 months corrected age, using air displacement
plethysmography (PEA POD®/BOD POD®) and Dual Energy X-ray Absorptiometry
(DXA). (Figure 2 and Figure 3)
In addition cord blood analysis of infants with a gestational age of 24 to 42 weeks

born at the VU University Medical Center in the study period took place to establish
neonatal reference ranges for IGF I, IGF BP 3, Insulin, C-peptide, glucose, cortisol,
cortisone and lipid profiles. (Figure 2 and Figure 3)
17
Chapter 1
Figure 1. NUTRIE study recruitment and inclusion
18
Figure 2. NUTRIE Study: procedures during hospitalization
Figure 3. NUTRIE Study: procedures during follow-up
19
Chapter 1
Thesis outline
Part I of this thesis focuses on the role of nutrition and IGF-I on growth, body
composition and health outcomes in preterm infants in infancy. Chapter 2
elaborates on the role of dietary proteins and IGF-I on growth, body composition
and neurodevelopment. Chapter 3 aims to explore how the developing IGF-I axis,
in relation to nutrition, is associated with the occurrence of bronchopulmonary
dysplasia. Chapter 4 investigates the influence of the route of administration
and the type of nutrition on IGF-I levels and growth in preterm infants during
hospitalisation.
Part II of this thesis focuses on the determinants of body composition and

methods to assess body composition in preterm infants. Chapter 5 describes
the role of IGF-I and weight gain in determining body composition of preterm
infants at term equivalent age. Chapter 6 compares body composition, generated
by air displacement plethysmography and dual-energy X-ray absorptiometry.
Furthermore, it evaluates the potential predictive value of the sum of skinfolds for
body composition measured in preterm infants at term equivalent age. In Chapter
7 a systematic review is reported, comparing different methods to assess body
composition in preterm infants up to 6 months corrected age. Chapter 8 discusses
the conclusions of this thesis and the implications for future research.
20
References
1. Neu J. Gastrointestinal development and meeting the nutritional needs of premature infants.
Am J Clin Nutr. 2007;85(2):629s-34s.
1
2. Sharma AA, Jen R, Butler A, Lavoie PM. The developing human preterm neonatal immune
system: a case for more research in this area. Clin Immunol. 2012;145(1):61-8.
3. Hellstrom A, Ley D, Hansen-Pupp I, Hallberg B, Ramenghi LA, Lofqvist C, et al. Role of Insulinlike
Growth Factor 1 in Fetal Development and in the Early Postnatal Life of Premature Infants.
American journal of perinatology. 2016;33(11):1067-71.
4. Horbar JD, Ehrenkranz RA, Badger GJ, Edwards EM, Morrow KA, Soll RF, et al. Weight Growth
Velocity and Postnatal Growth Failure in Infants 501 to 1500 Grams: 2000-2013. Pediatrics.
2015;136(1):e84-92.
5. Lee SM, Kim N, Namgung R, Park M, Park K, Jeon J. Prediction of Postnatal Growth Failure
among Very Low Birth Weight Infants. Scientific reports. 2018;8(1):3729.
6. Euser AM, de Wit CC, Finken MJ, Rijken M, Wit JM. Growth of preterm born children. Horm Res.
2008;70(6):319-28.
7. Griffin IJ, Tancredi DJ, Bertino E, Lee HC, Profit J. Postnatal growth failure in very low
birthweight infants born between 2005 and 2012. Archives of disease in childhood Fetal and
neonatal edition. 2016;101(1):F50-5.
8. Cormack BE, Harding JE, Miller SP, Bloomfield FH. The Influence of Early Nutrition on Brain
Growth and Neurodevelopment in Extremely Preterm Babies: A Narrative Review. Nutrients.
2019;11(9).
9. Nakano Y. Adult-Onset Diseases in Low Birth Weight Infants: Association with Adipose Tissue
Maldevelopment. J Atheroscler Thromb. 2020;27(5):397-405.
10. Euser AM, Finken MJ, Keijzer-Veen MG, Hille ET, Wit JM, Dekker FW, et al. Associations
between prenatal and infancy weight gain and BMI, fat mass, and fat distribution in young
adulthood: a prospective cohort study in males and females born very preterm. Am J Clin Nutr.
2005;81(2):480-7.
11. Gicquel C, Le Bouc Y. Hormonal regulation of fetal growth. Horm Res. 2006;65 Suppl 3:28-33.
12. Yumani DF, Lafeber HN, van Weissenbruch MM. Dietary proteins and IGF I levels in preterm
infants: determinants of growth, body composition, and neurodevelopment. Pediatr Res.
2015;77(1-2):156-63.
13. Rowland KJ, Choi PM, Warner BW. The role of growth factors in intestinal regeneration and
repair in necrotizing enterocolitis. Semin Pediatr Surg. 2013;22(2):101-11.
14. Xu L, Zhang W, Sun R, Liu J, Hong J, Li Q, et al. IGF‑1 may predict the severity and outcome
of patients with sepsis and be associated with microRNA‑1 level changes. Experimental and
therapeutic medicine. 2017;14(1):797-804.
15. Tanner SM, Berryhill TF, Ellenburg JL, Jilling T, Cleveland DS, Lorenz RG, et al. Pathogenesis of
necrotizing enterocolitis: modeling the innate immune response. Am J Pathol. 2015;185(1):4-16.
16. Ohkawa N, Shoji H, Kitamura T, Suganuma H, Yoshikawa N, Suzuki M, et al. IGF-I, leptin and
active ghrelin levels in very low birth weight infants during the first 8 weeks of life. Acta
Paediatr. 2010;99(1):37-41.
17. Hernandez MI, Rossel K, Pena V, Cavada G, Avila A, Iniguez G, et al. Leptin and IGF-I/II during
the first weeks of life determine body composition at 2 years in infants born with very low birth
weight. J Pediatr Endocrinol Metab. 2012;25(9-10):951-5.
18. DJP B. Programming the baby. Mothers, Babies and Health in Later Life: Churchill Livingstone;
1998. p. 13-41.
19. Yeung MY, Smyth JP. Nutritionally regulated hormonal factors in prolonged postnatal growth
retardation and its associated adverse neurodevelopmental outcome in extreme prematurity.
Biol Neonate. 2003;84(1):1-23.
21
Chapter 1
20. Bloomfield F, Spiroski A, Harding J. Fetal growth factors and fetal nutrition. Semin Fetal
Neonatal Med. 2013.
21. Wojnar MM, Fan J, Li YH, Lang CH. Endotoxin-induced changes in IGF-I differ in rats provided
enteral vs. parenteral nutrition. The American journal of physiology. 1999;276(3):E455-64.
22. Yumani DFJ, Calor AK, van Weissenbruch MM. The Course Of IGF-1 Levels and Nutrient Intake
in Extremely and Very Preterm Infants During Hospitalisation. Nutrients. 2020;12(3).
23. Larnkjaer A, Molgaard C, Michaelsen KF. Early nutrition impact on the insulin-like growth
factor axis and later health consequences. Current opinion in clinical nutrition and metabolic
care. 2012;15(3):285-92.
24. Embleton ND, Cooke RJ. Protein requirements in preterm infants: effect of different levels of
protein intake on growth and body composition. Pediatr Res. 2005;58(5):855-60.
25. Roggero P, Gianni ML, Amato O, Liotto N, Morlacchi L, Orsi A, et al. Growth and fat-free
mass gain in preterm infants after discharge: A randomized controlled trial. Pediatrics.
2012;130(5):e1215-e21.
22
23
PART I
THE ROLE OF NUTRITION AND IGF-I
ON GROWTH, BODY COMPOSITION
AND HEALTH OUTCOMES IN PRETERM
INFANTS IN INFANCY
2
CHAPTER 2
Dietary proteins and IGF I levels in preterm
infants: determinants of growth, body
composition and neurodevelopment
Dana FJ Yumani, Harrie N Lafeber and Mirjam M van Weissenbruch.

Pediatr Res. 2015 Jan;77(1-2):156-63. doi:10.1038/pr.2014.172.
Chapter 2
Abstract
It has been demonstrated that a high protein diet in preterm born infants during
the first weeks of life may enable a growth rate equal to that seen in utero and may
also result in a better long term neurodevelopmental outcome. This diet may limit
immediate postnatal growth retardation and may hence lower the risk of increased
fat deposition after birth leading to the metabolic syndrome in later life. Insulin
like growth factor I (IGF I) has proven to play an important role in early postnatal
growth of preterm infants, but also seems to have a persisting influence on body
composition in childhood. Furthermore increased IGF I concentrations in preterm
infants have been associated with improved neurodevelopmental outcome.
This review will elaborate on the role of dietary proteins and IGF I on growth, body
composition and neurodevelopment of preterm infants. Possible causal pathways
will be explored and areas for future research will be proposed.
28
Dietary proteins and IGF I levels in preterm infants
Introduction
Postnatal growth restriction is a major problem faced in the care for preterm
infants. At 36 weeks postmenstrual age 91% of all preterm infants show postnatal
growth restriction (weight < -1.3 SD) (1). At term age approximately 30% of infants are
reported to still be growth restricted (2). As survival rates of preterm infants with an
increasingly younger gestational age rise, we are subsequently confronted with the
2
long term sequelae of preterm birth. At 11 years of age 40% of children born before 26
weeks of gestation have been reported to have serious neurocognitive impairment
and moderate to severe impairment of neuromotor function, vision and hearing was
reported in respectively 10%, 9% and 2% of cases (3). Preterm birth and postnatal
growth restriction have both been associated with impaired neurodevelopmental
outcome (4). However Franz and colleagues found that only a small percentage of
the variability, roughly 3%, of the mental processing composite score was explained
by growth (5). There might be a common factor leading to both poor growth and
poor neurodevelopment, e.g. a poor nutritional status or major neonatal morbidities.
Nonetheless several studies suggest that there might be independent pathways (5,
6). Either way these poor ourcomes warrant an intervention.
Furthermore preterm infants are prone to develop risk factors for the onset of the
metabolic syndrome. They are reported to have lower insulin sensitivity, increased
blood pressure and increased fat mass in childhood and young adulthood (7-
9). Nutritional interventions in these infants have been found to influence the
development of risk factors for the metabolic syndrome (10). Hence neonatologists
are challenged to compose and administer a diet which limits postnatal growth
restriction; yet with caution to also limit the development of risk factors for the
onset of the metabolic syndrome.
Dietary factors, endocrine function and the simple immaturity of organ systems
are entangled in the endeavour to optimize postnatal growth and metabolic
programming. Dietary proteins are essential in enabling a growth rate similar to
intrauterine growth (11). Nevertheless the balance between proteins and other
nutrients are essential to understand how growth and body composition in
preterm infants can be optimized. Insulin-like growth factors (IGF) are a key in the
endocrine regulation of growth. Notably IGF I has an anabolic and mitogenic effect
which is crucial for symmetric growth due to the presence of the IGF I receptor
in multiple cell types and tissues. Moreover IGF I synthesis in multiple peripheral
tissues causes it to function as an auto- and paracrine factor which does not
merely influence growth, but also organ functioning. IGF I’s possible influence on
neurodevelopmental outcome may be potentiated through its trophic effect or
through altering the functioning of the central nervous system.
In this review we aim to explore the possible pathways relating neonatal dietary
protein intake and IGF I levels to growth, body composition and neurodevelopmental
outcome in infancy, childhood and young adulthood.
29
Chapter 2
IGF I regulation
IGF I is a small polypeptide which is mainly synthesized in the liver. It stimulates cell
division, cell growth, cell motility, glucose uptake and protein synthesis. Furthermore
IGF I inhibits apoptosis. Prenatally it is secreted into the blood under control of
insulin. Postnatally growth hormone (GH) gradually takes over this role. In addition
malnutrition and hypothyroidism negatively influence IGF I plasma levels. IGF I is also
synthesized in multiple peripheral tissues, e.g. kidney, bone and muscle, where it is
released under control of GH and local factors. 99% of IGF I in plasma is bound to high
affinity IGF binding proteins which control IGF I transportation and distribution. (12)
The role of IGF I in growth and body composition

Fetal IGF I levels gradually increase during pregnancy to reach approximately 46
to 90 ng/ml at term age (13). After preterm birth IGF I levels slowly increase (14).
Meanwhile infants born at term show a quick surge in IGF-I levels (15). Figure 1
illustrates postnatal IGF I levels in preterm and term infants.
In preterm infants IGF I levels at birth are positively correlated with birth weight
(16). Until term age these infants IGF I levels’ are also positively correlated with their
preceding as well as their subsequent weight gain, indicating higher previous as
well as higher subsequent growth rate in those infants with higher IGF I levels (14).
Figure 1. Postnatal Insulin-like growth factor I (IGF I) levels in preterm and term infants.
Preterm infants: Hansen-Pupp (○) (14), Ohkawa (□) (17), Van de Lagemaat (△) (18), Giapros (◊) (19), Wang (▽)
(20). Term infants: Iniguez (▲) (21), Kurtoglu (●) (15), Wang (◆) (20), Larnkjaer (■) (22), Hyun (▼) (23), Ong (+) (24)
30
However after term age contradictory findings have been reported. Several studies
observed IGF I levels to positively correlate with current growth parameters and
preceding growth velocity in preterm as well as healthy term infants (18, 19, 25, 26). In
contrast findings concerning the correlation between IGF I levels and subsequent
2
growth velocity are inconclusive (Table 1).
Table 1. Associations between IGF I and growth
Van de Van de Giapros Chellakooty Ong (24) Socha (26)

Lagemaat Lagemaat (19) (25)
(18) (18)
Study Very preterm Very preterm Late Healthy term Healthy Healthy
population infants infants preterm infants term term infants
infants infants
Timing Term age 3 months 6 weeks, 3 & 3 months 3 months 6 months

IGF I blood 6 monthsb
draw
IGF I & ↗ Δ weight ↗ Δ weight ↗ weightc ↗ Δ weight Not ↗ ΔWFL

previous SDS SDS SDS reported SDS
growth (birth - term (birth - 3 (birth - 3 (birth - 6
velocity age) monthsa) months) months)
↗ Δ length ↗ Δ length ↗ lengthc ↗ Δ length

SDS SDS (birth - 3 SDS
(birth - term monthsa) (birth - 3
age) months)
IGF I & ↗ weight and ↗ weight and ↗ weight ↗ weight; ↗ weight; ↗ WFL SDS
current length SDS length SDS and length → length →length
growth SDS
IGF I & ↘ Δ weight → Δ weight → (growth ↘Δ weight →Δ weight →Δ WFL

subsequent SDS SDS parameters SDS (3 - 12 SDS
growth (term age - 6 (3 - 6 monthsa ) not (3 - 18 months) months) (6 – 12
velocity monthsa) specified) months)
↘ Δ length → Δ length →Δ length ↗ Δ length

SDS SDS SDS (3 - 12
(term age - 6 (3 - 6 monthsa) (3 - 18 months) months)
monthsa )
↗ = positive correlation; ↘ = negative correlation; → = no correlation; Δ = gain; a = corrected age; b = chronological

age; c = higher IGF I levels in infants with accelerated previous weight and length gain (a difference of more
than 0.67 SDS between two study points); IGF I = Insulin-like growth factor I; SDS = standard deviation score;
WFL = weight-for-length
Hypothesizing these findings might reflect that after term age a turning point
occurs. At this point infants with the lowest IGF I levels and thus the poorest
previous growth may tend to show accelerated growth. This hypothesis would
be in line with the negative correlation between IGF I and subsequent growth
velocity found in the above stated studies: infants with the lowest IGF I levels had
the highest subsequent growth velocity. Less comorbidity in preterm infants after
term age might create a less stressful environment in which catch-up growth
could occur, i.e. an increased growth rate compared to the infant’s previous growth
31
Chapter 2
rate enabling the infant to reach a body size comparable to that of healthy infants
at a corresponding age. Also further maturation of the neuroendocrine axes and
target organs may improve feedback mechanisms and thus optimize growth
control. Hypothesizing more into detail eventually a point may be reached where
growth velocity is fixed regardless of the previous growth pattern. This would then
correspond with the absence of a correlation between IGF I levels and subsequent
growth. At this point the neuroendocrine axes and target organs may be completely
programmed setting the growth rate at a fixed point.
Concluding these diverse associations may possibly illustrate a regulatory effect

to direct growth towards the mean. For IGF I to function optimally, i.e. to enable
growth to the full potential of each individual, certain conditions are paramount.
The neuroendocrine axes and target organs need to be matured to such an extent
that feedback mechanisms can reliably control growth, i.e. correct too slow as
well as too fast growth. Furthermore the environment needs to reinforce a steady
growth rate; meaning that sufficient nourishment needs to be available and
stressful factors, such as comorbidities, should be limited. Hypothesizing after term
age there may be an optimal combination of these conditions, creating a window
of opportunity to catch up in growth. In comparison, in children who are adopted
from impoverished and stressful situations growth restriction occurs. When they
are placed in a more nurturing environment through adoption catch-up growth
is observed. In a study by Miller et al. adopted children with the lowest IGF I had
4.9 times higher odds (95%CI: 1.1 – 22.9) of showing catch up growth in height than
children with the highest IGF I (mean age at adoption 20.1 months + 9.8) (27).
Simplified, preterm infants show three growth patterns: small size at birth and
persistent small size at term age (small for gestational age infants), appropriate
size at birth but small size at term age (appropriate for gestational age infants
with postnatal growth restriction) and appropriate size at birth and appropriate
size at term age (appropriate for gestational age infants without postnatal growth
restriction). In analogy with children following adoption it is hypothesized that
infants small at term age come into a less stressful period after term age which
enables them to catch up in growth. Indeed some infants, but not all, show ‘catch-
up’ growth. The majority of this ‘catch-up’ growth occurs in the first 6 to 12 months
of life (28). However there is a concern that rapid growth leads to an unfavourable
metabolic and cardiovascular outcome in later life. Indeed De Jong et al. found
increased length gain between 6 and 12 months corrected age and weight gain
between term and 2 years corrected age to be associated with increased systolic
blood pressure at 2 years corrected age (29). In addition Singhal and colleagues
found a poorer endothelial function in preterm born adolescents with the highest
rate of weight gain in the first two weeks after birth (30). In comparison in term
born young adults rapid weight gain in the first three months of life has also been
associated with decreased insulin sensitivity, a higher percentage of body fat and
more central adiposity (31). Moreover Hovi et al. demonstrated that a decrease in
32
weight z-score from birth to term was associated with a higher blood pressure in
adulthood. However, this association did not remain significant after adjusting for
gestational age at birth (32). Nonetheless a recent review by Lapillonne and Griffin
on the effect of postnatal growth on metabolic and cardiovascular outcomes in
2
preterm born adults concluded that, in contrast to growth during late infancy
and childhood, growth up to 1 year was not associated with adult blood pressure,
glucose tolerance or lipid profile (33). However, the studies described in the review
were heterogenic and did not all take a possible confounding effect of nutrition and
small versus appropriate for gestational age into account. Therefore the concern of
a possible negative impact of initial growth restriction and subsequent catch-up
growth expressed through several studies cannot yet be disregarded.
Currently there is no full understanding of which infants will and which ones will
not completely ‘catch-up’. As in children following adoption one might hypothesize
that infants with the lowest IGF I level in the early post term period are the ones
who will show catch up growth. However to our knowledge as of yet there is no
evidence supporting this hypothesis in preterm infants. This hypothesis regarding
catch-up growth might depend on the plasticity of the neuroendocrine axes and
target organs. In certain infants intrauterine or early-life insults may completely
program the neuroendocrine axes and target organs. However if there is some
plasticity left, alteration of the growth rate may occur. This might further depend
on environmental factors, e.g. lack of comorbidities and sufficient nourishment, in
combination with the genetic make-up of the infant.
Remarkably, in term infants several studies did not find a correlation between
IGF I and subsequent weight gain, while they did find a positive correlation
with subsequent length gain. Thus higher IGF I levels were associated with a
subsequent lower BMI (24, 34). This might imply that high IGF I levels protect
against adiposity. Growth restricted and preterm infants would then be at
increased risk of developing obesity, because of their presumably low IGF I levels.
Indeed, in small for gestational age very low birth weight infants IGF I levels up to
3 months corrected age have been positively associated with lean mass at 2 years
(35). Moreover preterms were found to have increased fat mass and decreased lean
mass in childhood (9). However a recent meta-analysis could not confirm that this
trend persists into adulthood (36).
Yet it remains to be clarified whether IGF I is primarily associated with change in

height or is equally related to change in weight. Surely IGF I is involved in bone
accretion, but it is also implied in adipogenesis (37, 38). Indeed Stigson et al. found
that higher IGF I levels at a postmenstrual age of 30 to 32 weeks were associated
with increased bone mass (9). Also, a trend of higher IGF I levels was found in
preterm infants who increased in bone strength compared to preterm infants
with a decrease in bone strength measured by bone speed of sound (39). In line
with these findings preterm infants born small for gestational age had decreased
bone accretion at 6 months corrected age. In addition 20 year olds who were born
33
Chapter 2
preterm, especially those small for gestational age, had decreased bone mineral
density and were shorter compared to term controls (40, 41). Interestingly, however,
others found normal bone mass in 4 year old children who were born preterm (9). As
stated by Stigson and colleagues, osteoblasts as well as adipocytes are derived from
the same progenitor cells and the IGF system could be important in directing the
differentiation to either adipocytes or osteoblasts (9). It may well be that states with
low IGF I levels, such as critical illness, nutrient restriction and extreme prematurity,
stimulate differentiation towards adipogenesis as a mechanism to ensure an easily
accessible energy store, in this relatively catabolic state, as compared to anabolic
states with high IGF I levels where sustainable growth through bone formation
might be obtained.
Nevertheless the role of IGF I in growth, body composition and development of the
metabolic syndrome remains complex. For instance lower IGF I levels in infancy
have been associated with higher IGF I levels in later life (22). This suggests that
events in early life can program IGF I and possibly metabolic outcomes in later life.
However low as well as high IGF I levels in adulthood have been associated with
the metabolic syndrome and cardiovascular disease (42). Therefore it is difficult to
give a clear-cut view of the role of IGF I alone in growth, body composition and the
development of the metabolic syndrome.
The role of IGF I in neurodevelopment

In clinical studies, brain and cranial growth have been associated with subsequent
neurodevelopment (5, 43). A polymorphism in the IGF I promotor gene, which is
known to regulate serum IGF I levels, has been related with slower cranial growth
from birth until 5 years of age (44). Moreover Hansen-Pupp and colleagues found
IGF I levels to correlate with brain volumes while there was no association with
cerebral spinal fluid volume. The authors hypothesize that this could imply that
IGF I does not limit atrophy secondary to brain damage, but rather stimulates brain
growth (45).
In premature infants a higher rate of increase of IGF I until 35 weeks postmenstrual

age has directly been related to a better neurodevelopmental outcome at 2
years of age (43). In line with that, Okuma and colleagues found that IGF I levels
were associated with white matter organization (46). Interestingly, mean IGF I
concentration was positively correlated to neurodevelopmental outcome during
a period, from 30 to 35 weeks postmenstrual age, when a surge in IGF I levels
occurred and infants started growing after a phase of postnatal growth restriction
(14, 43). This may suggest that IGF I has to reach a certain level before it can enhance
neurodevelopmental outcome. Even so, the premature disruption of the maternal-
placental-fetal unit alters more neuroendocrine factors than merely IGF I, which
also influences the final neurodevelopmental outcome.
34
In analogy with the development of retinopathy of prematurity the sudden decrease

in IGF I at birth could cause stagnation in vascular growth. It is hypothesized that
the surge in IGF I might lead to neovascularisation with abnormal vessel formation,
which could cause intracranial haemorrhage and consequently influence
2
neurodevelopmental outcome. In experimental studies it has been suggested
that IGF I may limit damage after hypoxic-ischemic brain injury and inflammation
(47, 48). Moreover mice treated with IGF I seemed to have increased proliferation
of immature oligodendrocytes, while the number of mature oligodendrocytes
remained the same. This was hypothesized to possibly promote myelination at
later stages when the immature oligodendrocytes mature and start myelinating
(49). In addition lipopolysaccharide induced brain inflammation in a mouse model
led to lower IGF I levels and impaired myelination in the subcortical white matter
(50). However in a rat periventricular leukomalacia model it was demonstrated
that exogenous IGF I limited lipopolysaccharide induced damage at a low dose,
while it increased damage at higher doses (51). Recently IGF I administration has
been investigated in a phase I study in preterm infants and showed to effectively
increase IGF I levels without any adverse events (52). In the near future this might
offer a therapeutic intervention potentially improving neurodevelopment as well
as growth and body composition.
The role of dietary proteins in growth and body

composition
In the first weeks of life preterm infants almost universally accumulate a protein
deficit and show postnatal growth restriction. In an attempt to achieve an
intrauterine-like growth rate neonatologists are challenged to administer the
right composition of amino acids and the optimal amount of proteins, combined
with sufficient fatty acids and carbohydrates, to optimize nitrogen accretion.
Currently the recommended range of protein intake for preterm infants is 3.5 to
4.5 g/kg/day (53). Over the past few years increasing amounts of parenteral amino
acids have been administered to preterm infants showing a consistent increase
in protein balance. Recent nutritional studies have actually demonstrated that
by administering high dose parenteral amino acids current recommendations
for protein intake and intrauterine-like growth rates can be achieved, nutritional
deficits can drastically be reduced and postnatal growth restriction can in part
be prevented (Figure 2) (11). It is recognized that high and early introduction of
proteins can limit the initial postnatal weight loss (54). By reducing initial weight
loss the tendency for rapid ‘catch-up’ growth might be reduced, which may lead to
more favourable metabolic programming. Indeed low protein levels are associated
with low IGF I levels (55), which in turn is associated with fat mass accretion in
childhood (35).
35
Chapter 2
Figure 2. Protein intake and change in weight z-sore.
Protein intake (a) and change in weight z-score (b) in the first 6 weeks of life achieved by standard practice
in our neonatal intensive care unit in Amsterdam, the Netherlands (■) (unpublished data) compared with
that reported by Embleton (●) (56) before current ESPGHAN guidelines and Senterre (▲) (57) using current
ESPGHAN guidelines.
However after the initial period of weight loss, in which parenteral feeding is
the primary source of nutrition, growth sets in. When growth occurs, protein
requirements can be re-evaluated and slowly tapered off to reach 2-3 g/kg/day
at term age (58). Caution is warranted to maintain an appropriate protein intake
when transitioning from parenteral to enteral nutrition. However enteral and
parenteral protein intake might not be similar. For instance, bypassing the enteral
route is likely to lower the systemic availability of certain amino acids which are
metabolised from other amino acids in the intestine and/or liver (54).
In spite of several studies which did not find an increased growth after augmenting
protein intake (59, 60), most studies demonstrate that increased protein intake
in the neonatal period positively influences growth up to term age (61-65). These
studies report improved absolute and standardized measures of weight, length
and head circumference as well as an increased growth velocity. No intolerance of
high protein diets has been reported (58, 66) and glycaemic control might actually
be improved with a high protein intake (59, 60). However protein intake in the
neonatal period will not necessarily have an impact on growth indices in childhood
36
(65). A high protein intake in the in-hospital as well as in the post discharge period
seems to decrease fat mass and increase lean mass up to 6 months corrected age
(67-69). Whether this trend persists into childhood is not known.
2
Using bone transmission time Scattolin and colleagues found protein intake to
positively correlate with bone mineral status at 36 weeks postmenstrual age (66).
However Fewtrell was unable to correlate protein intake in infancy with peak bone
mass or bone turnover in young adulthood (41).
Thus a persisting beneficial effect of early protein intake on growth and body
composition in later life has not yet been confirmed.
Even though protein is vital to optimize growth, its relation to other nutrient
components and the administration of specific amino acids are equally important.
Indeed Mcleod and colleagues demonstrated that an increased protein/energy ratio
reduced adipose tissue accretion as compared to muscle accretion. Surely energy
from another source than protein itself is necessary for net protein gain. However
when non-protein caloric intake surpasses 60 kcal/kg/day, protein intake itself is
the primary determinant of protein gain. Nevertheless it should be questioned to
which level the protein/energy ratio should be increased. The ESPGHAN committee
on nutrition recommends a ratio of 3.2 to 4.1 g protein/100 kcal (53). Yet there is a
need of supportive evidence as to which ratio should be maintained at specific
points in time. Several studies found that when preterm infants were on complete
enteral nutrition increasing the protein/energy ratio above 3 g protein/ 100 kcal did
not improve fat free mass accretion compared to a ratio of 2.7-2.8 g protein/100 kcal
(67, 70). To our knowledge studies conducted so far have not assessed the effects on
body composition of various protein/energy ratios in the first two weeks of life.
Because preterm infants have limited ability to synthesize certain non-essential

amino acids those amino acids become conditionally essential. Some have proposed
that the addition of these so called semi-essential amino acids to the diet of preterm
infants will improve growth. Cysteine for example, has been implied to be one of the
key factors which potentiate the trophic effect of high protein diets (71).
The role of dietary proteins in neurodevelopment

During hospitalisation increased protein intake improves head growth in preterm
infants (72, 73). Even so total energy and lipid intake also have been positively
correlated with head growth (73, 74). Nonetheless Hansen-Pupp and colleagues
could not associate protein and caloric intake with brain volumes (43) and in
several studies protein-enriched nutrition failed to improve neurodevelopmental
outcome up to 18 months corrected age (59, 72, 75, 76). Macro- and microstructural
brain analyses could not be correlated to intake of protein or other nutritional
components either (46). Yet, two studies by Stephens et al. and Cormack et al.
37
Chapter 2
showed that protein intake in the first weeks of life was positively correlated with
the cognitive and motor score on the Bayley Scales of Infant Development (77, 78).
In addition Biasini and colleagues found that increased protein intake in extremely
low birth weight infants improved performance and hearing/language scores on
the Griffith Development Mental Score at 3 and 12 months corrected age (79).
Moreover increased fat free mass, which is claimed to reflect protein accretion, was
associated with faster neuronal processing at 4 months corrected age (80). Also,
perinatal protein restriction in mice altered the intracerebral dopamine circuit
which caused altered reward-processing and hyperactivity (81). The authors suggest
that this could possibly be translated to adverse neurodevelopmental outcome,
such as ADHD, in growth restricted infants. Furthermore in preterm infants who
were fed a high nutrient diet larger caudate volumes and higher verbal IQ were
found in adolescence (82).
So far pathways explaining the association between neurodevelopment and protein

intake are still speculative. Compared to other nutrient components the unique
feature of protein might just lie in the alteration of neuronal processing. Perhaps
that the underlying mechanism works through increased neurotransmitter- and
receptor synthesis. Indeed increased lactalbumin intake in rats increased cortex
tryptophan. Nevertheless casein had a negative effect on tryptophan (83). Also de
Kieviet and colleagues found an increased oral glutamine intake to be associated
with increased white matter-, hippocampus-, and brain stem volumes in very
preterm children at school age (84). Since glutamine has been shown to reduce
the number of serious infections in very preterm children, they hypothesized that
increased glutamine intake indirectly influences neurodevelopment by reducing
infections in the neonatal period.
IGF I and dietary proteins

As stated earlier IGF I levels are related to nutritional intake. Socha et al.
demonstrated that infants fed high-protein follow-up formula had higher IGF
I levels than those fed low-protein follow-up formula (26). Moreover a minimal
caloric as well as a minimal protein intake has to be reached to maintain normal
IGF I levels (12). Furthermore there is a strong negative effect of breast milk on IGF I
levels. Next to the lower protein content other, yet to be determined, factors might
play a role in establishing this effect (85). Given the reciprocal relation between
IGF I and nutrition, nutritional interventions might be the key factor in improving
growth, body composition and neurodevelopmental outcome of preterm infants.
Interestingly, Hansen-Pupp and colleagues found that IGF I and nutritional
intake only correlated after a postmenstrual age of 30 weeks (14). The timing of a
nutritional intervention may therefore be crucial for the sustainability of its effect.
38
Conclusions
Altogether preterm and small for gestational age infants are at risk for impaired
growth and a suboptimal body composition, making them prone to risk factors for
the metabolic syndrome. Low early postnatal IGF I levels seem to be at the origin of
this problem. Increased early dietary protein intake has shown to improve growth
and body composition in infancy. However it is yet to be elucidated whether this
2
trend persists in later life, thus calling for long-term follow-up studies.
Higher IGF I levels and increased dietary protein intake have been found to also
improve neurodevelopmental outcome of preterm infants. Evidence supports
a trophic role of IGF I in the development of the central nervous system. So far
signs of a direct mechanism which limits damage from hypoxic-ischemic and
inflammatory insults have only been found in experimental studies. However it
is plausible that improved neuronal processing due to higher IGF I levels and an
increased dietary protein intake may play a role in preterm infants.
Since IGF I levels are related to dietary protein intake it would be valuable to
investigate whether a nutritional intervention could improve the long-term
outcome of preterm infants by optimizing IGF I levels as well as optimally using
the potential of IGF I. It may be argued whether initial IGF I levels are sufficient or
should be further increased by increasing nutrient intake in the early postnatal
period. On the other hand nutrient intake might have to be reduced once IGF I
reaches the level where it’s trophic and neurodevelopment enhancing potential
becomes effective, creating a more favourable setting for further development.
Moreover, assessment of the optimal protein/energy ratio in this period may be
a key to improve metabolic programming and studies on specific amino acids
could ameliorate dietary advices. In addition it has to be questioned whether
IGF I administration to preterm infants could offer a potential future therapeutic
intervention. Altogether the above illustrates the important gap of knowledge in
potential causal pathways between dietary protein intake, IGF I levels and long-
term outcomes of preterm infants that needs to be explored in future research.
39
Chapter 2
References
1. Fanaroff AA, Stoll BJ, Wright LL, Carlo WA, Ehrenkranz RA, Stark AR, et al. Trends in neonatal
morbidity and mortality for very low birthweight infants. Am J Obstet Gynecol. 2007;196(2):147-8.
2. Leppanen M, Lapinleimu H, Lind A, Matomaki J, Lehtonen L, Haataja L, et al. Antenatal
and postnatal growth and 5-year cognitive outcome in very preterm infants. Pediatrics.
2014;133(1):63-70.
3. Johnson S, Fawke J, Hennessy E, Rowell V, Thomas S, Wolke D, et al. Neurodevelopmental
disability through 11 years of age in children born before 26 weeks of gestation. Pediatrics.
2009;124(2):e249-e57.
4. Cooke RW, Foulder-Hughes L. Growth impairment in the very preterm and cognitive and
motor performance at 7 years. Arch Dis Child. 2003;88(6):482-7.
5. Franz AR, Pohlandt F, Bode H, Mihatsch WA, Sander S, Kron M, et al. Intrauterine, early neonatal,
and postdischarge growth and neurodevelopmental outcome at 5.4 years in extremely
preterm infants after intensive neonatal nutritional support. Pediatrics. 2009;123(1):e101-e9.
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(4):1253-61.
7. Rotteveel J, van Weissenbruch MM, Twisk JW, HA D-VdW. Infant and childhood growth
patterns, insulin sensitivity, and blood pressure in prematurely born young adults. Pediatrics.
2008;122(2):313-21.
8. Kerkhof GF, Breukhoven PE, Leunissen RW, Willemsen RH, Hokken-Koelega AC. Does preterm
birth influence cardiovascular risk in early adulthood? J Pediatr. 2012;161(3):390-6.
9. Stigson L, Kistner A, Sigurdsson J, Engstrom E, Magnusson P, Hellstrom A, et al. Bone and fat
mass in relation to postnatal levels of insulin-like growth factors in prematurely born children
at 4 y of age. Pediatr Res. 2014;75(4):544-50.
10. Lafeber HN, van de Lagemaat M, Rotteveel J, van Weissenbruch M. Timing of nutritional
interventions in very-low-birth-weight infants: optimal neurodevelopment compared with the
onset of the metabolic syndrome. Am J Clin Nutr. 2013;98(2):556S-60S.
11. Rigo J, Senterre T. Intrauterine-like growth rates can be achieved with premixed parenteral
nutrition solution in preterm infants. J Nutr. 2013;143(12 Suppl):2066S-70S.
12. DR C. Physiology of insuline-like growth factor I. http://www uptodate com/contents/
physiology-of-insulin-like-growth-factor-i [Internet]. 2014 2/5/2014. Available from: http://www.
uptodate.com/contents/physiology-of-insulin-like-growth-factor-i.
13. Hawkes CP, Grimberg A. Measuring growth hormone and insulin-like growth factor-I in infants:
what is normal? Pediatr Endocrinol Rev. 2013;11(2):126-46.
14. Hansen-Pupp I, Lofqvist C, Polberger S, Niklasson A, Fellman V, Hellstrom A, et al. Influence of
insulin-like growth factor I and nutrition during phases of postnatal growth in very preterm
infants. Pediatr Res. 2011;69(5 Pt 1):448-53.
15. Kurtoglu S, Kondolot M, Mazicioglu MM, Hatipoglu N, Akin MA, Akyildiz B. Growth hormone,
insulin like growth factor-1, and insulin-like growth factor-binding protein-3 levels in the
neonatal period: a preliminary study. J Pediatr Endocrinol Metab. 2010;23(9):885-9.
16. Martos-Moreno GA, Barrios V, Saenz de PM, Pozo J, Dorronsoro I, Martinez-Biarge M, et al.
Influence of prematurity and growth restriction on the adipokine profile, IGF1, and ghrelin
levels in cord blood: relationship with glucose metabolism. Eur J Endocrinol. 2009;161(3):381-9.
17. Ohkawa N, Shoji H, Kitamura T, Suganuma H, Yoshikawa N, Suzuki M, et al. IGF-I, leptin and
active ghrelin levels in very low birth weight infants during the first 8 weeks of life. Acta
Paediatr. 2010;99(1):37-41.
18. van de Lagemaat M, Rotteveel J, Heijboer AC, Lafeber HN, van Weissenbruch MM. Growth in
preterm infants until six months postterm: the role of insulin and IGF-I. Horm Res Paediatr.
2013;80(2):92-9.
40
19. Giapros VI, Schiza V, Challa AS, Pantou C, Theocharis PD, Andronikou SK. Serum insulin-like
growth factor I (IGF-I), IGF-binding proteins-1 and -3, and postnatal growth of late preterm
infants. Horm Metab Res. 2012;44(11):845-50.
20. Wang XL, Ge MR, Wu WY, Zhang J. [Insulin-like growth factor 1 levels and their association
with growth and development in infants aged 1-24 months]. Zhongguo Dang Dai Er Ke Za Zhi.
21.
2010;12(6):459-61.
Iniguez G, Ong K, Bazaes R, Avila A, Salazar T, Dunger D, et al. Longitudinal changes in insulin- 2
like growth factor-I, insulin sensitivity, and secretion from birth to age three years in small-for-
gestational-age children. J Clin Endocrinol Metab. 2006;91(11):4645-9.
22. Larnkjaer A, Ingstrup HK, Schack-Nielsen L, Hoppe C, Molgaard C, Skovgaard IM, et al. Early
programming of the IGF-I axis: negative association between IGF-I in infancy and late
adolescence in a 17-year longitudinal follow-up study of healthy subjects. Growth Horm IGF
Res. 2009;19(1):82-6.
23. Hyun SE, Lee BC, Suh BK, Chung SC, Ko CW, Kim HS, et al. Reference values for serum levels of
insulin-like growth factor-I and insulin-like growth factor binding protein-3 in Korean children
and adolescents. Clin Biochem. 2012;45(1-2):16-21.
24. Ong KK, Langkamp M, Ranke MB, Whitehead K, Hughes IA, Acerini CL, et al. Insulin-like growth
factor I concentrations in infancy predict differential gains in body length and adiposity: the
Cambridge Baby Growth Study. Am J Clin Nutr. 2009;90(1):156-61.
25. Chellakooty M, Juul A, Boisen KA, Damgaard IN, Kai CM, Schmidt IM, et al. A prospective study
of serum insulin-like growth factor I (IGF-I) and IGF-binding protein-3 in 942 healthy infants:
associations with birth weight, gender, growth velocity, and breastfeeding. The Journal of
clinical endocrinology and metabolism. 2006;91(3):820-6.
26. Socha P, Grote V, Gruszfeld D, Janas R, Demmelmair H, Closa-Monasterolo R, et al. 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.
27. Miller BS, Kroupina MG, Mason P, Iverson SL, Narad C, Himes JH, et al. Determinants of
catch-up growth in international adoptees from eastern europe. Int J Pediatr Endocrinol.
2010;2010:107252.
28. Ong KK. Catch-up growth in small for gestational age babies: good or bad? Current opinion in
endocrinology, diabetes, and obesity. 2007;14(1):30-4.
29. de Jong M, Lafeber HN, Cranendonk A, van Weissenbruch MM. Components of the metabolic
syndrome in early childhood in very-low-birth-weight infants. Horm Res Paediatr. 2014;81(1):43-9.
30. Singhal A, Cole TJ, Fewtrell M, Deanfield J, Lucas A. Is slower early growth beneficial for long-
term cardiovascular health? Circulation. 2004;109(9):1108-13.
31. Leunissen RW, Kerkhof GF, Stijnen T, Hokken-Koelega A. Timing and tempo of first-year rapid
growth in relation to cardiovascular and metabolic risk profile in early adulthood. JAMA.
2009;301(21):2234-42.
32. Hovi P, Andersson S, Raikkonen K, Strang-Karlsson S, Jarvenpaa AL, Eriksson JG, et al. Ambulatory
blood pressure in young adults with very low birth weight. J Pediatr. 2010;156(1):54-9.
33. Lapillonne A, Griffin IJ. Feeding preterm infants today for later metabolic and cardiovascular
outcomes. J Pediatr. 2013;162(3 Suppl):S7-16.
34. Madsen AL, Larnkjaer A, Molgaard C, Michaelsen KF. IGF-I and IGFBP-3 in healthy 9 month
old infants from the SKOT cohort: breastfeeding, diet, and later obesity. Growth Horm IGF Res.
2011;21(4):199-204.
weight. Journal of Pediatric Endocrinology and Metabolism. 2012;25(9-10):951-5.
36. Parkinson JR, Hyde MJ, Gale C, Santhakumaran S, Modi N. Preterm birth and the metabolic
syndrome in adult life: a systematic review and meta-analysis. Pediatrics. 2013;131(4):e1240-e63.
41
Chapter 2
37. Bikle DD, Wang Y. Insulin-like growth factor-I and bone. IBMS BoneKEy. 2011;8(7):328-41.
38. Ali AT, Hochfeld WE, Myburgh R, Pepper MS. Adipocyte and adipogenesis. Eur J Cell Biol.
2013;92(6-7):229-36.
39. Eliakim A, Nemet D, Ahmad I, Zaldivar F, Koppel R, Grochow D, et al. Growth factors,
inflammatory cytokines and postnatal bone strength in preterm infants. J Pediatr Endocrinol
Metab. 2009;22(8):733-40.
40. van de Lagemaat M, Rotteveel J, van Weissenbruch MM, Lafeber HN. Small-for-gestational-age
preterm-born infants already have lower bone mass during early infancy. Bone. 2012;51(3):441-6.
41. Fewtrell MS. Does early nutrition program later bone health in preterm infants? American
Journal of Clinical Nutrition. 2011;94(6 SUPPL):1870S-3S.
42. van Bunderen CC, Oosterwerff MM, van Schoor NM, Deeg DJ, Lips P, Drent ML. Serum IGF1,
metabolic syndrome, and incident cardiovascular disease in older people: a population-based
study. Eur J Endocrinol. 2013;168(3):393-401.
43. Hansen-Pupp I, Hovel H, Lofqvist C, Hellstrom-Westas L, Fellman V, Huppi PS, et al. Circulatory
insulin-like growth factor-I and brain volumes in relation to neurodevelopmental outcome in
very preterm infants. Pediatr Res. 2013;74(5):564-9.
44. Euser AM, Finken MJ, Kharagjitsingh AV, Alizadeh BZ, Roep BO, Meulenbelt I, et al. IGF1
promoter polymorphism and cranial growth in individuals born very preterm. Horm Res
Paediatr. 2011;76(1):27-34.
45. Hansen-Pupp I, Hovel H, Hellstrom A, Hellstrom-Westas L, Lofqvist C, Larsson EM, et al.
Postnatal decrease in circulating insulin-like growth factor-I and low brain volumes in very
preterm infants. J Clin Endocrinol Metab. 2011;96(4):1129-35.
46. Okuma C, Isabel Hernandez M, Rodriguez P, Flores R, Avila A, Cavada G, et al. Microstructural
brain and multivoxel spectroscopy in very low birth weight infants related to insulin-like growth
factor concentration and early growth. Hormone Research in Paediatrics. 2013;79(4):197-207.
47. Lin S, Fan LW, Rhodes PG, Cai Z. Intranasal administration of IGF-1 attenuates hypoxic-ischemic
brain injury in neonatal rats. Exp Neurol. 2009;217(2):361-70.
48. Pang Y, Campbell L, Zheng B, Fan L, Cai Z, Rhodes P. Lipopolysaccharide-activated microglia
induce death of oligodendrocyte progenitor cells and impede their development. Neuroscience.
2010;166(2):464-75.
49. Hellgren G, Han W, Wang X, Lofqvist C, Hagberg H, Mallard C, et al. Safety aspects of
longitudinal administration of IGF-I/IGFBP-3 complex in neonatal mice. Growth Horm IGF Res.
2011;21(4):205-11.
50. Wang X, Hellgren G, Lofqvist C, Li W, Hellstrom A, Hagberg H, et al. White matter damage after
chronic subclinical inflammation in newborn mice. J Child Neurol. 2009;24(9):1171-8.
51. Pang Y, Zheng B, Campbell LR, Fan LW, Cai Z, Rhodes PG. IGF-1 can either protect against or
increase LPS-induced damage in the developing rat brain. Pediatr Res. 2010;67(6):579-84.
52. Ley D, Hansen-Pupp I, Niklasson A, Domellof M, Friberg LE, Borg J, et al. Longitudinal infusion
of a complex of insulin-like growth factor-I and IGF-binding protein-3 in five preterm infants:
pharmacokinetics and short-term safety. Pediatr Res. 2013;73(1):68-74.
53. Agostoni C, Buonocore G, Carnielli VP, De CM, Darmaun D, Decsi T, et al. Enteral nutrient supply
for preterm infants: commentary from the European Society of Paediatric Gastroenterology,
Hepatology and Nutrition Committee on Nutrition. J Pediatr Gastroenterol Nutr. 2010;50(1):85-91.
54. Koletzko B, Goulet O, Hunt J, Krohn K, Shamir R. 1. Guidelines on Paediatric Parenteral Nutrition of
the European Society of Paediatric Gastroenterology, Hepatology and Nutrition (ESPGHAN) and
the European Society for Clinical Nutrition and Metabolism (ESPEN), Supported by the European
Society of Paediatric Research (ESPR). J Pediatr Gastroenterol Nutr. 2005;41 Suppl 2:S1-87.
55. Yeung MY, Smyth JP. Nutritionally regulated hormonal factors in prolonged postnatal growth
retardation and its associated adverse neurodevelopmental outcome in extreme prematurity.
Biol Neonate. 2003;84(1):1-23.
42
56. Embleton NE, Pang N, Cooke RJ. Postnatal malnutrition and growth retardation: an inevitable
consequence of current recommendations in preterm infants? Pediatrics. 2001;107(2):270-3.
57. Senterre T, Rigo J. Reduction in postnatal cumulative nutritional deficit and improvement of
growth in extremely preterm infants. Acta Paediatr. 2012;101(2):e64-e70.
58. Hay WW, Thureen P. Protein for preterm infants: how much is needed? How much is enough?
59.
How much is too much? Pediatr Neonatol. 2010;51(4):198-207.
Burattini I, Bellagamba MP, Spagnoli C, D’Ascenzo R, Mazzoni N, Peretti A, et al. Targeting 2.5
2
versus 4 g/kg/day of amino acids for extremely low birth weight infants: a randomized clinical
trial. J Pediatr. 2013;163(5):1278-82.
60. Yang S, Lee BS, Park HW, Choi YS, Jeong SH, Kim JH, et al. Effect of high vs standard early
parenteral amino acid supplementation on the growth outcomes in very low birth weight
infants. JPEN J Parenter Enteral Nutr. 2013;37(3):327-34.
61. Alan S, Atasay B, Cakir U, Yildiz D, Kilic A, Kahvecioglu D, et al. An intention to achieve better
postnatal in-hospital-growth for preterm infants: Adjustable protein fortification of human
milk. Early Hum Dev. 2013;89(12):1017-23.
62. Cormack BE, Bloomfield FH. Increased protein intake decreases postnatal growth faltering in
ELBW babies. Archives of disease in childhood Fetal and neonatal edition. 2013;98(5):F399-F404.
63. Olsen IE, Harris CL, Lawson ML, Berseth CL. Higher Protein Intake Improves Length, Not
Weight, z Scores in Preterm Infants. J Pediatr Gastroenterol Nutr. 2014;58(4):409-16.
64. Stoltz SE, Ohlund I, Ahlsson F, Engstrom E, Fellman V, Hellstrom A, et al. Nutrient intakes
independently affect growth in extremely preterm infants: Results from a population-based
study. Acta Paediatrica, International Journal of Paediatrics. 2013;102(11):1067-74.
65. Peiler A, Woelfle J, Stutte S, Schreiner F, Bartmann P, Gohlke B. Postnatal nutrition in extremely
low birth weight infants and its impact on growth until the age of 6 years. Acta Paediatr.
2013;103(2):e61-8.
66. Scattolin S, Gaio P, Betto M, Palatron S, De TF, Intini F, et al. Parenteral amino acid intakes:
Possible influences of higher intakes on growth and bone status in preterm infants. Journal of
Perinatology. 2013;33(1):33-9.
67. Costa-Orvay JA, Figueras-Aloy J, Romera G, Closa-Monasterolo R, Carbonell-Estrany X. The
effects of varying protein and energy intakes on the growth and body composition of very low
birth weight infants. Nutr J. 2011;10:140.
68. Amesz EM, Schaafsma A, Cranendonk A, Lafeber HN. Optimal growth and lower fat mass in
preterm infants fed a protein-enriched postdischarge formula. J Pediatr Gastroenterol Nutr.
2010;50(2):200-7.
69. Roggero P, Gianni ML, Amato O, Liotto N, Morlacchi L, Orsi A, et al. Growth and fat-free
mass gain in preterm infants after discharge: A randomized controlled trial. Pediatrics.
2012;130(5):e1215-e21.
70. Embleton ND, Cooke RJ. Protein requirements in preterm infants: effect of different levels of
protein intake on growth and body composition. Pediatr Res. 2005;58(5):855-60.
71. Van Den Akker CH, te Braake FW, Rovekamp-Abels WW, Van Goudoever JB. Quality of amino
acid solutions for preterm infants. Pediatrics. 2008;121(4):865-6.
72. Ergenekon E, Hirfanonulllu I, Soysal S, Gucuyener K, Bas V, Turan O, et al. Physical and mental
developmental effects of additional enteral protein supplementation in preterm newborns.
Early Hum Dev. 2010;86 SUPPL. 1:S85-S6.
73. Morgan C, McGowan P, Herwitker S, Hart AE, Turner MA. Postnatal head growth in preterm
infants: a randomized controlled parenteral nutrition study. Pediatrics. 2014;133(1):e120-e8.
74. Stutte S, Peiler A, Schreiner F, Bartmann P, Woelfle J, Gohlke B. Is postnatal nutrition associated
with growth parameters until at the age of 6 years in infants born with extremely low birth
weight? Hormone Research in Paediatrics. 2010;74:32.
43
Chapter 2
75. Young L, Morgan J, McCormick FM, McGuire W. Nutrient-enriched formula versus standard
term formula for preterm infants following hospital discharge. Cochrane Database Syst Rev.
2012;3:CD004696.
76. Young L, Embleton ND, McCormick FM, McGuire W. Multinutrient fortification of human
breast milk for preterm infants following hospital discharge. Cochrane database of systematic
reviews. 2013;2:CD004866.
77. Stephens BE, Walden RV, Gargus RA, Tucker R, McKinley L, Mance M, et al. First-week protein
and energy intakes are associated with 18-month developmental outcomes in extremely low
birth weight infants. Pediatrics. 2009;123(5):1337-43.
78. Cormack BE, Bloomfield FH, Dezoete A, Kuschel CA. Does more protein in the first week of life
change outcomes for very low birthweight babies? Journal of Paediatrics and Child Health.
2011;47(12):898-903.
79. Biasini A, Marvulli L, Neri E, China M, Stella M, Monti F. Growth and neurological outcome in
ELBW preterms fed with human milk and extra-protein supplementation as routine practice:
do we need further evidence? J Matern Fetal Neonatal Med. 2012;25 Suppl 4:72-4.
80. Pfister KM, Gray HL, Miller NC, Demerath EW, Georgieff MK, Ramel SE. Exploratory study of
the relationship of fat-free mass to speed of brain processing in preterm infants. Pediatr Res.
2013;74(5):576-83.
81. Vucetic Z, Totoki K, Schoch H, Whitaker KW, Hill-Smith T, Lucki I, et al. Early life protein restriction
alters dopamine circuitry. Neuroscience. 2010;168(2):359-70.
82. Isaacs EB, Gadian DG, Sabatini S, Chong WK, Quinn BT, Fischl BR, et al. The effect of early
human diet on caudate volumes and IQ. Pediatr Res. 2008;63(3):308-14.
83. Choi S, Disilvio B, Fernstrom MH, Fernstrom JD. Meal ingestion, amino acids and brain
neurotransmitters: effects of dietary protein source on serotonin and catecholamine synthesis
rates. Physiol Behav. 2009;98(1-2):156-62.
84. de Kieviet JF, Oosterlaan J, Vermeulen RJ, Pouwels PJ, Lafeber HN, van Elburg RM. Effects
of glutamine on brain development in very preterm children at school age. Pediatrics.
2012;130(5):e1121-e7.
care. 2012;15(3):285-92.
44
45
Associations between bronchopulmonary 3
CHAPTER 3
dysplasia, Insulin-like growth factor I

and nutrition
Dana FJ Yumani, Floor H Walschot, Harrie N Lafeber, Mirjam M van Weissenbruch.

Submitted
Chapter 3
Abstract
Objective
Insulin-like growth factor I (IGF-I) has been suggested as an important factor in the
pathogenesis of bronchopulmonary dysplasia (BPD). In turn, nutrition has been
associated with IGF-I levels and could be of importance in the pathogenesis of
BPD. This study aimed to explore the association between nutrition, the IGF-I axis
and the occurrence BPD.
Design
An observational study was conducted from 2015 to 2018 in a level III NICU. Eighty-
six infants were included (44 male, mean gestational age 29.0 weeks (standard
deviation 1.7 weeks) Serum IGF-I (µgram/L) and Insulin-like growth factor binding
protein 3 (IGFBP-3; mg/L) was measured at birth and at 2, 4 and 6 weeks postnatal
age and BPD was diagnosed at 36 weeks postmenstrual age. (https://www.
trialregister.nl/trial/5171)
Results
Twenty-nine out of 86 infants were diagnosed with BPD. The odds ratio (OR) for
the occurrence of BPD was 0.68, 95% CI 0.48 – 0.96 for every µgram/L per week
increase in IGF-I, corrected for gestational age. The change in IGF-I in µgram/L/
week, gestational age in weeks and a week of predominant donor human milk
were associated with the occurrence of BPD in multivariable analysis (respectively
OR 0.63 (0.43 – 0.92), OR 0.44 (0.26 – 0.76) and 7.6 (1.2 – 50.4)). IGFBP-3 was not
associated with the occurrence of BPD in multivariable analysis.
Conclusion
A slow increase in IGF-I levels and a lower gestational age increase the odds of BPD.
Donor human milk might also increase the odds of BPD and should be further
explored.
48
Bronchopulmonary dysplasia, IGF-I & nutrition
Introduction
With reported incidences up to 30%, bronchopulmonary dysplasia (BPD) is one of
the most frequent complications following preterm birth. 1-3 Lately, more insight
has been gained in the pathogenesis of BPD.4 However, therapeutic options to
prevent BPD and current drug therapies only show a moderate reduction of the
incidence of BPD. 1 2
Previous research has shown that Insulin-like Growth Factor I (IGF-I) plays a role in
the development and differentiation of several different lung cells 5 and has anti- 3
inflammatory and anti-oxidative effects. 6-8 Therefore, IGF-I could play an important
role in the development of BPD. Indeed, a recent trial targeting the prevention of
retinopathy of prematurity (ROP) by IGF-1 infusion showed a concomitant decline
in BPD occurrence. 9
Various factors, however, e.g. nutritional intake, genetic predisposition, and

comorbidities have been associated with IGF-I levels 10 and could therefore be of
importance in the association between IGF-I and the occurrence of BPD. Price
and colleagues demonstrated that protein intake was associated with Insulin-
like Growth Factor Binding Protein 3 (IGFBP-3).11 Nutrition might be of particular
interest to further investigate, because, in contrast to more invasive interventions
such as IGF-I administration, it is generally feasible to implement nutritional
interventions.
This study aims to explore whether IGF-I levels and the rate of increase in IGF-I
are associated with the occurrence of BPD. Moreover it is explored whether these
potential associations are impacted by the type nutrition.
Methods
Study population
The study population consisted of preterm infants who were admitted to the
neonatal intensive care unit of Amsterdam UMC, location VU University medical
center. The infants were born between 2015 and 2018, with a gestational age of
24 to 32 weeks and participated in the NUTRIE study, a longitudinal observational
study on nutrition in relation to the endocrine regulation of preterm growth.12
Power calculations were done for the primary outcome of the NUTRIE study. No
power calculations were done for the results presented in this paper.
Informed consent was obtained in the first week of life and participants were
followed up from birth to two years corrected age. Infants with substantial congenital
anomalies based on a chromosomal disorder or syndrome were excluded. Infants
could be enrolled in other studies during inclusion. These studies did not have
overlapping outcomes. The study was approved by the medical research ethics
49
Chapter 3
committee of the VU University Medical Center (approval number 2014.491) and

was conducted according to the good clinical practice guidelines and in line with
the Declaration of Helsinki. The study was registered at the Dutch Trial Register.12
Measures of endocrine parameters

Blood was drawn every other week up to 36 weeks postmenstrual age (PMA) for
the measurement of IGF-I and IGFBP-3. Analyses of IGF-I were conducted using
chemiluminescence immunoassays (LIAISON®, DiaSorin, Italy) with an intra-assay
percent coefficient of variation (%CV) of 8% and an inter-assay %CV of 7%. IGFBP-3
was analyzed with a sandwich ELISA (DRG Instruments GmbH, Germany) with an
intra-assay %CV of 5% and an inter-assay %CV of 13%.
Morbidities
BPD was defined as the need for supplemental oxygen at 36 weeks PMA. Grade 1
BPD was defined as respiratory support through a nasal cannula with a flow of 2L/
minute or less, grade 2 as a flow of 2L/minute or more through a nasal cannula or
noninvasive positive airway pressure, grade 3 as invasive mechanical ventilation.13 In
addition data on the following comorbidities was collected (see web appendices for
definitions): necrotizing enterocolitis (NEC), late-onset sepsis (LOS), intraventricular
hemorrhage (IVH), ROP, post hemorrhagic ventricle dilatation (PHVD), infant
respiratory distress syndrome (IRDS) and patent ductus arteriosus (PDA).
Nutrition
Nutrition was initially provided through total parenteral nutrition and minimal
enteral feeding.14 Within seven to ten days, 160 ml per kg per day human milk was
scheduled to be given fortified with breast milk fortifier. Either own mother’s milk
or pasteurized donor human milk was given. Donor human milk underwent Holder
pasteurization (heated at 62.0°C to 62.5°C for 30 minutes, followed by fast cooling
to under 4°C).15 If necessary, due to poor growth, up to 2% Nenatal Human Milk
Protein Fortifier (Nutricia, Wageningen, The Netherlands) or fat emulsion Calogen
(Nutricia, Wageningen, The Netherlands), was added to the fortified human milk.
If own mother’s milk was not available, donor human milk was supplied up to 32
weeks PMA, followed by preterm starters formula till discharge home. If parents
declined the use of donor human milk, infants were fed preterm starters formula
from birth.
Data was obtained from hospital records to calculate the daily macronutrient
intake based on the reference values for human milk. (table 1) Macronutrient
intake, type of enteral nutrition, and parenteral nutrition were assessed as potential
confounders. Type of enteral nutrition was defined as predominant own mother’s
milk if at least 60% of the enteral intake consisted of own mother’s milk. Likewise
predominant donor human milk and predominant formula feeding were defined
50
as at least 60% of enteral intake consisting of respectively donor human milk or

formula.
Table 1. Reference values used for the nutritional composition of human milk per 100 ml
OMM OMM + BMF DHM DHM + BMF

(4.4g/100ml) (4.4g/100ml)
Energy (kcal) 68.5 83.8 60 75
Protein (g)
Carbohydrates (g)
1.5
7.3
2.6
10.0
0.8
7.5
1.9
10.2
3
Fat (g) 3.3 3.3 2.9 2.9
BMF: Breast milk fortifier, DHM: donor human milk, OMM: own mother’s milk
Reference values for OMM were derived from meta-analyses.32 33 Donor human milk composition was
based on analyses of the donor milk batches administered to the first 23 study participants.
Statistical analysis
Comparisons between infants with and without BPD were analyzed, depending on
the data distribution, with either the independent sample T-test or Mann-Whitney
U test and the Chi-square test or the Fisher-Exact test. Logistic regression was used
to correct hypothesized predictors and potential confounders for gestational age.
A final prediction model was designed which included every factor which was
significantly associated with BPD after correction for gestational age. A backward
regression analysis was used to design this prediction model.
A mixed model with a random intercept and slope, IGF-I (µgram/L) as the dependent
variable and postmenstrual age (weeks) as a covariate was used to estimate the
mean change of IGF-I in µgram/L/week per infant. The same was done for IGFBP-3
levels.
Analyses were conducted using IBM® SPSS® Statistics 22 for Windows (IBM Corp.,
Armonk, NY). Two-sided statistical significance was assumed at p-values less than
0.05.
Results
Eighty-six infants were included in the analyses. 14 (table 2) Twenty-nine infants
were classified as having BPD: 14 infants had grade 1 BPD, 15 infants had grade 2
BPD and there were no cases of grade 3 BPD.
51
Chapter 3
Table 2. Baseline characteristics
All BPD No BPD P value

(n=86)a (n=29) (n=57)
Gender, n male (%) 44 (51.2) 17 (58.6) 27 (47.4) 0.324b
Ethnicity, n white (%) 65 (75.6) 22 (75.9) 43 (75.4) 0.966b
Gestational age (weeks), mean (SD) 29.0 (1.7) 27.9 (1.7) 29.6 (1.5) <0.001c
Birthweight (g), mean (SD) 1217 (312) 1055 (271) 1298 (301) 0.001c
Birthweight SDS, mean (SD) 0.0 (0.7) 0.0 (0.7) 0.0 (0.7) 0.765c
Birthweight SDS < -1.3, n (%) 3 (3.5) 1 (3.4) 2 (3.5) 1.000d
Antenatal steroids e, n (%) 56 (69.1) 17 (63.0) 39 (72.2) 0.395b
Postnatal steroids r, n (%) 8 (9.3) 5 (17.2) 3 (5.3) 0.113d
Ventilation days, median (IQR) 0 (0.0 - 2.0 (0.0 – 0.0 (0.0 – 0.007e
5.0) 9.0) 2.3)
IRDS, n (%) <0.001b

IRDS stage I - II 24 (27.9) 14 (48.3) 10 (17.5)
IRDS stage III - IV 19 (22.1) 9 (31.0) 10 (17.5)
ROP, n (%) 0.532d
No ROP 81 (94.2) 27 (93.1) 54 (94.7)
ROP stage I 4 (4.7) 1 (3.4) 3 (5.3)
ROP stage III 1 (1.2) 1 (3.4) 0 (0.0)
PDA requiring treatment, n (%) 8 (9.3) 5 (17.2) 3 (5.3) 0.113d
NEC, n (%) 6 (7.0) 4 (13.8) 2 (3.5) 0.173d
LOS, n (%) 30 (34.9) 14 (48.3) 16 (28.1) 0.063b
IVH grade > III, n (%) 3 (3.5) 3 (10.3) 0 (0) 0.036d
PHVD, n (%) 8 (9.3) 3 (10.3) 5 (8.8) 1.000d
PVL, n (%) 3 (3.5) 1 (3.4) 2 (3.5) 1.000d
a
See web appendices for inclusion flow chart
b
Chi-square Test
c
Independent sample T-test
d
Fisher-Exact test
e
Mann-whitney U test
f
Antenatal steroids were defined as at least 2 doses of bethametason
g
Postnatal steroids were defined as at least 3 days of hydrocortisone treatment
Values in bold are statistically significant
BPD: Bronchopulmonary dysplasia, IRDS: Infant respiratory stress syndrome, IVH: intraventricular
hemorrhage, LOS: Late-onset sepsis; NEC: Necrotizing enterocolitis; PDA: patent ductus arteriosus,
PHVD: post-hemorrhagic ventricular dilatation, PVL: periventricular leukomalacia, ROP: retinopathy of
prematurity, SD: standard deviation, SDS: standard deviation score
52
Infants who developed BPD, had a lower gestational age and accordingly a lower
birth weight, as well as a longer duration of mechanical ventilation and a higher
incidence of comorbidities. (Table 2) However, after correction for gestational age,
IRDS remained the only comorbidity significantly associated with the occurrence
of BPD: OR 4.2 (95% CI 1.4 – 13.0, p = 0.012).
In infants who developed BPD the mean change in IGF-I was 2.8 + 2.2 µgram/L per
week from birth through to 34 weeks PMA versus 4.5 + 1.8 µgram/L per week in
infants without BPD. (Figure 1) Between birth and 34 weeks PMA every µgram/L
per week increase in IGF-I lowered the odds of BDP 0.68 times (95% CI 0.48 − 0.96,
3
p = 0.026; corrected for gestational age). The change in IGF-I between birth and 36
weeks PMA was not significantly different between infants with and without BPD.
(Figure 1)
Figure 1. Mean change in IGF-I and IGFBP-3 serum levels in infants with and without BPD up to 36 weeks
postmenstrual age
The mean change in IGF-I and IGFBP-3 was estimated for every individual using a mixed model.
53
Chapter 3
BPD: bronchopulmonary dysplasia, IGFBP-3: Insulin-like growth factor binding protein 3, IGF-I: Insulin-like
growth factor I
The mean change in IGFBP-3 from birth through to 34 weeks PMA was 0.3 + 0.01
mg/L per week in infants who developed BPD compared to 0.4 + 0.01 mg/L per
week in infants who did not develop BPD. After correction for gestational age, the
change in IGFBP-3 did not predict the occurrence of BPD. (Figure 1)
Infants with and without BPD had a comparable macronutrient and caloric intake
during hospitalization. (Figure 2)
Figure 2. Nutritional intake in preterm infants with and without BPD up to 6 weeks postnatal age
After correction for gestational age there were no significant differences in intake in infants with or
without BPD.
Both groups were mainly fed own mother’s milk. (Figure 3) Corrected for
gestational age, the percentage of own mother’s milk, donor human milk and
formula did not significantly differ between those with and without BPD. Median
54
duration of predominant donor human milk feeding was 4 days (IQR 0 – 11.5) in
infants who did not develop BPD compared to 7 days (2.5 – 18.5) in infants who
developed BPD (not significant after correction for gestational age). Throughout
hospitalization 13 infants were predominantly fed donor human milk for at least
a week. Of these infants 10 (76.9%) developed BPD compared to 19 out of 73 (26%)
of infants who were predominantly fed donor milk for a shorter period (p 0.001).
Corrected for gestational age, being predominantly fed donor milk for at least one
week increased the odds of BPD by 4.6 (95% CI 1.03 - 20.55, p = 0.046) compared to
infants fed donor milk for less than a week. 3
Figure 3. Enteral intake in preterm infants with and without BPD during hospitalization
After correction for gestational age there were no significant differences the type of enteral nutrition in
infants with or without BPD.
Median duration of parenteral nutrition was 12 days (IQR 9 – 22) in infants who
developed BPD compared to 10 days (IQR 8 – 13) in infants who did not develop
BPD (not significant after correction for gestational age). In the first two weeks of
life infants who developed BPD had a larger proportion of their intake as parenteral
nutrition compared to infants who did not develop BPD: 37.9 % (95% CI 29.4 – 49.5)
versus 32.3% (95% CI 23.2 – 42.7), (not significant after correction for gestational age).
The change in IGF-I in μgram/L per week, gestational age in weeks, IRDS (yes or
no) and predominant donor human milk use for at least one week compared to
less than a week were included in our final predictive model. A slow increase in
IGF-I, lower gestational age and at least one week of predominant donor human
milk feeding compared to less than a week of predominant donor human milk
feeding increased the odds of BPD. (table 3) The macronutrient and caloric intake
did significantly impact this association. (see web appendices)
55
Chapter 3
Table 3. Multivariable logistic regression for the occurrence of BPD
B (SE) p-value Odds ratio (95% CI)

Included variables
Constant 23.9 (8.0) 0.003
Change in IGF-I (µgram/L per week) -0.5 (0.2) 0.018 0.63 (0.43 – 0.92)
Gestational age at birth (weeks) -0.8 (0.3) 0.003 0.44 (0.26 – 0.76)
Predominant donor human milk for at least 2.0 (1.0) 0.035 7.6 (1.2 – 50.4)
1 weeka
R² = 0.358 (Cox & Snell), 0.498 (Nagelkerke). Model χ² (3) = 26.63, p < 0.001
a
Predominant donor human milk for at least one week compared to less than one week predominant
donor human milk feeding. Predominant donor milk feeding was defined as at least 60% of total enteral
intake consisting of donor human milk.
Variables removed in backward regression: IRDS
IGF-I: Insulin-like growth factor I, IRDS: Infant respiratory distress syndrome
Discussion
This study showed that, in particular before 35 weeks PMA, low IGF-I levels and
a slower increase in IGF-I increased the odds of BPD in preterm infants. IGFBP-3
showed a similar pattern, but wasn’t significant in multivariable analysis. Gestational
age and donor human milk consumption were significant confounders in the
association between IGF-I and the occurrence of BPD.
In line with our findings, other studies have shown that preterm infants with BPD
have lower postnatal IGF-I levels and a slower increase in IGF-I, which was associated
with an increased risk of developing BPD. 16-18 Also, in cases of intrauterine growth
restriction IGF-I levels are decreased and the odds of BPD increase. 19 20 Previous
research has suggested that IGF-I levels are a representation of the grade of infant
immaturity, and the link between IGF-I and BPD may be an indirect connection 17.
However, after adjustment for gestational age and other potential confounders,
which significantly differed between infants with and without BPD, the association
remained significant. This makes a functional direct mechanism more plausible.
Potential pathways through which IGF-I influences the occurrence of BPD

A number of factors may play a role in the association between IGF-I and the
occurrence of BPD. IGF-I signaling influences the development and differentiation
of several types of lung cells 5. Lower IGF-I could therefore arrest lung development.
Furthermore, IGF-I is known to be an important factor in the lung injury and repair
process. During lung injury, IGF-I increases the proliferation of lung fibroblasts
and enhances collagen products 16 21. In addition, pre-and postnatal inflammation
contributes to lung injury and subsequently to the development of BPD. It has
been shown by previous studies that the incidence of BPD, independently
increases with postnatal sepsis due to increased oxidative stress, inflammation
56
and endothelial lung injury 3 22. IGF-I is known to have a protective effect on
inflammation and to show anti-oxidative effects, protecting cells from oxidative-
stress induced apoptosis 6-8. The protective effect on inflammation and the anti-
oxidative effects of IGF-I may be important in decreasing the risk to develop BPD.
In support of this, a recent study aiming to reduce the occurrence of retinopathy of
prematurity by administering human recombinant IGF-I, showed that an increase
in pro-inflammatory cytokines, e.g. Interleukin-6, is associated with a subsequent
decrease in IGF-1 and IGF-I administration reduces the occurrence of BPD. 9 23
3
Possible interactions between nutrition, IGF-I levels and the occurrence
of BPD
In line with others, in our study, infants with BPD had a comparable macronutrient
intake to infants without BPD. 16 Milanesi and colleagues recently reported that
infants who developed BPD received less than the recommended daily protein/
energy ratio. 24 Nevertheless in our population infants reached the recommend
daily intake by the second week of life. However, our study did show that infants
who were predominantly fed donor human milk for at least a week had a higher
odds of BPD compared to those who were predominantly fed donor human milk
for a shorter period. Hypothetically, this could be due to the lower energy and
macronutrient content of donor human milk compared to own mother’s milk.
Yet, in our study population nutrient intake did not alter the association between
donor human milk intake and the occurrence of BPD. (see web appendices) On
the other hand, it has been described that human milk contains IGF-I and human
milk IGF-I levels have been associated with growth in term infants. 25 Furthermore,
Holder pasteurization can reduce IGF-I levels up to 40%. 26 Therefore, it could be
hypothesized that donor human milk could lead to lower IGF-I levels in preterm
infants. In addition to a possible direct uptake of IGF-I from human milk, it has
been described that micronutrients and branched amino acids in human milk
may stimulate the infant’s IGF-I axis. 27 Pasteurization also has been reported to
affect these factors 28 and thus supports a potential negative effect of pasteurized
donor human milk on endogenous IGF-I levels. In the future novel techniques such
as High-Temperature-Short-Time pasteurization, high pressure processing and
ultraviolet-C irradiation may offer alternatives to Holder pasteurization. 29
However, in our study participants were mainly fed own mother’s milk. (figure 3)
The absolute quantity of donor human milk was small and did not correlate with
the occurrence of BPD. Therefore, our findings remain speculative.
Strengths and limitations

Data on actual nutrient intake, as opposed to the prescribed amount, was collected
prospectively and changes in IGF-I over time were calculated for every participant.
Nevertheless, blood samples were taken every other week leading to a relatively
low sample size in the analyses per week PMA.
57
Chapter 3
Our population showed a relatively high BPD incidence. The definition used for
BPD, categorizes BPD severity according to the respiratory support at 36 weeks
PMA which could lead to the inclusion of infants who had a relatively short overall
duration of respiratory support. 13 Previous studies and several clinics still use the
older definition criteria which define BPD as having had a need for supplemental
oxygen for at least 28 days. 30 Using the older criteria 8 infants no longer would
have been classified as having BPD, while 9 infants classified as not having BPD
would have been classified otherwise. Using the definition by Job et al. our results
still showed a similar pattern with regards to the relationship between IGF-I, BPD
and nutrition. In addition, the incidence in our population would have stayed
relatively high despite average antenatal corticosteroid use 31 and a short duration
of mechanical ventilation.
Moreover, only a limited number of extremely preterm infants were enrolled,

limiting the statistical power of analyses below 30 weeks PMA. Lastly, it remains
noteworthy that causality could not be examined due to the observational nature
of this study.
Conclusions
In conclusion, this study provided further insight into the relationship between
the development of the endocrine axis and the occurrence of BPD: low IGF-I
levels and a slower increase in IGF-I are associated with the development of BPD.
Furthermore, this study pointed towards a possible effect of donor human milk
on the development of BPD. However, more research is needed to investigate
whether there truly is an increased risk of BPD in infants fed donor human milk.
58
References
1. Jobe AH. Mechanisms of Lung Injury and Bronchopulmonary Dysplasia. American journal of
perinatology 2016;33(11):1076-8. doi: 10.1055/s-0036-1586107 [published Online First: 2016/09/08]
2. Shahzad T, Radajewski S, Chao CM, et al. Pathogenesis of bronchopulmonary dysplasia: when
inflammation meets organ development. Molecular and cellular pediatrics 2016;3(1):23. doi:
10.1186/s40348-016-0051-9 [published Online First: 2016/07/01]
3. Thekkeveedu RK, Guaman MC, Shivanna B. Bronchopulmonary dysplasia: A review of
pathogenesis and pathophysiology. Respiratory medicine 2017;132:170-77.
4. Mandell EW, Kratimenos P, Abman SH, et al. Drugs for the Prevention and Treatment of
Bronchopulmonary Dysplasia. Clin Perinatol 2019;46(2):291-310. doi: 10.1016/j.clp.2019.02.011
3
[published Online First: 2019/04/24]
5. Wang J, Dong W. Oxidative stress and bronchopulmonary dysplasia. Gene 2018;678:177-83. doi:
10.1016/j.gene.2018.08.031 [published Online First: 2018/08/12]
6. Rowland KJ, Choi PM, Warner BW. The role of growth factors in intestinal regeneration
and repair in necrotizing enterocolitis. Semin Pediatr Surg 2013;22(2):101-11. doi: 10.1053/j.
sempedsurg.2013.01.007
7. Xu L, Zhang W, Sun R, et al. IGF‑1 may predict the severity and outcome of patients with sepsis
and be associated with microRNA‑1 level changes. Experimental and therapeutic medicine
2017;14(1):797-804.
8. Tanner SM, Berryhill TF, Ellenburg JL, et al. Pathogenesis of necrotizing enterocolitis: modeling
the innate immune response. Am J Pathol 2015;185(1):4-16. doi: 10.1016/j.ajpath.2014.08.028
9. Ley D, Hallberg B, Hansen-Pupp I, et al. rhIGF-1/rhIGFBP-3 in Preterm Infants: A Phase 2
Randomized Controlled Trial. J Pediatr 2019;206:56-65 e8. doi: 10.1016/j.jpeds.2018.10.033
10. Engstrom E, Niklasson A, Wikland KA, et al. The role of maternal factors, postnatal nutrition,
weight gain, and gender in regulation of serum IGF-I among preterm infants. Pediatr Res
2005;57(4):605-10. doi: 10.1203/01.PDR.0000155950.67503.BC [published Online First: 2005/02/08]
11. Price WA, Lee E, Maynor A, et al. Relation between serum insulinlike growth factor-1, insulinlike
growth factor binding protein-2, and insulinlike growth factor binding protein-3 and nutritional
intake in premature infants with bronchopulmonary dysplasia. J Pediatr Gastroenterol Nutr
2001;32(5):542-9. doi: 10.1097/00005176-200105000-00010 [published Online First: 2001/06/29]
12. Yumani DF, Lafeber HN, van Weissenbruch M. Study protocol Nutrition in relation to the
endocrine regulation of preterm growth 2015 [updated 14-04-2019. Available from: https://
www.trialregister.nl/trial/5171 accessed 17-12-2021.
13. Jensen EA, Dysart K, Gantz MG, et al. The Diagnosis of Bronchopulmonary Dysplasia in Very
Preterm Infants. An Evidence-based Approach. Am J Respir Crit Care Med 2019;200(6):751-59.
doi: 10.1164/rccm.201812-2348OC [published Online First: 2019/04/18]
in Extremely and Very Preterm Infants During Hospitalisation. Nutrients 2020;12(3) doi: 10.3390/
nu12030675 [published Online First: 2020/03/07]
15. de Waard M, Mank E, van Dijk K, et al. Holder-Pasteurized Human Donor Milk: How
Long Can It Be Preserved? J Pediatr Gastroenterol Nutr 2018;66(3):479-83. doi: 10.1097/
MPG.0000000000001782 [published Online First: 2017/10/12]
16. Lofqvist C, Hellgren G, Niklasson A, et al. Low postnatal serum IGF-I levels are associated with
bronchopulmonary dysplasia (BPD). Acta paediatrica (Oslo, Norway : 1992) 2012;101(12):1211-6.
doi: 10.1111/j.1651-2227.2012.02826.x [published Online First: 2012/08/29]
17. Hellstrom A, Engstrom E, Hard AL, et al. Postnatal serum insulin-like growth factor I deficiency
is associated with retinopathy of prematurity and other complications of premature birth.
Pediatrics 2003;112(5):1016-20. [published Online First: 2003/11/05]
59
Chapter 3
18. Perez-Munuzuri A, Couce-Pico ML, Bana-Souto A, et al. Preclinical screening for retinopathy
of prematurity risk using IGF1 levels at 3 weeks post-partum. PloS one 2014;9(2):e88781. doi:
10.1371/journal.pone.0088781 [published Online First: 2014/02/14]
19. Gortner L, Reiss I, Hilgendorff A. Bronchopulmonary dysplasia and intrauterine growth
restriction. Lancet 2006;368(9529):28. doi: 10.1016/S0140-6736(06)68964-2 [published Online
First: 2006/07/04]
20. Kurata H, Ochiai M, Inoue H, et al. Inflammation in the neonatal period and intrauterine growth
restriction aggravate bronchopulmonary dysplasia. Pediatr Neonatol 2019;60(5):496-503. doi:
10.1016/j.pedneo.2018.11.007 [published Online First: 2019/07/16]
21. Capoluongo E, Ameglio F, Zuppi C. Insulin-like growth factor-I and complications of
prematurity: a focus on bronchopulmonary dysplasia. Clinical chemistry and laboratory
medicine 2008;46(8):1061-66.
22. Jensen EA, Schmidt B. Epidemiology of bronchopulmonary dysplasia. Birth defects research
Part A, Clinical and molecular teratology 2014;100(3):145-57. doi: 10.1002/bdra.23235 [published
Online First: 2014/03/19]
23. Klevebro S, Hellgren G, Hansen-Pupp I, et al. Elevated levels of IL-6 and IGFBP-1 predict low
serum IGF-1 levels during continuous infusion of rhIGF-1/rhIGFBP-3 in extremely preterm
infants. Growth Horm IGF Res 2019;50:1-8. doi: 10.1016/j.ghir.2019.11.001 [published Online First:
2019/11/23]
24. Milanesi BG, Lima PA, Villela LD, et al. Assessment of early nutritional intake in preterm infants
with bronchopulmonary dysplasia: a cohort study. Eur J Pediatr 2021 doi: 10.1007/s00431-020-
03912-0 [published Online First: 2021/01/04]
25. Kon IY, Shilina NM, Gmoshinskaya MV, et al. The study of breast milk IGF-1, leptin, ghrelin and
adiponectin levels as possible reasons of high weight gain in breast-fed infants. Ann Nutr
Metab 2014;65(4):317-23. doi: 10.1159/000367998 [published Online First: 2014/11/18]
26. Goelz R, Hihn E, Hamprecht K, et al. Effects of different CMV-heat-inactivation-methods
on growth factors in human breast milk. Pediatr Res 2009;65(4):458-61. doi: 10.1203/
PDR.0b013e3181991f18 [published Online First: 2009/01/08]
27. Hoeflich A, Meyer Z. Functional analysis of the IGF-system in milk. Best Pract Res Clin Endocrinol
Metab 2017;31(4):409-18. doi: 10.1016/j.beem.2017.10.002 [published Online First: 2017/12/10]
28. Peila C, Moro GE, Bertino E, et al. The Effect of Holder Pasteurization on Nutrients and
Biologically-Active Components in Donor Human Milk: A Review. Nutrients 2016;8(8) doi:
10.3390/nu8080477 [published Online First: 2016/08/05]
29. Moro GE, Billeaud C, Rachel B, et al. Processing of Donor Human Milk: Update and
Recommendations From the European Milk Bank Association (EMBA). Front Pediatr 2019;7:49.
doi: 10.3389/fped.2019.00049 [published Online First: 2019/03/16]
30. Jobe AH, Bancalari E. Bronchopulmonary dysplasia. Am J Respir Crit Care Med 2001;163(7):1723-
9. doi: 10.1164/ajrccm.163.7.2011060 [published Online First: 2001/06/13]
31. Travers CP, Clark RH, Spitzer AR, et al. Exposure to any antenatal corticosteroids and outcomes
in preterm infants by gestational age: prospective cohort study. BMJ 2017;356:j1039. doi: 10.1136/
bmj.j1039 [published Online First: 2017/03/30]
32. Gidrewicz DA, Fenton TR. A systematic review and meta-analysis of the nutrient content of
preterm and term breast milk. BMC Pediatr 2014;14:216. doi: 10.1186/1471-2431-14-216 [published
33. Boyce C, Watson M, Lazidis G, et al. Preterm human milk composition: a systematic literature
review. Br J Nutr 2016;116(6):1033-45. doi: 10.1017/S0007114516003007 [published Online First:
2016/08/16]
60
Supplemental material
Methods
Definitions of diseases included as clinical data
Necrotizing enterocolitis was defined according to the modified Bell criteria

(Walsh 1986).
I. Stage 1A Suspected NEC based on temperature instability, apnea, bradycardy,
or lethargy combined with gastric retention, abdominal distention, emesis, 3
or heme-positive stool and normal or mild intestinal dilation, or mild ileus on
radiography.
II. Stage 1B Suspected NEC based on temperature instability, apnea,
bradycardy, or lethargy combined with grossly bloody stool and normal or
mild intestinal dilation, or mild ileus on radiography.
III. Stage 2A definite NEC (mildly ill) based on temperature instability, apnea,
bradycardy, or lethargy combined with grossly bloody stool and absent
bowel sounds with or without abdominal tenderness and Intestinal dilation,
ileus, or pneumatosis intestinalis on radiography.
IV. Stage 2B definite NEC (moderately ill) based on temperature instability,
apnea, bradycardy, or lethargy, as well as mild metabolic acidosis and
thrombocytopenia, combined with grossly bloody stool and absent bowel
sounds, definite tenderness, with or without abdominal cellulitis or right
lower quadrant mass and Intestinal dilation, ileus, pneumatosis intestinalis
and ascites on radiography.
V. Stage 3A advanced NEC (severely ill, intact bowel) based on temperature
instability, severe apnea, bradycardia, hypotension, lethargy, respiratory
and metabolic acidosis, disseminated intravascular coagulation, and
neutropenia combined with grossly bloody stool and absent bowel sounds,
signs of peritonitis, marked tenderness, and abdominal distention, with or
without abdominal cellulitis or right lower quadrant mass and Intestinal
dilation, ileus, pneumatosis intestinalis and ascites on radiography.
VI. Stage 3B advanced NEC (severely ill, perforated bowel) based on
temperature instability, severe apnea, bradycardia, hypotension, lethargy,
respiratory and metabolic acidosis, disseminated intravascular coagulation,
and neutropenia combined with grossly bloody stool and absent
bowel sounds, signs of peritonitis, marked tenderness, and abdominal
distention, with or without abdominal cellulitis or right lower quadrant
mass and Intestinal dilation, ileus, pneumatosis intestinalis, ascites and
pneumoperitoneum on radiography.
Walsh MC, Kliegman RM. Necrotizing enterocolitis: treatment based on staging
criteria. Pediatr Clin North Am 1986;33(1):179-201. doi: 10.1016/s0031-3955(16)34975-6
61
Chapter 3
Late onset sepsis was defined as sepsis occurring 72 hours after birth with a positive
blood culture or clinical signs of sepsis and relevant antimicrobial treatment (Dong
2015).
Dong Y, Speer CP. Late-onset neonatal sepsis: recent developments. Arch Dis
Child Fetal Neonatal Ed 2015;100(3):F257-F63. doi: 10.1136/archdischild-2014-306213
Intraventricular hemorrhage and post hemorrhagic ventricle dilatation were

defined according to Papile (Papile 1978).
I. Grade 1: subependymal hemorrhage
II. Grade 2: intraventricular hemorrhage without ventricular dilatation
III. Grade 3: intraventricular hemorrhage with ventricular dilation
IV. Grade 4: intraventricular hemorrhage with parenchymal hemorrhage
Papile LA, Burstein J, Burstein R, et al. Incidence and evolution of subependymal
and intraventricular hemorrhage: a study of infants with birth weights less than
1,500 gm. J Pediatr 1978;92(4):529-34. doi: 10.1016/s0022-3476(78)80282-0 [published
Retinopathy of prematurity was defined according to the 2005 International

Classification of Retinopathy of Prematurity
I. Stage I: Demarcation line
II. Stage 2: Ridge
III. Stage 3: Extraretinal fibrovascular proliferation, i.e. neovascularization from
the ridge into the vitreous
IV. Stage 4: Partial retinal detachment
International Committee for the Classification of Retinopathy of P. The Interna-
tional Classification of Retinopathy of Prematurity revisited. Arch Ophthalmol
2005;123(7):991-9. doi: 10.1001/archopht.123.7.991 [published Online First: 2005/07/13]
Infant respiratory distress syndrome was graded according to Giedion (Giedion 1973).
I. Grade 1: Reticulogranular pattern
II. Grade 2: Reticulogranular patterin and air bronchogram
III. Grade 3: Reticulogranular pattern, air bronchogram and loss of distinct heart
borders
IV. Grade 4: Diffuse opacification and obscuration of the cardiac silhouette
Giedion A, Haefliger H, Dangel P. Acute pulmonary X-ray changes in hyaline
membrane disease treated with artificial ventilation and positive end-expiratory
pressure (PEP). Pediatr Radiol 1973;1(3):145-52. doi: 10.1007/BF00974058 [published
62
Patent ductus arteriosus was defined as a patent ductus on ultrasound which

was treated with medication or required surgical intervention. Treatment was
considered when an infant had respiratory distress, a patent ductus with left-right
shunting and:
• 3 characteristics of a moderate left-right shunting with a duct diameter of
at least 1.6 mm and a pulsatile flow pattern
OR
3
• 2 characteristics of a large left-right shunt and a duct diameter of at least 2
mm and a pulsatile flow pattern
Small left-right Moderate left-right Large left-right

shunt shunt shunt
Diameter of the ductus < 1.6 1.6 – 2.0 > 2.0
arteriosus (mm)
Flow pattern in the Closing Growing/pulsatile Pulsatile

ductus arteriosus
End diastolic flow in left < 0.20 0.20 – 0.40 > 0.40
pulmonary artery (m/s)
Flow in the aorta Antegrade flow No or minimally Reversed flow

descendens reversed flow
Flow speed in ductus > 2.0 1.0 – 2.0 < 1.0

arteriosus
Left atrial to aortic root < 1.4 1.4 – 1.6 > 1.6
ration
Source: local protocol Amsterdam University Medical Centers Neonatal Intensive Care Unit
63
Chapter 3
Results
Inclusion/drop-out flow chart
* The infant was born at 24 weeks gestational age and deceased in the third week of life. Reason of death
was a progressive PHVD, combined with a NEC stage IIA and severe pulmonary interstitial emphysema.
** All 86 participants were included in the analyses presented in the study. 6 infants discontinued study
participation before term age (study burden to high (n=5), unresponsive to term age follow-up invite), but
permission was given to continue data collection from the file and blood draws were done before study
discontinuation.
64
Association between donor human milk intake and BPD corrected for the nutrient
intake
Association between donor human milk, protein intake and BPD

Included variables
Constant -6.2 (2.7) 0.021
Predominant donor human milk for at least

1 weeka
2.7 (0.8) 0.001 15.4 (3.3 – 72.2)
3
Protein intake (g/kg/day) 1.4 (0.7) 0.051 4.0 (1.0 – 16.4)
a
Association between donor human milk, carbohydrate intake and BPD

Included variables
Constant -4.1 (3.6) 0.255
1 weeka
Carbohydrate intake (mg/kg/min) 0.3 (0.4) 0.394 1.4 (0.7 – 2.8)
R² = 0.140 (Cox & Snell), 0.194 (Nagelkerke). Model χ² (2) = 12.945, p 0.002
a
Association between donor human milk, fat intake and BPD

Included variables
Constant -2.8 (2.6) 0.283
1 weeka
Fat intake (g/kg/day) 0.4 (0.6) 0.496 1.5 (0.5 – 4.5)
a
65
Chapter 3
Association between donor human milk, caloric intake and BPD

Included variables
Constant -5.9 (3.8) 0.003
1 weeka
Caloric intake (kcal/kg/day) 0.0 (0.0) 0.198 1.0 (1.0 – 1.1)
a
Multivariable logistic regression for the occurrence of BPD including nutrient intake

Included variables
Constant 37.9 (13.9) 0.006
Change in IGF-I (µgram/L per week) -0.4 (0.2) 0.040 0.67 (0.45 – 0.98)
Gestational age at birth (weeks) -1.0 (0.4) 0.006 0.38 (0.19 – 0.76)
1 weeka
IRDS 1.5 (0.9) 0.111 4.47 (0.71-28.16)
Mean fat intake (g/kg/day) -2.3 (1.2) 0.066 0.10 (0.01 – 1.16)
a
When caloric, protein and carbohydrate intake were included in the model, the terms for the nutrient
intake and IRDS were removed in the backward regression, resulting in the same regression model as
depicted in table 3 of the manuscript.
66
67
4 CHAPTER 4
The course of IGF-1 levels and nutrient
intake in extremely and very preterm
infants during hospitalisation
Dana FJ Yumani, Alexandra K Calor and Mirjam M van Weissenbruch.

Nutrients. 2020 Mar 2;12(3):675. doi: 10.3390/nu12030675.
Chapter 4
Abstract
Background
Insulin-like growth factor 1 (IGF-1) plays an important role in the complex association
between nutrition, growth, and maturation in extremely and very preterm infants.
Nevertheless, in this population, research on associations between IGF-1 and
nutrition is limited. Therefore this study aimed to evaluate the possible associations
between the course of IGF-1 levels and nutrient intake between preterm birth and
36 weeks postmenstrual age (PMA).
Methods
87 infants born between 24 and 32 weeks gestational age were followed up to 36
weeks PMA. Actual daily macronutrient intake was calculated, and growth was
assessed weekly. IGF-1 was sampled from umbilical cord blood at birth and every
other week thereafter.
Results
There was an inverse relationship between the amount of parenteral nutrition in
the second week of life and IGF-1. Total protein, fat, and carbohydrate intake, as
well as total energy intake, primarily showed a positive association with IGF-1 levels,
particularly between 30 and 33 weeks PMA. Gestational age, bronchopulmonary
dysplasia (BPD), and weight were significant confounders in the association
between nutrient intake and IGF-1 levels.
Conclusion
Parenteral nutrition was found to be a negative predictor of IGF-1 levels, and there
could potentially be a time frame in which macronutrient intake is unable to
impact IGF-1 levels. Future research should aim to narrow down this time frame
and to gain more insight into factors enhancing or decreasing the response of IGF-
1 to nutrition, e.g., age and inflammatory state, to align nutritional interventions
accordingly.
70
IGF-I and nutrition in preterm infants
Introduction
Preterm birth leads to an abrupt disruption of fetal development, leaving preterm
infants in a precarious situation where they need to thrive despite an immature
gastrointestinal tract and not fully developed immune and endocrine functions.
Insulin-like growth factor 1 (IGF-1) stimulates growth and plays a crucial role in the
complex association between early nutrient intake, growth, and maturation (1). In
preterm infants, IGF-1 is mainly stimulated by insulin and nutrition (1, 2). However,
to what extent various macronutrients impact IGF-1 levels in different phases of
postnatal life is yet to be elucidated.
In the few studies relating actual nutrient intake to IGF-1 levels between preterm birth
and hospital discharge, protein and energy intake had a positive association with
IGF-1 levels (3-5). Remarkably, one previous study reported that in preterm infants,
4
the positive association between IGF-1 and nutrient intake was only apparent after
30 weeks postmenstrual age (PMA) (4). This suggests that there might be a limited
window of opportunity for nutrition to influence early postnatal growth.
It is to be noted that, to the best of our knowledge, only protein and energy intake
have been studied in relation to IGF-1 levels in preterm infants. Interestingly, in
adults, studies assessing fat and carbohydrate intake in relation to IGF-1 have been
inconclusive. This leaves us with a gap in knowledge concerning the potential impact
of dietary fat and carbohydrate intake on IGF-1 levels in preterm infants (6-8).
In addition, the route of nutrient administration is another largely uncharted research

area in relation to IGF-1 levels in preterm infants. Animal studies have shown that in a
state of inflammation or poor nutrient intake, enteral feeding results in higher IGF-1
levels than parenteral feeding. This is thought to be due to a reduction in inflammatory
cytokine levels after enteral feeding (9, 10). These findings suggest that the route
of nutrient administration could mediate cytokine production and consequently
influence IGF-1 levels. To our knowledge, this is yet to be investigated in preterm infants.
Given the impact of poor growth and subsequent accelerated growth on later health
outcomes in infants born preterm (11-13), it is pertinent to gain insight into factors
influencing early postnatal growth, in order to obtain potential interventions to avert
later adverse outcomes. In this light, the association between nutrition and IGF-1 is of
particular interest, because nutrition is a factor which lends itself well for intervention
and could lead to changes in clinical practice. Nevertheless, research on the
relationship between nutrition and IGF-1 in preterm infants is scarce, and most studies
were published over a decade ago. Meanwhile, nutrition and neonatal intensive care
have significantly changed. In addition, in previous studies, the infants were either on
full enteral feeds or the relationship with the proportion of parenteral feeding was not
taken into account. In this explorative observational study, associations between the
macronutrient intake, the proportion of parenteral feeding, and IGF-1 were assessed in
very and extremely preterm infants between birth and 36 weeks PMA.
71
Chapter 4
Methods
Study Population
This paper describes the results of the “Nutrition in relation to the endocrine
regulation of preterm growth” (NUTRIE) study, a longitudinal observational study
on nutrition in relation to the endocrine regulation of growth and body composition
in preterm infants. Eighty-seven participants were enrolled between September
2015 and July 2018. Infants born between 24 and 32 weeks of gestation were eligible
for study participation if they were born without substantial congenital anomalies,
and were admitted to the neonatal intensive care unit (NICU) of Amsterdam UMC,
location VU University Medical Center in Amsterdam, The Netherlands. Informed
consent was obtained in the first week of life. The study was approved by the
medical ethics committee of the VU Amsterdam and was registered at the Dutch
Trial Register (www.trialregister.nl; NTR5311).
Nutrition
Infants initially received total parenteral nutrition and minimal enteral feeding.
During total parental feeding, clinicians aimed to achieve an energy intake of 85–
100 kcal kg−1 day−1, a protein intake of 3–4 g kg−1 day−1, and a fat intake of 3–3.5 g kg−1
day−1. One the first day of life, parenteral glucose administration was targeted at
5.5–7 mg.kg−1 min−1, going up to maximum 12 mg kg−1 min−1 after the first week of
life, depending on blood glucose levels. Full enteral feeding (160 mL kg−1 day−1) was
aimed to be achieved within 7 to 10 days after birth with a total protein intake of 3.5
to 4.5 g kg−1 day−1 and a total energy intake of 110 to 140 kcal kg−1 day−1. Infants were
primarily fed human milk. If own mother’s milk was insufficient or unavailable,
donor human milk was administered up to 32 weeks PMA, followed by preterm
starters formula until discharge home. If parents declined the use of donor human
milk, infants were fed preterm starters formula from birth, whenever own mother’s
milk was unavailable.
Clinicians aimed to achieve 15–20 g weight gain kg−1 day−1, with a weight SD score
above −1 SD. Head circumference growth was targeted at 1 cm per week and length
at 1.25 cm per week.
Breast milk fortifier (Nutrilon Nenatal Breast Milk Fortifier, Nutricia, Wageningen,
The Netherlands) was added to human milk once an enteral intake of 100 mL kg−1
day−1 was achieved. In case of poor growth, as assessed by the clinician in charge,
intake was increased to a maximum of 180 mL/kg, permitted that the infant’s
condition allowed for an increased fluid intake. If poor growth persisted, up to
1% protein fortifier (Nutrilon Nenatal Protein Fortifier, Nutricia, Wageningen, The
Netherlands) was added to the fortified human milk. Lastly, up to 4% of a high-
energy, long-chain triglyceride, fat emulsion (Calogen, Nutricia, Wageningen, The
Netherlands) was added if growth remained restricted despite fortification. In
72
case of growth restriction in formula-fed infants, intake was increased to 180 mL

kg−1 day−1 and an additional 1.5 g of preterm starters formula was added per 100
mL formula (Nutrilon Nenatal Start, Nutricia, Wageningen, The Netherlands). In
addition, protein fortifier and a fat emulsion could be added to the formula if poor
growth persisted.
Study Procedures
All participants were admitted to the NICU of Amsterdam UMC, location VU
University medical center within 24 h from birth. Infants in good clinical condition
were discharged to step-down units in other hospitals at a PMA of 30 weeks and a
4
weight of at least 1000 g.
Obstetric data, clinical condition and intake up to 36 weeks PMA were collected
from hospital records.
Growth
Growth was assessed weekly between birth and 36 weeks PMA. Weight was
measured on an electronic scale to the nearest gram, length was measured on
a length board to the nearest 0.1 cm, and occipital-frontal head circumference
was measured with a nonstretchable measuring tape to the nearest 0.1 cm. The
measurements were done by the nursing staff.
Standard deviation scores (SDS) of weight, length and head circumferences were
calculated according to Fenton (14).
Intake
Daily macronutrient intake was calculated from actual intake data obtained from
hospital records. Own mother’s milk composition was based on reference values
(15, 16) (Table 1). Donor human milk composition was based on analyses of the
donor milk batches administered to the first 23 study participants.
Table 1. Reference values used for the nutritional composition of human milk per 100 mL.
Variables OMM OMM + BMF DHM DHM + BMF

(4.4g/100mL)
Energy (kcal) 68.5 83.8 60 75
Protein (g) 1.5 2.6 0.8 1.9
Protein/energy ratio (g/100 kcal) 2.2/100 1.3/100
Carbohydrates (g) 7.3 10.0 7.5 10.2
Fat (g) 3.3 3.3 2.9 2.9
BMF: Breast milk fortifier, DHM: donor human milk, OMM: own mother’s milk.
73
Chapter 4
Endocrine Parameters
IGF-1 was sampled from umbilical cord blood at birth and from venipuncture
or capillary puncture every other week between birth and 36 weeks PMA. A
chemiluminescence immunoassay (LIAISON®, DiaSorin, Saluggia, Italy) was used
to analyze IGF-1 (intra-assay percent coefficient of variation (% CV): 8%, inter-assay
% CV: 7%). The number of samples per week PMA is depicted below (Table 2).
Table 2. Sample size for IGF-1 analyses per postmenstrual age.
PMA 24 25 26 27 28 29 30 31 32 33 34 35 36
N Total 2 3 7 17 21 33 30 34 25 33 25 26 29
N Postnatal 0 0 1 3 6 12 16 22 25 33 25 26 29
N Total reflects the total number of samples taken. N Postnatal reflects the number of samples excluding
umbilical cord blood.
Potential Confounders
The following comorbidities were assessed as potential confounders in the
association between nutrient intake and IGF-1:
• Bronchopulmonary dysplasia (BPD); defined as having had a need for
supplemental oxygen for at least 28 days at 36 weeks PMA or discharge home
(whichever came first) (17).
• Necrotizing enterocolitis (NEC); classified according to the Modified Bell’s
staging criteria (18).
• Late-onset sepsis (LOS), defined as sepsis occurring 72 h after birth with a
positive blood culture or a full course of antibiotic treatment (19).
• Retinopathy of prematurity (ROP), classified according to the International
Classification for Retinopathy of Prematurity (20).
• Intraventricular hemorrhage (IVH), classified according to the Papile grading
system (21).
• Patent ductus arteriosus (PDA), which was defined as hemodynamically
significant if treatment was prescribed (22).
In addition, gender, gestational age at birth, postmenstrual age at the time of blood
sampling, weight and weight SD score were assessed as potential confounders.
Statistical Analysis
The change in IGF-1 over time was predicted for each individual using mixed
models. The associations between nutrient intake, IGF-1, and potential confounders
were assessed with regression analyses. Analyses were conducted using IBM®
SPSS® Statistics 26 for Windows (IBM Corp., Armonk, NY, USA). Two-sided statistical
significance was assumed at p-values less than 0.05.
74
Results
Eighty-seven infants were included in primary analysis (Figure 1). Baseline
characteristics are shown in Table 3.
Figure 1. Flow diagram of participants included in the study.
75
Chapter 4
Table 3. Characteristics of the study population.
Variables (n = 87)
Gender, n (%)
Male 44 (50.6)
Female 43 (49.4)
Ethnicity, n (%)
White 66 (75.9)
Other 21 (24.1)
Gestational age (weeks), mean (SD) 29.0 (1.8)
Extremely preterm, n (%) 25 (28.7)
Very preterm, n (%) 62 (71.3)
Birthweight (g), mean (SD) 1210 (216)
Birthweight SDS, mean (SD) 0.0 (0.7)
Birthweight SDS < −1.3, n (%) 3 (3.4)
BPD, n (%) 30 (34.5)
NEC, n (%) 8 (9.2)
LOS, n (%) 30 (34.5)
PDA, n (%)
Hemodynamically Insignificant PDA 11 (12.6)
Hemodynamically Significant PDA 8 (9.2)
ROP, n (%)
ROP stage I 4 (4.6)
ROP stage III 1 (1.1)
IVH, n (%)
IVH grade I 8 (9.2)
IVH grade II 11 (12.6)
IVH grade III 4 (4.6)
BPD: Bronchopulmonary dysplasia, IVH: intraventricular hemorrhage, LOS: Late-onset sepsis; NEC:
Necrotizing enterocolitis; PDA: patent ductus arteriosus, ROP: retinopathy of prematurity.
Changes in IGF-1 During Hospitalisation

Between birth and the second week of life, IGF-1 levels dropped from 4.8 nmol/L
to 3.2 nmol/L in extremely preterm infants (mean decrease −1.5, 95% CI −5.2–2.2, p
= 0.314). In very preterm infants, IGF-1 showed a mean decrease of 0.3 nmol/L (95%
CI −2.1–1.4, p = 0.675) between birth and the second week of life. From the second
week of life, IGF-1 showed a mean (SD) increase of 0.6 (0.2) nmol/L per week in very
preterm infants and 0.7 (0.1) nmol/L per week in extremely preterm infants (mean
difference 0.1, 95% CI 0.0–0.2, p = 0.143) (Figure 2).
76
Figure 2. Insulin-like growth factor 1 (IGF-1) levels in extremely preterm and very preterm infants.
Compared to boys, at birth, IGF-1 levels were lower in girls. In addition, IGF-1 levels
had an inverse relationship with gestational age at birth and PMA at the time
of sampling. Postnatal age in days at the time of sampling did not predict IGF-1
levels. After correcting for weight, IGF-1 levels were no longer predicted by gender,
gestational age at birth, and postmenstrual age at the time of blood sampling
either, and weight remained the only significant predictor of IGF-1 levels.
IGF-1 Levels in Relation to Growth

Mean birth weight SDS was 0.04, with three out of 87 infants being small for
gestational age (weight SDS < −1.3) (Figure 3). At 36 weeks PMA 17 out of 80 infants
had a weight SDS below −1.3 SDS. Between the second week of life and 36 weeks
PMA, five out of 80 infants showed catch-up growth (increase in weight SDS > 0.67).
IGF-1 positively correlated with previous, concurrent, and subsequent weight and
weight SDS. When weight SDS was corrected for absolute weight in grams, only
weight remained a significant predictor of IGF-1 levels. Compared to infants with
a weight of 1000 g or more, infants with a weight below 1000 g had a 2.5 nmol/L
lower IGF-1 at two weeks postnatal age (95% CI −3.7–−1.3, p < 0.001).
77
Chapter 4
Figure 3. Weight SD score in extremely and very preterm infants.
IGF-1 Levels and Route of Administration

On the fourteenth day of life, 80% of infants received full enteral feeding. Figure 4
displays the ratio between parenteral and enteral intake in the first two weeks of
life. At two weeks postnatal age, 73 out of 87 infants were fed more than 90% own
mother’s milk, 10 were fed donor human milk, and 4 were formula-fed. From the
second week of life, mean nutrient intake was within the references of our local
protocol (Figure 5). The percentage of parenteral intake from the eighth through to
parental energy intake
the twelfth day of life, expressed as total energy intake , was associated with lower
IGF-1 levels at two weeks postnatal age. Gestational age, weight, BPD, and
hemodynamically significant PDA were confounders in the relationship between
parenteral intake and IGF-1 levels. Based on the F-change, the best predictive
model included weight, BPD, and hemodynamically significant PDA. After
correcting for these confounders, the association between the percentage of
parenteral nutrition and IGF-1 levels remained significant (Table 4).
78
Figure 4. Parenteral and enteral nutrient intake in preterm infants in the first two weeks of life.
Day 1 is not equal to 24 h for all study subjects.
79
Chapter 4
Figure 5. Nutrient intake in preterm infants in the first six weeks of life.
Table 4. Regression analyses of parenteral nutrition as a predictor of IGF-1 levels at 2 weeks postnatal age.
Variables B (SE) β p-Value

Included variables
Constant 1.482 (1.359) 0.281
Percentage parenteral intake on day 8 −0.027 (0.011) −0.234 0.019
Weight on day 8 (grams) 0.004 (0.001) 0.478 <0.001

BPD −1.134 (0.516) −0.233 0.032
Hemodynamic significant PDA −1.350 (0.793) −0.159 0.095
parental energy intake

R² = 0.574, p < 0.001; percentage parenteral intake: total energy intake , BPD: bronchopulmonary
dysplasia, PDA: persistent ductus arteriosus, IGF-1: insulin-like growth factor 1.
80
Nutrient Intake in Relation to Concurrent IGF-1 Levels

Positive associations were found between energy intake and IGF-1 levels at 30 to
33 weeks PMA (Table 5). BPD was a significant confounder from 32 weeks PMA.
Protein, carbohydrate, and fat intake showed a similar pattern (Table 5). In addition,
however, protein intake showed a positive association with IGF-1 levels at 28
weeks PMA: per gram increase in protein intake IGF-1 levels showed an increase
of 1.1 nmol/L, R² = 0.506, p = 0.032. This is in contrast to a lack of associations at 29
weeks PMA with a larger sample size (n = 12) than the sample size at 28 weeks
PMA. At 28 weeks PMA, IGF-1 was measured in six infants, of whom five had a
recent history of sepsis and required an erythrocyte transfusion within 24 h of the
blood sampling. Nutrient intake per kg body weight was not associated with IGF-1
levels at any point in time. After correcting for weight in multivariate analysis, the
associations between total nutrient intake and IGF-1 lost their significance (Table
4
5). In univariate analysis at 30 weeks PMA, weight explained 45% of the variance in
IGF-1 levels, compared to 33% of the variance that was explained by nutrient intake.
By 33 weeks PMA, these numbers declined to respectively 17% and 15%.
Nutrition in Relation to Changes in IGF-1 According to Postnatal Age

The change in IGF-1 levels in the first four weeks of life was positively associated
with protein, carbohydrate, fat, and total energy intake (after correction for
gestational age). IGF-1 levels increased with 0.01 nmol/L per 10 kcal, R² = 0.266, p <
0.001. Comorbidities were not a significant confounder. After correcting for weight,
total nutrient intake was no longer a significant predictor of change in IGF-1.
Nutrition in Relation to Changes in IGF-1 According to Postmenstrual Age

Looking at postmenstrual age, there was a positive association between total
nutrient intake from 28 through 31 weeks PMA and the change in IGF-1 between
birth and 32 weeks PMA (after correcting for gestational age). IGF-1 levels increased
with 0.2 nmol/L per 10 kcal, R² = 0.287, p = 0.002. Comorbidities were not a significant
confounder. However, after correcting for weight, energy intake could no longer
predict the change in IGF-1. All macronutrients showed a similar pattern.
81
Chapter 4
Table 5. Regression analyses of intake as a predictor of IGF-1 levels at 30 weeks postmenstrual age.
Variables Model R² Model p-Value B (SE) β p-Value

Energy intake model 1: 0.605 0.006
Constant 11.8 (6.7) 0.106
Energy intake (kcal/day) 0.05 (0.02) 0.6 0.015
Gestational age (weeks) −0.6 (0.2) −0.5 0.029
Energy intake model 2: 0.640 0.014
Constant 9.4 (7.1) 0.215
Energy intake (kcal/day) 0.03 (0.03) 0.3 0.395
Weight (grams) 0.003 (0.003) 0.36 0.348
Protein intake model 1: 0.578 0.009
Constant 14.5 (6.7) 0.053
Protein intake (g/day) 1.2 (0.4) 0.6 0.013
Protein intake model 2: 0.625 0.017
Constant 10.2 (7.6) 0.209
Protein intake (g/day) 0.5 (0.8) 0.2 0.561
Weight (grams) 0.004 (0.003) 0.4 0.289
Carbohydrate intake model 1: 0.593 0.014
Constant 9.1 (7.8) 0.268
Carbohydrate intake (g/day) 0.3 (0.1) 0.6 0.022
Carbohydrate intake model 2: 0.690 0.007
Constant 5.5 (6.9) 0.444
Carbohydrate intake (g/day) 0.2 (0.1) 0.3 0.144
Weight (grams) 0.004 (0.002) 0.5 0.052
Fat intake model 1: 0.581 0.008
Constant 12.3 (6.9) 0.102
Fat intake (g/day) 1.1 (0.4) 0.6 0.012
Fat intake model 2: 0.631 0.015
Constant 9.4 (7.2) 0.225
Fat intake (g/day) 0.5 (0.7) 0.3 0.494
Weight (grams) 0.003 (0.003) 0.4 0.273
IGF-1: Insulin-like growth factor 1.
82
Discussion
This study shows that the proportion of parenteral nutrition is negatively associated
with IGF-1 levels in extremely and very preterm infants. Gestational age, BPD, and
weight were significant confounders in the association between nutrient intake
and IGF-1 levels. Total protein, fat, and carbohydrate intake, as well as total energy
intake, showed a positive association with IGF-1 levels, particularly between 30 and
33 weeks PMA.
The Effect of the Various Macronutrients on IGF-1 Levels

Studies in preterm infants consistently show that protein intake is positively
associated with IGF-1 levels (3-5). However, not all studies could show that energy
intake was a predictor of IGF-1 after correction for confounders (3). In our study,
4
higher total protein intake and higher total energy intake were associated with
higher IGF-1 levels. In contrast to previous studies, our study did not find an
association between nutrient intake per kg bodyweight and IGF-1 levels. Weight
explained more of the variance in IGF-1 than nutrient intake. However, it should be
noted that the variance in nutrient intake per kg bodyweight may have been too
small to show significant differences in IGF-1 levels in our population. This is due to
the univocal application of our local nutrition protocol. For example, from 33 weeks
PMA the interquartile range in protein intake was between 3.7 and 4.1 g kg−1 day−1.
This range was notably smaller compared to previous research (4) and could have
limited the statistical power.
To our knowledge, the impact of fat and carbohydrate intake on IGF-1 levels in
preterm infants has not been studied previously. Meanwhile, studies in adults have
been inconclusive, with some finding positive associations (7), while others found
negative associations (6) or no association at all (8). In our study, total fat and total
carbohydrate intake showed a positive association with IGF-1 levels. Interestingly,
both carbohydrate and fat intake had a comparable impact on IGF-1 when
compared to protein intake—a one SD change in any of the macronutrients led to
a change of 0.6 SD in IGF-1 levels at 30 weeks PMA. Although it has been suggested
that the role of proteins is more important than that of carbohydrates and fat in
stimulating IGF-1 (3, 6), like dietary proteins, dietary fat and carbohydrates have
been found to increase hepatic IGF-1 expression in animal studies (23). Moreover,
fat and carbohydrates provide the majority of the total energy intake, which has
repeatedly been shown to have a positive association with IGF-1 levels and thus
supports our findings.
The Route of Nutrient Administration

Parenteral feeding was found to be negatively associated with IGF-1 levels. It
could be hypothesized that less exposure of the gastrointestinal tract to nutrition
enhances a pro-inflammatory state in the immature gut, which in turn could lead
83
Chapter 4
to lower IGF-1 levels. Indeed, a pro-inflammatory state in preterm infants has been
associated with decreased IGF-1 levels (24). It has also been demonstrated that
colostrum and maternal milk contain high concentrations of anti-inflammatory
cytokines (25). These anti-inflammatory cytokines could potentially lower the
relatively pro-inflammatory state in the immature gut and consequentially increase
IGF-1 levels. In one study, preterm infants who received own mother’s milk from
birth were shown to have higher levels of IGF-1 at term equivalent age compared to
those who were formula fed from birth (26). Moreover, animal studies found that in
a state of inflammation or nutrient deprivation, parenteral feeding was associated
with lower IGF-1 levels compared to enteral feeding. This appears to be due to a
decrease in pro-inflammatory cytokines after enteral feeding (9, 10). This leads us
to believe that the neutralizing effect of anti-inflammatory cytokines, which are
particularly abundant in colostrum and breast milk, is diminished and results in
lower IGF-1 levels when parenteral nutrition is increased.
It could also be suggested that infants who received relatively higher proportions
of parenteral nutrition were the more vulnerable, smaller, younger, and iller
infants, and thus the association with lower IGF-1 levels. However, after correcting
for gestational age, weight, and comorbidities, parenteral nutrition remained a
significant predictor of IGF-1 levels.
Window of Effect of Nutrient Intake on IGF-1 Levels

In our population, the influence of nutrition on IGF-1 levels seemed to be most
apparent between 30 and 33 weeks PMA. Hypothesizing, preterm infants may
have to reach a certain level of maturity before an impact of nutrition on the
IGF-1 axis can be noted. In support of this, Smith and colleagues found that the
magnitude of the rise in IGF-1 levels per gram protein increased with increasing
gestational and postnatal age. Hansen-Pupp and colleagues also found nutrient
intake not to influence IGF-1 levels at lower postmenstrual ages, but only from 32
weeks PMA onwards. Speculatively, other factors than maturity could influence
when IGF-1 levels start to respond to nutrient intake. Of note, in our study, in a
set of infants who were ill, a positive association between total protein intake and
IGF-1 levels was already found at 28 weeks PMA. This was in contrast to the other
macronutrients and total energy intake, which only showed positive associations
with IGF-1 levels from 30 weeks PMA onwards. Among the set of infants of whom
blood was sampled at 28 weeks, all but one had a recent history of sepsis and
anemia requiring erythrocyte transfusion. It could be speculated that the state
of inflammation triggered a higher sensibility of the IGF-1 axis to protein uptake.
Despite their IGF-1 levels still being low, 1 g of protein might have triggered more
increase in IGF-1 compared to infants who were not ill.
It is noteworthy that in our population, no associations between nutrient intake

and IGF-1 levels were found at 34 and 35 weeks PMA, in contrast to other studies
84
(3-5). However, Ëngstrom and colleagues found that in infants with a weight of less
than 2000 g, protein supplementation had a stronger association with IGF-1 levels
compared to infants over 2000 g. Perhaps this can explain why the positive trend
our study found at 34 and 35 weeks PMA was not statistically significant.
In contrast to the relationship between nutrient intake and concurrent IGF-1 levels
described above, our results showed nutrient intake to influence the change in IGF-1 at
a younger PMA. For every macronutrient, intake from 28 weeks PMA was associated
with the change in IGF-1 levels between birth and 32 weeks PMA. Hypothesizing,
total macronutrient intake before 30 weeks may not reach the threshold to increase
concurrent IGF-1 levels, but it might stimulate the IGF-1 axis to mature more rapidly
and in this way cause a more rapid increase in IGF-1 levels over time.
4
Strengths and Limitations
This is the first study to evaluate the contribution of all macronutrients in relation
to circulatory IGF-1 levels in preterm infants from birth until 36 weeks of gestation.
In addition, to the best of our knowledge, the proportion of parenteral nutrition
has not been investigated previously in relation to IGF-1 levels. In line with previous
research, our results demonstrate a slow increase in IGF-1 levels in the first weeks
of life (4). However, our results failed to support previous findings on nutrient
intake per kg body weight. As previously mentioned, our population had little
variation in nutrient intake per kg body weight. This could potentially explain the
lack of statistically significant findings. In addition, despite the considerable overall
sample size, this study had a relatively small sample size per week. This was due to
the low sample frequency (on alternating weeks) and the relatively small number
of extremely preterm infants, which resulted in a sample size ranging from 1 infant
at 26 weeks PMA to 33 infants at 33 weeks PMA. This may have contributed to our
findings. Moreover, it is important to note that this was an exploratory observational
study. Therefore, the findings should be interpreted with caution and strong
conclusions on potential causative relationships cannot be made.
Conclusions
Our findings further illustrate the complex association of maturation, concurrent
comorbidities, and nutrition in relation to IGF-1 levels. The proportion of parenteral
nutrition was found to be a negative predictor of IGF-1 levels, affirming the
importance of stimulating enteral nutrition and limiting parenteral nutrition as
much as possible in clinical practice. Our findings point towards a potential time
frame in which nutrition is unable to impact IGF-1 levels. Future research should aim
to narrow down this time frame and to gain more insight into factors enhancing or
decreasing the response of IGF-1 to nutrition, e.g., age and inflammatory state, to
align nutritional interventions accordingly.
85
Chapter 4
References
care. 2012;15(3):285-92.
2015;77(1-2):156-63.
3. Engstrom E, Niklasson A, Wikland KA, Ewald U, Hellstrom A. The role of maternal factors,
postnatal nutrition, weight gain, and gender in regulation of serum IGF-I among preterm
infants. Pediatr Res. 2005;57(4):605-10.
5. Smith WJ, Underwood LE, Keyes L, Clemmons DR. Use of insulin-like growth factor I (IGF-I) and
IGF-binding protein measurements to monitor feeding of premature infants. The Journal of
clinical endocrinology and metabolism. 1997;82(12):3982-8.
6. Giovannucci E, Pollak M, Liu Y, Platz EA, Majeed N, Rimm EB, et al. Nutritional predictors
of insulin-like growth factor I and their relationships to cancer in men. Cancer Epidemiol
Biomarkers Prev. 2003;12(2):84-9.
7. Kaklamani VG, Linos A, Kaklamani E, Markaki I, Koumantaki Y, Mantzoros CS. Dietary fat and
carbohydrates are independently associated with circulating insulin-like growth factor 1 and
insulin-like growth factor-binding protein 3 concentrations in healthy adults. J Clin Oncol.
1999;17(10):3291-8.
8. Norat T, Dossus L, Rinaldi S, Overvad K, Gronbaek H, Tjonneland A, et al. Diet, serum insulin-like
growth factor-I and IGF-binding protein-3 in European women. Eur J Clin Nutr. 2007;61(1):91-8.
9. O’Leary MJ, Xue A, Scarlett CJ, Sevette A, Kee AJ, Smith RC. Parenteral versus enteral nutrition:
effect on serum cytokines and the hepatic expression of mRNA of suppressor of cytokine
signaling proteins, insulin-like growth factor-1 and the growth hormone receptor in rodent
sepsis. Critical care (London, England). 2007;11(4):R79.
2019;11(9).
12. Embleton ND, Korada M, Wood CL, Pearce MS, Swamy R, Cheetham TD. Catch-up growth and
metabolic outcomes in adolescents born preterm. Arch Dis Child. 2016;101(11):1026-31.
13. Lapillonne A, Griffin IJ. Feeding preterm infants today for later metabolic and cardiovascular
outcomes. J Pediatr. 2013;162(3 Suppl):S7-16.
14. Fenton TR, Kim JH. A systematic review and meta-analysis to revise the Fenton growth chart
for preterm infants. BMC Pediatr. 2013;13:59.
15. Boyce C, Watson M, Lazidis G, Reeve S, Dods K, Simmer K, et al. Preterm human milk
composition: a systematic literature review. Br J Nutr. 2016;116(6):1033-45.
16. Gidrewicz DA, Fenton TR. A systematic review and meta-analysis of the nutrient content of
preterm and term breast milk. BMC Pediatr. 2014;14:216.
17. Jobe AH, Bancalari E. Bronchopulmonary dysplasia. Am J Respir Crit Care Med. 2001;163(7):1723-9.
18. Kliegman RM, Walsh MC. Neonatal necrotizing enterocolitis: pathogenesis, classification, and
spectrum of illness. Curr Probl Pediatr. 1987;17(4):213-88.
19. Bekhof J, Reitsma JB, Kok JH, Van Straaten IH. Clinical signs to identify late-onset sepsis in
preterm infants. Eur J Pediatr. 2013;172(4):501-8.
86
20. International Committee for the Classification of Retinopathy of P. The International

Classification of Retinopathy of Prematurity revisited. Arch Ophthalmol. 2005;123(7):991-9.
21. Papile LA, Burstein J, Burstein R, Koffler H. Incidence and evolution of subependymal and
intraventricular hemorrhage: a study of infants with birth weights less than 1,500 gm. J Pediatr.
1978;92(4):529-34.
22. Jain A, Shah PS. Diagnosis, Evaluation, and Management of Patent Ductus Arteriosus in
Preterm Neonates. JAMA Pediatr. 2015;169(9):863-72.
23. Bertucci JI, Blanco AM, Canosa LF, Unniappan S. Direct actions of macronutrient components
on goldfish hepatopancreas in vitro to modulate the expression of ghr-I, ghr-II, igf-I and igf-II
mRNAs. Gen Comp Endocrinol. 2017;250:1-8.
24. Hansen-Pupp I, Hellstrom-Westas L, Cilio CM, Andersson S, Fellman V, Ley D. Inflammation
at birth and the insulin-like growth factor system in very preterm infants. Acta Paediatr.
2007;96(6):830-6.
25. MohanKumar K, Namachivayam K, Ho TT, Torres BA, Ohls RK, Maheshwari A. Cytokines
and growth factors in the developing intestine and during necrotizing enterocolitis. Semin
4
Perinatol. 2017;41(1):52-60.
26. Alzaree FA, AbuShady MM, Atti MA, Fathy GA, Galal EM, Ali A, et al. Effect of Early Breast Milk
Nutrition on Serum Insulin-Like Growth Factor-1 in Preterm Infants. Open access Macedonian
journal of medical sciences. 2019;7(1):77-81.
87
PART II.
THE DETERMINANTS OF BODY
COMPOSITION AND METHODS TO
ASSESS BODY COMPOSITION IN
PRETERM INFANTS
5
CHAPTER 5
IGF-I, growth and body composition in
preterm infants up to term equivalent age
Dana FJ Yumani, Harrie N Lafeber and Mirjam M van Weissenbruch.

J Endocr Soc. 2021 Jun 18;5(7):bvab089. doi: 10.1210/jendso/bvab089
Chapter 5
Abstract
Context
There are concerns that a higher fat mass in the early life of preterm infants is
associated with adverse cardiometabolic outcomes in young adulthood.
Objective
To investigate the role of IGF-I and growth in determining body composition of
preterm infants at term equivalent age.
Design
An observational study was conducted from August 2015 to August 2018. From
birth to term equivalent age IGF-I levels were measured bi-weekly and growth was
assessed weekly. At term equivalent age body composition was assessed through
air displacement plethysmography.
Patients
Sixty-five infants with a gestational age of 24 to 32 weeks were assessed at term
equivalent age of whom 58 completed body composition measurement.
Main outcome measures

Fat (free) mass (g) and fat (free) mass percentage at term equivalent age.
Results
In the first month of life each 0.1 nmol/L per week increase in IGF-I was associated
465 g (S.E. 125 g) increase in fat free mass. A greater increase in weight SDS in the
first month of life was associated with a higher fat free mass percentage (B 200.9,
95% CI 12.1 - 389.6). A higher head circumference SDS was associated with more fat
free mass (r 0.46, 95% CI 0.21 – 0.65). However, a greater increase in weight SDS up
to term equivalent age was associated with a lower fat free mass percentage (B
-55.7, S.E. 9.4).
Conclusions
These findings suggest that impaired growth in the first month of life , is associated
with a less favorable body composition at term equivalent age.
92
IGF-I, growth and body composition in preterm infants
Introduction
Preterm birth abruptly interferes with a prime stage of development. Not
surprisingly, almost half of infants born preterm, show postnatal growth restriction
by the time they are discharged from the hospital. (1, 2) In addition, preterm infants,
compared to infants born at term, have been reported to have a higher fat mass
percentage at term equivalent age. (3) Postnatal growth restriction, as well as
increased fat mass in infancy, have been linked to impaired neurodevelopmental
outcome. (4, 5) In addition, there are concerns that higher fat mass in early life
is associated with adverse cardiometabolic outcomes in young adulthood. (6,
7) Therefore, it would be of interest to gain more insight in factors determining
growth and body composition in early life, as a means to ameliorate later health
outcomes in infants born preterm.
Insulin-like growth factor I (IGF-I) plays a key role in the regulation of growth and
body composition in early life. (8) Interestingly, the relationship between IGF-I and
body composition has been reported to vary, depending on the timing of IGF-I and
5
body composition measurement. Based on previous studies, it could be speculated
that high IGF-I levels between preterm birth and term equivalent age increase fat
free mass, while from term age onwards higher IGF-I could be associated with
more fat mass. (9-11) Nevertheless, only a few studies report on IGF-I in relation
to body composition in preterm infants. Therefore, definitive conclusions on the
relationship between IGF-I and body composition are yet to be drawn.
This study aimed to explore the relationship between IGF-I, growth and body
composition in preterm infants up to term equivalent age. To account for the
possible impact of timing, IGF-I levels were assessed in different time frames
between birth and term equivalent age and associations between changes in
IGF-I, growth and body composition were explored.
Methods
Study population
Ninety infants were enrolled between August 2015 and August 2018. (Figure 1)
The study participants were part of the NUTRIE study (Nutrition in relation to the
endocrine regulation of preterm growth study). This longitudinal observational
study collected data on nutritional intake, linear growth, body composition and
IGF-I levels between preterm birth and 2 years corrected age. Infants admitted to
the neonatal intensive care unit (NICU) of the VU University Medical Center were
assessed for eligibility if they were born at a gestational age of 24 weeks + 0/7 days
up to and including 31 weeks + 6/7 days. Infants who had substantial congenital
abnormalities were excluded. Informed consent was obtained within the first
week of life.
93
Chapter 5
The NUTRIE study was powered to detect a medium size effect (r = 0.35) of IGF-I on
fat mass percentage. To that end, a sample size of at least 62 infants was required
(power 80%, significance 5%). With an expected dropout rate of 10%, the aim was
to include 70 infants.
The study was approved by the medical research ethics committee of the VU
University Medical Center and was registered in the Dutch trial register (www.
trialregister.nl; NTR5311).
Study procedures
All participants were admitted to the NICU of the VU University medical center
within 24 hours of birth. Infants were discharged to step-down clinics in other
hospitals when the clinical condition permitted to do so and the infants had
reached a postmenstrual age (PMA) of at least 30 weeks and a weight of at least
1000 grams. Clinical data was collected up to 36 weeks PMA or discharge home,
whichever came first. Infants were invited for follow-up at term equivalent age.
Growth
From birth to hospital discharge, weight, length and head circumference were
measured weekly by the nursing staff. Weight was measured to the nearest gram
on an electronic scale. Length was measured with a length board to the nearest 0.5
cm. Occipital-frontal head circumference was measured to the nearest 0.1 cm with
a non-stretchable measuring tape. At the follow-up visit growth was measured in
the same way by two investigators. Fenton growth charts were used to calculate
weight SDS. (12) Growth restriction was defined as a weight SDS under the tenth
percentile (-1.3 SD). (1)
Body composition
Fat mass and fat-free mass were measured at term equivalent age through air
displacement plethysmography (ADP), using the PeaPod®. Infants were measured
nude and hair was smoothened using hair oil. Movement was allowed, but in case
of excessive crying the measurement was stopped. All measurements were done
by two investigators. In line with the manufacturer's guideline, quality control
checks were done daily and the PeaPod® scale was calibrated every two weeks. A
more detailed description of the PeaPod® measurements can be found elsewhere.
(13) The coefficient of variation for repeated measurements has been reported to
lie between 0.02% and 0.09%.(13)
IGF-I
IGF-I was sampled from umbilical cord blood at birth, followed by venous or
capillary blood sampling from the infant every other week up to 36 weeks PMA
94
and once at term equivalent age. A chemiluminescence immunoassay (LIAISON®,

DiaSorin, Italy) was used for the analysis (intra-assay percent coefficient of variation
(%CV): 8%, inter-assay %CV: 7%).
Potential critical windows

Based on previous studies investigating the influence of IGF-I on body composition,
the following time frames were assessed to analyze the predictive value of (changes
in) IGF-I and weight, length and head circumference SDS on body composition at
term equivalent age:
• 4 weeks postnatal age (PNA) and the change from birth to 4 weeks PNA (10);
• 36 weeks PMA and change from birth to 36 weeks PMA (9);
• The change from 30 weeks PMA to term equivalent age (14);
• Term equivalent age (11).
Potential confounders
5
Nutrition
Actual daily macronutrient intake was calculated from the intake documented in
hospital records between birth and 36 weeks PMA and total macronutrient intake
was assessed as a potential confounder.
Infants were started off with minimal enteral feeding and total parenteral nutrition
shortly after birth. Full enteral feeding (160 ml.kg-1.day-1) was aimed to be achieved
within 7 to 10 days after birth. In case of poor growth, as assessed by the clinician in
charge, enteral nutrition was supplemented with protein fortifier (Nutrilon Nenatal
Protein Fortifier, Nutricia, Wageningen, The Netherlands) or fat emulsion (Calogen,
Nutricia, Wageningen, The Netherlands).
Infants were primarily fed their own mother’s milk. The average intake of own
mother’s milk was 90% of the total enteral intake, supplemented with donor
human milk or preterm starters formula, if parents declined donor human milk
use. Donor human milk was only administered up to 32 weeks PMA, thereafter
preterm starters formula was given if own mother’s milk did not suffice. During
hospitalization no infants were exclusively formula fed.
Comorbidities
Data was collected on the occurrence of bronchopulmonary dysplasia (BPD),
necrotizing enterocolitis (NEC), late-onset sepsis (LOS), intraventricular hemorrhage
(IVH), persistent ductus arteriosus (PDA) and retinopathy of prematurity (ROP). The
occurrence and severity of these conditions were assessed as potential confounders.
Other confounders
Lastly, gender and ethnicity were assessed as a potential confounders.
95
Chapter 5
Statistical analysis
Continuous variables were presented as mean and standard deviations (SD).
Dichotomous variables were presented as count and percentages.
Mixed models were used to predict the change in IGF-I and the SDS of weight,
length and head circumference for every individual in the previously mentioned
time frames .
Partial correlations, controlling for gestational age at birth and postmenstrual age
at the time of body composition measurements, were reported for associations
between the (change in) SD scores for weight, length and head circumference and
fat mass, and fat free mass. Likewise, partial correlations were reported for IGF-I
and the aforementioned outcomes.
Due to the non-normal distribution of fat (free) mass percentage the association
between the (change in) SD scores for weight, length, and head circumference
and fat (free) mass percentage was explored using linear regression modeling.
Gestational age at birth and PMA at the time of body composition measurement
were entered in the regression model as covariates.
Potential confounders were evaluated with stepwise (backward) regression analysis.
All statistical analyses were done using IBM® SPSS® Statistics 24 for Windows
(IBM Corp., Armonk, NY, USA). Two-sided statistical significance was assumed at
p-values less than 0.05.
Results
Eighty-seven infants were included in the primary analysis for growth and IGF-I.
Sixty-five infants were assessed at term equivalent age of whom 58 completed body
composition measurement. (Figure 1) The sample size for IGF-I measurements per
postmenstrual age is shown in table 1. Baseline characteristics are shown in table 2.
96
Figure 1. Enrollment and follow-up of study participants
Table 1. IGF-I sample size per postmenstrual age
Postmenstrual age (weeks) 24 25 26 27 28 29 30 31 32 33 34 35 36 TEA
Postnatal samples, n 0 0 1 3 6 12 16 22 25 33 25 26 29 40
Number of infants born at 2 3 6 14 15 21 14 12 0 0 0 0 0 0

respective gestational ages
Total number of participants 2 5 10 24 39 60 74 86 81 81 81 81 81 65
Postnatal samples reflects the number of samples excluding umbilical cord blood
TEA: term equivalent age, mean (SD) postmenstrual age at the term equivalent visit was 43.3 (2.6) weeks
97
Chapter 5
All (n=87)
Gender, n (%)
Male 44 (50.6)
Female 43 (49.4)
Ethnicity, n (%)
White 66 (75.9)
Other 21 (24.1)
Gestational age (weeks), mean (SD) 29.0 (1.8)
Birthweight (g), mean (SD) 1210 (216)
Birthweight SDS, mean (SD) 0.0 (0.7)
Birthweight SDS < -1.3, n (%) 3 (3.4)
Birth length (cm), mean (SD) 37.4 (3.2)
Birth length SDS, mean (SD) 0.0 (0.9)
Birth head circumference (cm), mean (SD) 26.7 (2.2)
Birth head circumference SDS, mean (SD) 0.3 (1.0)
BPD, n (%)
Yes 30 (34.4)
No 57 (65.5)
NEC, n (%)
Yes 8 (9.2)
No 79 (90.8)
LOS, n (%)
Yes 30 (34.5)
No 57 (65.5)
PDA, n (%)
No PDA 68 (78.2)
Hemodynamically insignificant PDA 11 (12.6)
Hemodynamically significant PDA 8 (9.2)
ROP, n (%)
No ROP 82 (94.3)
ROP stage I 4 (4.6)
ROP stage III 1 (1.1)
IVH, n (%)
No IVH 64 (73.6)
IVH grade I 8 (9.2)
IVH grade II 11 (12.6)
IVH grade III 4 (4.6)
98
BPD: Bronchopulmonary dysplasia, IRDS: Infant respiratory stress syndrome, IVH: intraventricular
hemorrhage, LOS: Late-onset sepsis; NEC: Necrotizing enterocolitis; PDA: patent ductus arteriosus, PHVD:
post-hemorrhagic ventricular dilatation, PVL: periventricular leukomalacia, ROP: retinopathy of prematurity
Growth and body composition

At birth mean (SD) weight SDS was 0.0 (0.7) and 3 infants were growth restricted
(3.4 %). Between birth and 36 weeks PMA 25 infants (28.7%) showed more than 1 SD
decline in weight SDS. At 36 weeks PMA 17 infants (19.5%) were growth restricted
(SDS < -1.3). Figure 2 depicts the SDS for weight, length and head circumference
over time. While the SDS for weight and head circumference increased after 32
weeks PMA, the SDS for length showed a continued decline up to 36 weeks PMA.
At the term equivalent age visit, infants had a mean (SD) PMA of 43.3 (2.6) weeks.
The mean (SD) SDS for weight, length and head circumference at term equivalent
age were -0.5 (1.0), -0.7 (1.0) and 0.5 (1.0) respectively. The mean (SD) fat free mass
and percentage at the term age visit were 3301 (448) g and 79.5 (4.0) % respectively.
5
The 17 infants who were growth restricted at 36 weeks PMA had a comparable fat
free mass percentage compared to those who were not growth restricted: median
(IQR) 80 (79-82) and 78 (77-82) % respectively, p 0.147.
Potential critical windows for growth

From birth to 4 weeks PNA, more gain in weight SDS was associated with a higher
fat free mass percentage at term equivalent age. (table 3) On the contrary, more
weight gain from birth to 36 weeks PMA and from 30 weeks PMA onwards, was
associated with a lower fat free mass percentage at term equivalent age. (table
3) Of the potential critical windows for weight gain, the strongest predictor of
fat free mass percentage at term equivalent age was the change of weight SDS
from 30 weeks PMA onwards. (Table 6) More gain in weight SDS in this period was
associated with more fat free mass and more fat mass, but a lower fat free mass
percentage at term equivalent age. (Figure 3)
Correlations between length indices and fat (free) mass at term equivalent age
showed a positive trend. Likewise, correlations between head circumference and
fat (free) mass at term equivalent age showed a positive trend. Nevertheless,
only a few potential critical windows had a statistically significant impact on
body composition at term equivalent age. (table 4 and table 5) In particular, the
change in length and head circumference SDS in the first 4 weeks of life did not
significantly impact fat free mass percentage.
Of the potential critical windows for length, the strongest predictor of fat free
mass percentage was length SDS at term equivalent age: a higher length SDS
was associated with a lower fat free mass percentage. For head circumference,
the change in head circumference SDS from 30 weeks PMA to term equivalent
99
Chapter 5
age was the strongest predictor. (table 6) Similarly to weight gain, more gain in
head circumference in this time frame was associated with a lower fat free mass
percentage at term equivalent age. (table 6)
Table 3. Associations between weight gain and body composition at term age
Fat mass Fat free mass Fat mass Fat free mass
term age term age percentage percentage term
r (95% CI)¹ r (95% CI)¹ term age age
B (95% CI)² B (95% CI)²
Weight SDS 0.57 0.75 2.2 -2.2

4 wks PNA (0.37 – 0.72) (0.60 – 0.84) (0.4 – 4.0)a (-4.0 – -0.4)b
Weight SDS 0.64 0.82 1.8 -1.8

36 wks PMA (0.46 – 0.78) ( 0.71 – 0.89) (0.5 – 3.1)c (-3.1 – -0.5)d
Weight SDS 0.62 0.62 1.7 -1.7

Term age (0.42 – 0.76) (0.42 – 0.76) (0.7 – 2.8)e (-2.8 – -0.7)f
Change in weight SDS -0.54 -0.72 -198.9 200.9

birth – 4 wks PNA (-0.70 – -0.32) (-0.82 – -0.57) (-383.3 – -14.6)g (12.1 – 389.6)h
Change in weight SDS 0.57 0.68 32.3 -32.9

birth – 36 weeks PMA (0.37 – 0.72) (0.51 – 0.80) (9.2 – 55.5) i (-56.6 – -9.1)j
Change in weight SDS 0.77 0.72 52.4 -53.9

from 30 weeks PMA onwards (0.65 – 0.86) (0.56 – 0.82) (34.7 – 70.1)k (-72.0 – -35.9)l
PMA: postmenstrual age, PNA: postnatal age

All associations were statistically significant p < 0.05.
¹ Correlations were controlled for gestational age at birth and postmenstrual age at time of body
composition measurement
² Gestational age at birth and postmenstrual age at the time of body composition measurement were
entered in the regression model as covariates
a
R² 0.176, p 0.015; b R² 0.179, p 0.013; c R² 0.199, p 0.009; d R² 0.202, p 0.008; e R2 0.210, p 0.008; f R2 0.210, p 0.008;
g
R² 0.151, p 0.028; h R² 0.154, p 0.026; i R² 0.193, p 0.008; j R² 0.197, p 0.007; k R² 0.438, p < 0.001; l R² 0.445, p <
0.001;
100
Table 4. Associations between length gain and body composition at term age
B (95% CI)² B (95% CI)²
Length SDS 0.34 0.44 0.9 -0.9

4 wks PNA (0.06 – 0.56) (0.18 – 0.64) (-0.5 – 2.3)a (-2.4 – 0.5)b
Length SDS 0.54 0.72 1.3 -1.3

36 wks PMA (0.29 – 0.72) (0.53 – 0.84) (0.2 – 2.4)c (-2.4 – -0.2)d
Length SDS 0.60 0.82 1.4 -1.4

Term age (0.39 – 0.75) (0.70 – 0.89) (0.4 – 2.4)e (-2.5 – -0.4)f
Change in length SDS -0.073 -0.03 -6.2 5.4

birth – 4 wks PNA (-0.33 – 0.20) (-0.30 – 0.24) (-32.6 – 20.2)g (-21.6 – 32.4)h
Change in length SDS 0.13 0.39 0.2 -0.4

birth – 36 weeks PMA (-0.14 – 0.39) (0.13 – 0.60) (-12.9 – 13.4) i (-13.9 – 13.0)j
Change in length SDS 0.24 0.47 13.5 -14.6

from 30 weeks PMA onwards (-0.02 – 0.47) (0.23 – 0.65) (-16.9 – 43.8)k (-45.6 – 16.4)l
5
Associations in bold were statistically significant p < 0.05.
a
g
R² 0.033, p 0.190; h R² 0.088, p 0.171; i R² 0.047, p 0.138; j R² 0.197, p 0.007; k R² 0.042, p 0.148; l R² 0.049, p 0.124;
101
Chapter 5
Table 5. Associations between head circumference gain and body composition at term age
B (95% CI)² B (95% CI)²
Head circumference SDS 0.23 0.46 0.3 -0.2

4 wks PNA (-0.05 – 0.47) (0.21 – 0.65) (-1.0 – 1.5)a (-1.5 – 1.0)b

36 wks PMA (0.25 – 0.70) (0.35 – 0.75) (0.2 – 2.6)c (-2.7 – -0.2)d

Term age (0.31 – 0.70) (0.33 – 0.71) (0.6 – 2.7)e (-2.7 – -0.7)f
Change in head -0.11 0.09 -6.5 5.8

circumference SDS (-0.36 – 0.17) (-0.18 – 0.35) (-19.5 – 6.5)g (-7.5 – 19.2)h
birth – 4 wks PNA
Change in head 0.14 0.20 5.5 -6.3

circumference SDS (-0.14 – 0.39) (-0.08 – 0.44) (-7.4 – 18.4) i (-19.4 – 6.9)j
birth – 36 weeks PMA
Change in head 0.15 0.00 17.5 -19.1

circumference SDS (-0.12 – 0.39) (-0.27 – 0.27) (-0.6 – 35.6)k (-37.5 – -0.7)l
from 30 weeks PMA onwards

a
g
R² 0.047, p 0.136; h R² 0.048, p 0.131; i R² 0.069, p 0.079; j R² 0.197, p 0.007; k R² 0.138, p 0.042; l R² 0.104, p 0029;
102
Table 6. Predictive value of growth for fat free mass percentage at term equivalent age
B (SE) ß p-value
Model 1 weight gain
Constant 98.9 (9.8) < 0.001

PMA (weeks) -0.4 (0.2) -0.22 0.056
Change in weight SDS from 30 weeks PMA -55.7 (9.4) -0.67 < 0.001
onwards (SDS per week)
Model 2 length gain
Constant 87.0 (14.6) < 0.001

PMA (weeks) -0.6 (0.3) -0.29 0.031
Gestational age (weeks) 0.6 (0.3) 0.25 0.064
Length SDS at term equivalent age -1.7 (0.6) 0.58 0.005
5
Model 3 head circumference growth
Constant 106.1 (11.4) < 0.001

PMA (weeks) -0.6 (0.3) -0.28 0.033
Head circumference SDS at 36 weeks PMA -1.4 (0.6) -0.31 0.020
Change in head circumference SDS from -23.1 (8.9) -0.33 0.013

30 weeks PMA onwards
PMA: postmenstrual age

Model 1:
R² = 0.425, p < 0.001
Backward regression model. Variables not included in the final model: gestational age at birth, week 4
weight SDS, change in weight SDS from birth to week 4, weight SDS at 36 weeks PMA, change in weight
SDS from birth to 36 weeks PMA, weight SDS at term equivalent age.
Model 2:
R² = 0.260, p 0.010
Backward regression model. Variables not included in the final model: Length z-score at 36 weeks
postmenstrual age.
Model 3:
R² = 0.280, p 0.014
Backward regression model. Variables not included in the final model: gestational age at birth, head
circumference SDS at term equivalent age.
103
Chapter 5
Figure 2. Postnatal growth and IGF-I levels in preterm infants

A: Mean weight z-scores (Fenton 2013) up to term equivalent age.
B: Mean length z-scores (Fenton 2013) up to term equivalent age.
C: Mean head circumference z-scores (Fenton 2013) up to term equivalent age.
D: Mean IGF-I levels up to term equivalent age. Umbilical cord blood samples were excluded from this
graph.
104
Figure 3. Weight gain in relation to IGF-I and body composition in preterm infants
105
Chapter 5

A: Change in IGF-I in relation to change in weight SDS up to 36 weeks PMA. r 0.62 (95% CI 0.46 – 0.74, p <
0.001)
B: Change in weight SDS from 30 weeks PMA onwards in relation to fat free mass and fat mass at term
equivalent age.
Partial correlation with fat free mass: r 0.72 (95% CI 0.56 – 0.82, p < 0.001)
Partial correlation with fat mass: r 0.77 (95% CI 0.65 – 0.86, p < 0.001)
(Correlations were controlled for gestational age at birth and postmenstrual age at time of body
composition measurement)
C: Change in weight SDS from 30 weeks PMA onwards in relation to fat free mass percentage at term
equivalent age.
Linear regression model: R²: 0.45 p < 0.001, B for change in weight SDS -54.0 (95% CI -72.0 − -35.9, p < 0.001)
(gestational age at birth and postmenstrual age at the time of body composition measurement were
entered in the regression model as covariates)
IGF-I in relation to growth and body composition

IGF-I levels are shown in figure 2. Between birth and term equivalent age IGF-I
showed a mean (SD) increase of 0.4 (0.2) nmol/L per week. The change in IGF-I over
time was independent of gestational age at birth (r – 0.07, 95% CI -0.27 – 0.16, p 0.608).
The increase in IGF-I between birth and 36 weeks PMA positively correlated with
the concurrent change in weight and head circumference SDS (respectively r 0.61,
95% CI 0.45 – 0.73, p < 0.001 and r 0.33, 95% CI 0.11 – 0.52, p 0.004), but not with the
concurrent change in length SDS (r 0.20, 95% CI -0.02 – 0.41, p 0.081).
106
Potential critical windows for IGF-I

From birth to 4 weeks PNA a greater increase in IGF-I was associated with more fat
free mass at term equivalent age. A similar association was seen for the change in
IGF-I from birth to 36 weeks PMA. (table 7). Of the potential critical windows the
change in IGF-I from birth to 4 weeks PNA, was the strongest predictor. (table 8)
IGF-I levels were not associated with fat or fat free mass percentage. (table 7)
Table 7. Associations between IGF-I and body composition at term age
B (95% CI)² B (95% CI)²
IGF-I 0.14 0.53 -0.2 0.2

4 wks PNA (-0.18 – 0.43) (0.27 – 0.72) (-1.0 – 0.5)a (-0.6 – 1.0)b
IGF-I 0.18 0.16 0.4 0.4

36 wks PMA
IGF-I
(-0.18 – 0.48)
0.24
(-0.19 – 0.47)
0.43
(-0.4 – 1.1)c
0.0
(-1.1 – -0.4)d
0.0
5
Term age (-0.16 – 0.57)3 (0.06 – 0.69)3 (-0.5 – 0.5)e (-0.5 – 0.5)f
Change in IGF-I birth – 4 wks 0.22 0.59 -2.5 2.2

PNA (-0.12 – 0.51) (0.33 – 0.77) (-35.7 – 30.7)g (-32.1 – 36.5)h
Change in IGF-I 0.23 0.33 4.4 -4.6

birth – 36 weeks PMA (-0.04 – 0.47)4 (0.07 – 0.55)4 (-4.2 – 13.1) i4 (-13.5 – 4.2)j4
Change in IGF-I 0.06 0.03 2.0 -2.2

from 30 weeks PMA onwards (-0.21 – 0.32)4 (-0.24 – 0.30)4 (-3.0 – 7.1)k4 (-7.4 – 3.0)l4

3
Also corrected for postmenstrual age at blood sampling
4
Not controlled for gestational age at birth, because gestational age did not correlate with change in
IGF-I across PMA nor did it correlate with body composition.
a
g
R² 0.066, p 0.530; h R² 0.073, p 0.483; i R² 0.069, p 0.160; j R² 0.075, p 0.136; k R² 0.062, p 0.195; l R² 0.069, p 0.159;
107
Chapter 5
Table 8. Predictive value of IGF-I for fat free mass at term equivalent age
B (SE) ß p-value
Included variables
Constant -5204 (1370) 0.001
PMA (weeks) 146 (30) 0.59 < 0.000

Change in IGF-I up to week 4 (nmol/L per week) 4652 (1247) 0.44 0.001
R² = 0.581, p < 0.001

Backward regression model. Variables not included in the final model: gestational age at birth, week 4
IGF-I, change in IGF-I from birth to 36 weeks PMA.
Total caloric intake (kcal/kg/day) up to 36 weeks PMA was associated with fat free
mass at term equivalent age (r 0.364, p 0.044, corrected for PMA at the time of body
composition measurement). In multivariate analysis, including the change in IGF-I
up to 4 weeks PNA and the change in weight SDS from 30 weeks onwards, caloric
intake was no longer a significant predictor. There was no association between
caloric intake and fat (free) mass percentage. Protein, fat and carbohydrate intake
up to 36 weeks was not associated with body composition at term equivalent age.
Comorbidities (NEC, LOS, BPD, ROP, IVH and PDA) were not associated with body
composition at term equivalent age.
Gender and ethnicity were not associated with body composition at term equi-
valent age.
Discussion
This study confirmed that, in preterm infants, the postnatal growth and IGF-I are
associated with body composition at term equivalent age. In particular, higher IGF-I
levels in the first month of life and more weight gain in that period were associated
with a more favorable body composition at term equivalent age. Meanwhile a
greater increase in weight, length or head circumference SDS after this window
was associated with a higher fat mass percentage at term equivalent age.
IGF-I and growth in the first month of life

After preterm birth, nutrient supply is abruptly disrupted. Moreover, as the
transplacental supply of essential growth factors stops, the preterm infant now
solely relies on their own production of these factors. This leads to a drop in IGF-I
and consequent decrease in the SDS for weight, length and head circumference.
(14) Accordingly, our population showed low IGF-I levels and a decline in growth
rate in the first weeks of life.
108
In our study cohort a greater increase in IGF-I in the first month of life was associated
with more fat free mass at term equivalent age. In addition, our results showed that
less decrease in weight SDS in the first month of life was associated with a higher
fat free mass percentage at term equivalent age. The change in length and head
circumference SDS were not associated with body composition. Nevertheless,
while a higher length SDS at 4 weeks postnatal age was associated with both a
higher fat mass as well as a higher fat free mass, a higher head circumference SDS
at 4 weeks postnatal age was associated with higher fat free mass alone.
In line with our findings, Hernandez and colleagues found higher IGF-I levels in the
first week of life to be associated with increased lean mass at 2 years corrected age
in small for gestational age preterm infants. (10)
Therefore, our results suggest that impaired growth in the first month of life, i.e.
lower IGF-I levels and weight and head circumference SDS, is associated with a less
favorable body composition at term equivalent age. 5
Interestingly, we observed that the postnatal change in IGF-I was independent of
gestational age at birth, i.e. the rate of increase of IGF-I depended on postmenstrual
age at the time of blood sampling regardless of the gestational age at birth. Likewise,
Hansen-Pupp and colleagues found IGF-I to increase at 30 weeks gestational age,
irrespective of gestational age at birth(14) Accordingly, in our population IGF-I
showed an increase from 31 weeks PMA. Taking into account that our findings
suggest early IGF-I levels are important for a more favorable body composition,
there may be a window of opportunity for interventions to increase IGF-I levels
in this early phase. Indeed, trials with IGF-I administration between preterm birth
and 30 weeks PMA show promising results with regards to the occurrence of major
comorbidities such as BPD. However, IGF-I is expected to affect various organ
systems. Current studies have not shown a decrease in other comorbidities such
as ROP and the effects of IGF-I administration on body composition are yet to be
investigated . (15) Furthermore, IGF-I administration is an invasive and burdensome
procedure, which might limit its implementation in clinical practice.
IGF-I levels and growth up to term equivalent age

From birth to 36 weeks PMA and at term equivalent age there was still a positive,
but weak, association between change in IGF-I and fat free mass. Nevertheless,
there was no association with the change from 30 weeks PMA onwards. Taking into
account, the stronger association in the first month of life, it could be suggested that
early changes in IGF-I are more important in influencing body composition at term
equivalent age. Nevertheless, given the sample size and observational nature of
this study, the evidence for this hypothesis is inconclusive and remains speculative.
Moreover, this is in contrast to reports by Ruys and colleagues who found a positive
association between IGF-I and fat free mass as well as fat mass measured in
109
Chapter 5
preterm infants at term equivalent age, implying IGF-I relates to overall growth
rather than body composition. (11) Albeit, in relation to later measurements of body
composition, others did find IGF-I to relate to body composition. For example,
Cooper et al. reported a higher increase in IGF-I between hospital discharge and
1 year corrected age to be associated with a concurrent higher increase in fat free
mass in infants born prematurely. (9)
After correcting for gestational age at birth, more weight gain between birth and
36 weeks PMA, and in particular from 30 weeks PMA onwards, was associated with
a lower fat free mass percentage at term equivalent age. Hypothesizing this would
imply that, in contrast to weight gain in the first month of life, high rates of weight
gain are not desirable up to term equivalent age. In line with that, we found that
a higher SD score for length and head circumference at 36 weeks PMA and term
equivalent age, was associated with a lower fat mass percentage. Furthermore, the
increase in head circumference SDS from 30 weeks onwards was associated with
a lower fat free mass percentage. Our data suggest that increased growth rates,
in particular after the first month of life, are associated with a lower fat free mass
percentage and thus a less favorable body composition. We speculate that it might
be beneficial to prevent the decrease in SDS in the first weeks of life. Then there
might be a less rapid increase in SDS afterwards and potentially a more favorable
body composition at term equivalent age. Nevertheless, the fat deposition may be
an adaption to extra-uterine life to enable adequate thermoregulation and provide
energy stores.(16) In that light it might be valuable to develop normative data, with
cut off points where fat accumulation becomes undesirable, for example based on
later cardiometabolic outcomes.
Study limitations
Our study cohort had a relatively low incidence of postnatal growth restriction and
a relatively high mean fat mass percentage. Others previously showed that preterm
infants with postnatal growth restriction (weight SD at term age < -2 SD), had a
lower fat mass percentage compared to preterm infants without postnatal growth
restriction. (17, 18) This could partially explain the higher fat mass percentage in our
cohort. Yet, it makes our results less generalizable. In addition, our effect size was
limited. Despite the sample size of this cohort, the number of IGF-I measurements
per week was low. Blood sampling took place on alternating weeks, and combined
with a relatively small number of extremely preterm infants, this resulted in a
sample size ranging between 1 infant at 26 weeks PMA to 33 infants at 33 weeks
PMA. These conditions may have contributed to our findings.
110
Conclusions
IGF-I plays an important role in growth and body composition in preterm infants.
Higher IGF-I levels in the first month of life are associated with a more favorable
body composition at term equivalent age. It remains to be elucidated what the
optimal body composition would be at term equivalent age. Future studies taking
into account long term health outcomes are warranted to establish useful guidance
for growth and body composition in early life.
111
Chapter 5
References
1. Horbar JD, Ehrenkranz RA, Badger GJ, Edwards EM, Morrow KA, Soll RF, et al. Weight Growth
Velocity and Postnatal Growth Failure in Infants 501 to 1500 Grams: 2000-2013. Pediatrics.
2015;136(1):e84-92.
3. Johnson MJ, Wootton SA, Leaf AA, Jackson AA. Preterm birth and body composition at term
equivalent age: a systematic review and meta-analysis. Pediatrics. 2012;130(3):e640-9.
2019;11(9).
5. Scheurer JM, Zhang L, Plummer EA, Hultgren SA, Demerath EW, Ramel SE. Body Composition
Changes from Infancy to 4 Years and Associations with Early Childhood Cognition in Preterm
and Full-Term Children. Neonatology. 2018;114(2):169-76.
6. Euser AM, Finken MJ, Keijzer-Veen MG, Hille ET, Wit JM, Dekker FW, et al. Associations
between prenatal and infancy weight gain and BMI, fat mass, and fat distribution in young
adulthood: a prospective cohort study in males and females born very preterm. Am J Clin Nutr.
2005;81(2):480-7.
7. Nakano Y. Adult-onset diseases in low birth weight infants: Association with adipose tissue
maldevelopment. Journal of Atherosclerosis and Thrombosis. 2020;27(5):397-405.
2015;77(1-2):156-63.
9. Cooper DM, Girolami GL, Kepes B, Stehli A, Lucas CT, Haddad F, et al. Body composition and
neuromotor development in the year after NICU discharge in premature infants. Pediatr Res.
2020.
weight. J Pediatr Endocrinol Metab. 2012;25(9-10):951-5.
11. Ruys CA, van de Lagemaat M, Lafeber HN, Rotteveel J, Finken MJJ. Leptin and IGF-1 in relation
to body composition and bone mineralization of preterm-born children from infancy to 8
years. Clin Endocrinol (Oxf). 2018;89(1):76-84.
for preterm infants. BMC Pediatr. 2013;13:59.
13. Urlando A, Dempster P, Aitkens S. A new air displacement plethysmograph for the
measurement of body composition in infants. Pediatr Res. 2003;53(3):486-92.
15. Ley D, Hallberg B, Hansen-Pupp I, Dani C, Ramenghi LA, Marlow N, et al. rhIGF-1/rhIGFBP-3 in
Preterm Infants: A Phase 2 Randomized Controlled Trial. J Pediatr. 2019;206:56-65 e8.
16. Sauer PJ. Can extrauterine growth approximate intrauterine growth? Should it? The American
Journal of Clinical Nutrition. 2007;85(2):608S-13S.
17. Roggero P, Gianni ML, Liotto N, Taroni F, Orsi A, Amato O, et al. Rapid recovery of fat mass in
small for gestational age preterm infants after term. PLoS One. 2011;6(1):e14489.
18. Bruckner M, Khan Z, Binder C, Morris N, Windisch B, Holasek S, et al. Extremely Preterm Infants
Have a Higher Fat Mass Percentage in Comparison to Very Preterm Infants at Term-Equivalent
Age. Front Pediatr. 2020;8:61.
112
113
A comparative study using Dual-energy
X-ray absorptiometry, air displacement
plethysmography and skinfolds to assess fat
mass in preterms at term-equivalent age
6
CHAPTER 6
Dana FJ Yumani, Dide de Jongh, Harrie N Lafeber and Mirjam M van Weissenbruch.
Eur J Pediatr. 2021 Mar;180(3):919-927. doi: 10.1007/s00431-020-03812-3.
Chapter 6
Abstract
The aim of this study was to compare whole body composition, generated by
air displacement plethysmography (ADP) and dual-energy x-ray absorptiometry
(DXA), and to evaluate the potential predictive value of the sum of skinfolds (∑SFT)
for whole body composition, in preterm infants at term equivalent age.
A convenience sample of sixty-five preterm infants with a mean (SD) gestational

age of 29 (1.6) weeks was studied at term equivalent age. Fat mass measured by
DXA and ADP were compared and the ability of the ∑SFT to predict whole body fat
mass was investigated.
There was poor agreement between fat mass percentage measured with ADP
compared to DXA (limits of agreement: -4.8% and 13.7%). A previously modeled
predictive equation with the ∑SFT as a predictor for absolute fat mass could not
be validated. Corrected for confounders the ∑SFT explained 42% (ADP, p 0.001)
and 75% (DXA, p 0.001) of the variance in fat mass percentage.
Conclusions
The ∑SFT was not able to accurately predict fat mass and ADP and DXA did not
show comparable results. It remains to be elucidated whether or not a properly
executed DXA provides more accurate assessment of whole body fat mass than
ADP in preterm infants.
116
DXA, ADP & SFT in preterm infants
Introduction
Preterm infants are prone to develop risk factors for the metabolic syndrome in
later life (1). Adolescents and adults born preterm have been shown to have a higher
fat mass, a higher blood pressure and an increased risk of dysglycemia compared
to adolescents and adults born at term (2, 3). While some report no differences in
fat distribution at younger ages (4), others have did find difference in infancy when
comparing the body compostion of infants born preterm to those born at term
(5). For instance, at term equivalent age, premature infants have been reported
to have an increased fat mass compared to term infants (5). Term equivalent age
is an important benchmark for the development of the preterm infant: a point to
evaluate whether any disparities in extra-uterine development and normal fetal
development bear short or long term consequences. Since in adulthood, the fat
mass percentage and the fat mass index have been related to the occurrence of
metabolic syndrome components (6, 7), monitoring body composition in infancy
and childhood could help to signal early signs of increased disease risk. Therefore,
to ensure the timely implementation of preventive measures, it is pertinent to have
a validated method to assess body composition, in particular fat mass.
The most frequently used methods to estimate fat mass are air displacement 6
plethysmography (ADP) and dual-energy X-ray absorptiometry (DXA). There is no
consensus on which reference method should preferentially be used and at the same
time studies in term infants show poor agreement between fat mass measured
with ADP compared to DXA (8-10). To our knowledge, there is no published data on
the comparison of ADP and DXA in preterm infants. Therefore the purpose of the
present study was to compare DXA-generated and ADP-generated whole body
composition in preterm infants at term equivalent age. Even so, ADP and DXA are
both expensive and immobile instruments. Therefore, it would be valuable to have a
reliable and low-cost point-of-care instrument. Skinfold measurements have been
suggested as a low budget tool for measuring fat mass in infants, in particular in
low income countries (11, 12). Nevertheless, there are questions about the reliability
and reproducibility of skinfold measurements. Moreover, despite several studies
assessing predictive equations including skinfolds or the sum of skinfolds (∑SFT)
to estimate fat mass, to our knowledge, only a few included preterm infants (12-
15). In addition, the limited publications on the predictive value of SFT for fat mass
percentage in preterm infants, included mainly late preterm infants (16). All in all,
at this time there are no validated predictive equations including SFT for extremely
and very preterm infants. Therefore, this study assessed the potential predictive
value of the ∑SFT for fat mass and fat mass percentage in preterm infants. In
conclusion, the aim of this study was to assess the agreement between fat and
fat-free mass measured with ADP, DXA, and estimated by the sum of skinfolds, in
preterm born infants at term equivalent age.
117
Chapter 6
Methods
Study cohort
The study cohort consisted of a convenience sample of 65 preterm infants born
between 2015 and 2018, with a gestational age of 24 to 32 weeks, admitted to
the neonatal intensive care unit of the Amsterdam University Medical Centers,
location Vrije Universiteit University Medical Center. The preterm infants were part
of the NUTRIE study, a longitudinal observational study on nutrition in relation to
the endocrine regulation of preterm growth and body composition. The NUTRIE
study was powered to detect a medium size effect (r = 0.35) of Insulin-like growth
factor 1 on fat mass percentage.. No power calculations were done for the primary
outcomes presented in this paper. To demonstrate that the maximum allowed
difference in fat mass measured by two different methods is < 200 g, 10 pairs would
be needed based on a mean difference in the population of 100 g (+ 25) (8).
Informed consent was obtained in the first week of life and participants were
followed up from birth to two years corrected age. Infants with substantial congenital
anomalies based on a chromosomal disorder or syndrome were excluded.
The study was approved by the medical research ethics committee of the Vrije
Universiteit University Medical Center and was conducted according to the good
clinical practice guidelines and in line with the Declaration of Helsinki. The study
was registered at the Dutch Trial Register where an audit trail of changes to the
design was kept (www.trialregister.nl; NTR5311).
Assessment of growth and body composition

Growth and body composition were assessed on the same day in the same order
in all participants. Follow-up at term equivalent age was planned between 38 and
46 weeks postmenstrual age (mean 43.8 + 1.9 weeks). SFT were measured first,
followed by ADP and finally DXA. Infants were fed before the DXA in case the child
was too agitated.
Growth was assessed from birth until 36 weeks postmenstrual age and at term
equivalent age. Measurements of weight, length, and head circumference were
taken by two investigators. Infants were weighed nude on an electronic scale to
the nearest 5 gram and length was measured with a length board to the nearest
0.5 cm. Occipital-frontal head circumference was measured to the nearest 0.1 cm
with a non-stretchable measuring tape. Standard deviation scores (SDS) of weight,
length and head circumferences were calculated according to Fenton (17). Small for
gestational age (SGA) was defined as a birth weight below the tenth percentile (-1.3
SD) and postnatal growth restriction was assumed if, at 36 weeks postmenstrual
age, there was a decrease in weight z-score of more than 1 SD compared to the
birth weight z-score (18).
118
Skinfolds were measured (to the nearest mm) at biceps, triceps, subscapular and
supra-ilial positions with a Harpenden® skinfold caliper by two investigators. One
measurement was taken bilaterally for every position. The bilateral measurements
were averaged to come to one skinfold thickness for every position. According to
previous studies, the intraobserver coefficient of variation is below 3%, however the
inter-observer coefficient is up to 10%(15).
The anthropometric formula which was used to estimate fat mass at term
equivalent age was that of Schmelzle and Fusch (12). This formula was originally
modeled to predict fat mass, measured with DXA, in infants 34 weeks gestational
age and older, using ∑SFT (mm) and length (cm): Fat mass (g) = 68.2 x ∑SFT ^
(0.0162 x length) – 172.8. Skinfolds were measured at the same site as our study.
This formula was selected because it was the only predictive equation based on
a population that included preterm infants and gave a high explanation of the
variance in fat mass (12, 14).
The Pea Pod® (PEA POD Infant Body Composition System, Cosmed, Ltd, Concord,
CA, USA) was used to assess whole body fat mass and fat-free mass through ADP.
The measurements were performed by two investigators. Infants were measured
naked and hair was flattened using hair oil. Infants were allowed to move during
6
the measurement. In case of excessive crying the measurement was stopped.
Measurements were done briefly before feeding time, i.e. approximately 3 hours
after the last feeding. In line with the manufacturer's guideline, daily quality
control checks were done which included chamber calibration. Every two weeks
the scale was calibrated. A detailed description of the Pea Pod® measurement
is described elsewhere(19). As previously reported, the coefficient of variation for
repeated volume measurements lies between 0.02 and 0.09% (19). Fat mass and fat
free mass were calculated using gender-specific equations developed by Fomon
and colleagues (20).
The Hologic QDR 4500 A, using Infant Whole Body Software version 13.5.3:3 (Hologic
Inc, Bedford, MA, USA) was used to assess whole body fat mass and fat-free mass
through dual-energy x-ray absorptiometry. During the procedure, the infants were
required not to move. The infants were swaddled in a blanket of the same size
and type supplied by the investigators, without any clothing or diapers. Infants
were swaddled in supine position with the soles of the feet together and knees
bent (frog-leg-position) and the arms stretched beside the body. Infants were
positioned in the center of the scanning bed with their head near the head end
of the bed. The measurement was done after feeding. Typically infants remained
awake, but lights were dimmed and a video was played from a mobile device
outside of the scanning field. The preparation and positioning of the infants was
performed by two experienced investigators. Calibration was done daily using an
anthropomorphic spine phantom and a geometric block phantom. In addition a
radiographic uniformity test was done once a week and software was regularly
119
Chapter 6
updated. All images were analyzed by one radiologist. Images with excessive
movement artifacts were excluded at the judgement of the radiologist.
The following factors are known to relate to body composition and were assessed
as potential confounders: gestational age, gender, ethnicity, type of nutrition :
human milk (60% or more of total diet) vs formula (60% or more of total diet), waist
circumference and absolute weight, length and head circumference at birth and
term age and their corresponding z-scores (15).
The statistical analysis

Characteristics of the study group were first summarized using descriptive
statistics, stratified by sex. Mean and standard deviations (SD) were calculated for
all continuous variables and presented as mean + SD. Percentages were reported
for dichotomous variables. The median and the interquartile range were reported
if the variable was not normally distributed.
The level of agreement and potential bias between fat mass percentage obtained
via ADP and DXA was examined using the Bland-Altman analysis (12). Agreement
between the formula of Schmelzle et al (12). and fat mass was also examined using
Bland Altman plots. Based on the normal variation in fat mass the maximum
allowed difference was set at 200 g for absolute fat mass and at 2% for fat mass
percentage (21-23).
Prediction models were developed for the predicting variable ∑SFT and absolute
fat mass and fat mass percentage measured with ADP at term equivalent age.
Potential confounders which showed significant correlations with fat mass
(percentage) in univariate analysis were added together in a multivariate model.
The final model was determined through a backward stepwise regression analysis.
The removal criterion was: F-to-remove > 0.10.
All statistical analyses were conducted using IBM® SPSS® Statistics 22 for Windows
(IBM Corp., Armonk, NY, USA). Two-sided statistical significance was assumed at
p-values less than 0.05 with a 95% confidence interval.
Results
Sixty-five infants were assessed for growth and body composition at term age.
Baseline characteristics are shown in table 1. Measurements of skinfolds were
successfully completed in 63 infants, ADP was successful in 58 infants and DXA in
32 infants (see figure 1).
120
Total Male Female p-valuea

n = 65 n = 35 n = 30
Characteristics at birth
Gestational age (weeks), mean + SD 29.0 + 1.6 29.3 + 1.6 28.7 + 1.6 0.132
Race, n (%)
White 52 (80.0) 27 (77.1) 25 (83.3) 0.534
Non-white 13 (20.0) 8 (22.9) 5 (16.7)
Birth weight (grams), mean + SD 1170 + 316 1347 + 295 1108 + 293 0.002
Birth length (cm), mean + SD 37.0 + 3.2 38.6 + 3.0 36.3 + 3.0 0.004
Birth head circ. (cm), mean + SD 26.3 + 2.2 27.4 + 2.3 26.1 + 1.9 0.016
Birth weight SDS, median (IQR) 0.2 (-0.3 – 0.6) 0.3 (-0.1 – 0.9) -0.1 (-0.4 – 0.5) 0.065
Birth length SDS, median (IQR) 0.3 (-0.5 – 0.6) 0.3 (-0.3 – 0.6) 0.1 (-0.6 – 0.5) 0.229
Birth head circ. SDS, median (IQR) 0.4 (-0.2 – 1.0) 0.3 (-0.2 – 1.2) 0.4 (-0.3 – 1.0) 0.246
Small for gestational age ( < p10), n (%) 3 (4.6) 1 (2.9) 2 (6.7) 0.782
Characteristics at term age visit
PMA at term age visit (weeks), mean + SD
Weight at term age visit (grams), mean

43.8 + 1.9
4078 + 662
43.8 + 2.1
4320 + 683
43.8 + 1.6
3795 + 517
0.962
0.001
6
+ SD
Length at term age visit (cm), mean + SD 52.8 + 2.7 53.9 + 2.4 51.5 + 2.5 0.000
Head circ. at term age visit (cm), mean + 37.4 + 1.6 37.8 + 1.6 37.0 + 1.4 0.040
SD
Weight SDS at term age visit, median -0.5 (-1.3 – 0.1) -0.2 (-1.1 – 0.3) -0.7 (-1.5 – 0.0) 0.104
(IQR)
Length SDS at term age visit, median -0.4 (-1.1 – 0.1) -0.2 (-0.7 – 0.3) -1.0 (-1.7 – -0.2) 0.003
(IQR)
Head circ. SDS at term age visit, median 0.5 (-0.1 – 1.3) 0.7 (0.0 – 1.6) 0.3 (-0.1 – 1.2) 0.272
(IQR)
Postnatal growth restrictionb, n (%) 18 (27.7) 7 (20.0) 11 (36.7) 0.134
Type of nutrition
Human milk 29 (44.6) 18 (51.4) 11 (36.7)
Formula 36 (55.4) 17 (48.6) 19 (63.3)
: p-value for females compared to males based on t-test for normal distributions, Mann-Whitney U test
a
for non-parametric variables and Pearson chi-square or Fisher’s exact test for categorical variables.
: Postnatal growth restriction was assumed if, at 36 weeks postmenstrual age, there was a decrease in
b
weight z-score of more than 1 SD compared to the birth weight z-score.

IQR: interquartile range, PMA: postmenstrual age, SD: standard deviation, SDS: standard deviation score
121
Chapter 6
Figure 1. Nutrie study flow diagram
DXA compared to ADP

Compared to ADP, fat mass measured with DXA was higher (254.7 g (165.7 – 343.9).
Likewise, fat mass percentage measured with DXA was 4.5 % (2.7 – 6.2) higher
than ADP. (table 2) There was no agreement between fat mass measured with
DXA compared to ADP. The mean difference was 255 + 234 g with a lower limit
of agreement of -212 g and an upper limit of agreement of 723 g. For fat mass
percentage the mean difference between DXA and ADP was 4.5 + 4.7%, with a
lower limit of -4.8% and an upper limit of 13.7 %. The Bland Altman plot showed
a proportional bias: as the mean fat mass percentage increased the absolute
difference in fat mass percentage between the two methods increased. Based
on a maximum allowed difference of 2%, no agreement was found. (Figure 2) For
122
both ADP and DXA, fat mass and fat mass percentage at term age did not differ
significantly between gender, ethnicity or type of nutrition at term age.
Table 2. Body composition at term equivalent age
Total Male Female p-valuea

N = 65 N = 35 N = 30
Sum of skinfolds (mm)b, mean + SD 22.4 + 4.2 22.4 + 4.4 22.5 + 4.1 0.872
Waist circumference (mm) , median

c
36.0 (34.0 – 37.5) 36.5 (34.5 – 37.8) 35.4 (33.3 – 37.0) 0.167
(IQR)
ADP fat mass (grams)d, mean + SD 864 + 253 910 + 280 799 + 198 0.100
DXA fat mass (grams)e, mean + SD 1078 + 417 1187 + 475 1012 + 376 0.258
ADP fat mass percentage, median 20.7 (18.4 – 23.0) 21.2 (18.4 – 23.4) 20.4 (18.1 – 22.6) 0.594
(IQR)
DXA fat mass percentage, median 25.0 (21.5 – 30.5) 25.1 (22.4 -31.8) 24.8 (19.8 – 29.5) 0.684
(IQR)
ADP fat-free mass (grams), mean 3309 + 462 3447 + 430 3114 + 442 0.006
+ SD
6
DXA fat-free mass (grams), mean 3138 + 400 3316 + 422 3032 + 356 0.051
+ SD
ADP fat-free mass percentage, 79.3 (77.1 – 81.6) 78.8 (76.6 – 81.7) 79.7 (70.5 – 81.9) 0.594
median (IQR)
DXA fat-free mass percentage, 75.1 (69.6 – 78.5) 74.9 (68.2 – 77.7) 75.2 (70.5 – 80.2) 0.686
median (IQR)
a
: p-value for females compared to males based on t-test for normal distributions, Mann-Whitney U test
for non-parametric variables and Pearson chi-square or Fisher’s exact test for categorical variables.
b
: Sum of skinfolds measured at biceps, triceps, subscapular and supra-iliac positions, n = 63 (34 male, 39
female)
c
:
Waist circumference, n = 63 (34 male, 39 female)
d
:
ADP, n = 58 (34 male, 24 female)
e
:
DXA, n = 32 (12 male, 24 female)
ADP: air displacement plethysmography, DXA: Dual-energy x-ray, IQR: interquartile range, SD: standard
deviation
123
Chapter 6
Figure 2. Bland-Altman plot of fat mass percentage measured by DXA compared with ADP
ADP air displacement plethysmography, DXA dual-energy X-ray

Average of fat mass measured with DXA and ADP is depicted on the x-axis and the difference between
the fat mass percentage measured with DXA and ADP is depicted on the y-axis. Mean difference: 4.5 ±
4.7%, lower limit of agreement: - 4.8%, upper limit of agreement: 13.7%, maximum allowed difference: 2%
DXA compared to skinfolds

The difference between fat mass estimated through the model of Schmelzle and
Fusch, and the fat mass measured with DXA exceeded the limits of agreement.
The mean difference was 272 + 240 g, with a lower limit of agreement of -742
g and an upper limit of agreement of 199 g. Fat mass percentage derived from
skinfolds did not agree with fat mass percentage measured with DXA and showed
a proportional bias with a larger difference in fat mass percentage with increasing
mean fat mass percentage. (Figure 3)
Predictive model for fat mass measured with DXA

Within our cohort, fat mass, measured with DXA, could be estimated with
gestational age, waist circumference, length and the ∑SFT: Fat mass (g) = -4649.1
+ 23.5*∑SFT + 64.4*length + 77.6*waist circumference – 33.7*gestational age (∑SFT
in mm, length and waist circumference in cm and gestational age in weeks). These
factors explained 89% of the variance (R2 = 0.893, S.E. of the estimate 146 g, , p <
0.001). In addition, 75% of the variance in fat mass percentage, measured with DXA,
could be explained by waist circumference, head circumference and the ∑SFT (R2
= 0.753, S.E. of the estimate 3.5 % 4, p < 0.001).
124
Figure 3. Bland-Altman plot of fat mass percentage measured by DXA compared with fat mass percentage
estimated based on the formula by Schmelzle and Fusch [18].
DXA dual-energy X-ray, ΣSFT sum of skinfold measurements

Average of fat mass percentage measured with DXA and estimated with the formula is depicted on the
x-axis and the difference between the fat mass percentage measured with DXA and estimated by the
formula is depicted on the y-axis. Mean difference: 4.9 ± 5.4%, lower limit of agreement − 5.6%, upper limit
of agreement 15.5%, maximum allowed difference 2%
6
Predictive model for fat mass measured with ADP
Within our cohort, fat mass, measured with ADP, could be estimated with
gestational age, waist circumference, head circumference, weight SDS, head
circumference SDS and the ∑SFT: Fat mass (g) = -3013.0 – 9.4*gestational
age + 39.1*waist circumference + 65.9*head circumference + 67.6*weight SDS
– 59.3*head circumference sds + 15.1*∑SFT (Gestational age in weeks, waist and
head circumference in mm and ∑SFT in mm). These factors explained 72% of the
variance (R2 = 0.716, S.E. of the estimate 138.4 g, p <0.001).
Fourty-two percent of the fat mass percentage measured with ADP could be
explained by the ΣSFT and waist circumference(R2 =0.426, S.E. of the estimate 3.1 %
<0.001)
In multivariate analysis other potential confounders, were found to not be

significant.
125
Chapter 6
Discussion
This study showed that there is poor agreement between body composition
measured with ADP and body composition measured with DXA in preterm born
infants at term equivalent age. Compared to ADP, DXA showed higher fat mass
percentages. Furthermore, estimations of fat mass based on the ∑SFT showed
poor agreement with the actual fat mass measured with DXA.
Various studies in term infants report high correlations between fat mass
measured with ADP and fat mass measured with DXA (9, 10). Nevertheless, a
high correlation does not imply both methods found the same value and doesn’t
provide information about the test quality (24). Similar to studies performed in
full term infants, DXA gave higher estimates of fat mass in our cohort compared
to estimates by ADP (9, 10). In agreement with that, early animal studies showed
that DXA seems to overestimate fat mass (25, 26). To our knowledge, no data
has been published on the comparison of DXA and ADP in extremely preterm
infants. Nevertheless, one recent study in South-African term infants also showed
higher estimates of fat mass by ADP compared to DXA (8). Moreover, several
reviews have highlighted that both DXA and ADP have reasonable reproducibility,
but only modest accuracy. According to these reviews ADP actually seems to
underestimate fat-free mass percentage or fat-free mass expressed in grams
per liter (fat-free mass density). Especially, when the fat-free mass percentage or
density gets higher, the underestimation becomes larger (15, 27). In actual fact the
fat-free mass percentage or density may be a more relevant parameter to assess,
as in practice it may be more insightful to properly predict fat mass and fat-free
mass percentage than it is to predict absolute fat and fat-free mass. Nonetheless,
in all these studies it is to be questioned whether an appropriate reference method
for body composition has been used. In practice, both DXA and ADP, as well as
deuterated water, have been deemed as reliable methods, however, there seems
to be no universally accepted preferential reference method in living infants.
In contrast to DXA, ADP takes into account that hydration status is different in
infants as compared to adults. Particularly during the first week of life, infants’ fat-
free mass hydration is higher, and therefore DXA estimations of fat and fat-free
mass may not be as accurate as ADP estimations in this period (28). Moreover, the
algorithms used in DXA software are not open for critical analysis. In addition, DXA
quality is negatively influenced by movement, while moderate movement does
not affect body composition measurements taken using ADP. Moreover, infants
are exposed to a low dose of radiation. Therefore, in this study, and presumably in
others as well, ADP measurements may have been more reliable.
In line with others, we could not externally validate the model by Schmelze et al. for
fat mass prediction (29). To date, predictive models for the estimation of fat mass
using the SFT, have only been validated in term and late preterm infants and the
126
predictive value of SFT alone was generally low (16, 30). Moreover, these models were
mostly only internally validated and looked at the prediction of absolute fat mass and
not fat mass percentage (12, 13, 29). However, in our cohort it seems that the ∑SFT
could also explain an important part of the variance in fat mass percentage. Even
though the prediction model yielded a lower R² than the model for the prediction of
fat mass, fat mass percentage may be a more generalizable factor and worth further
exploration for external validation. Especially, in the light of resource poor settings,
the ∑SFT might still be useful as an indicator of fat mass percentage.
Our study was limited by the small sample size, which reduces the generalizability
of the prediction models. Moreover, body composition was measured between 38
and 46 weeks postmenstrual age, a period in which body composition alters (31). In
addition, the low number of successful DXA scans, with 1 in 4 scans not completed
because of excess movement or too much agitation prior to the measurement,
limited the assessment of agreement between different methods. Recent studies in
term infants have shown that placing infants in a vacuum cushion limits movement
artefacts and leads to more comparable results between DXA and ADP (32).
This study has not been able to robustly show that skinfold measurements qualify as
a reliable, low-cost, point of care instrument. However, it remains desirable to find an
6
easily accessible and reliable way of monitoring fat mass in light of possible adverse
cardiometabolic outcomes in later life(1-3). Nevertheless, currently available methods
for bedside assessment of body composition, such as bioelectrical impedance
analysis and body proportionality measures, have a questionable accuracy and
accurate, low-cost bedside methods are limited (30). To the best of our knowledge,
other predictive equations including weight and length indices and easily measured
clinical parameters are yet to be externally validated.(33-35) It would be of interest
to further investigate the potential of these predictive equations. Taking previous
findings into account, ADP seems to be more practical to assess body composition,
in particular fat mass, in preterm infants in early life. Nevertheless, it remains to
be elucidated whether or not a DXA without movement artifacts provides a more
accurate assessment of whole body composition than ADP in preterm infants.
127
Chapter 6
References
1. Raju TNK, Pemberton VL, Saigal S, Blaisdell CJ, Moxey-Mims M, Buist S. Long-Term Healthcare
Outcomes of Preterm Birth: An Executive Summary of a Conference Sponsored by the National
Institutes of Health. The Journal of pediatrics. 2017;181:309-18.e1.
2. Kerkhof GF, Breukhoven PE, Leunissen RW, Willemsen RH, Hokken-Koelega AC. Does preterm
birth influence cardiovascular risk in early adulthood? The Journal of pediatrics. 2012;161(3):390-
6.e1.
3. Sipola-Leppanen M, Vaarasmaki M, Tikanmaki M, Matinolli HM, Miettola S, Hovi P, et al.
Cardiometabolic risk factors in young adults who were born preterm. American journal of
epidemiology. 2015;181(11):861-73.
4. Forsum EK, Flinke E, Olhager E, body composition study g. Premature birth was not associated
with increased body fatness in four-year-old boys and girls. Acta paediatrica (Oslo, Norway :
1992). 2020;109(2):327-31.
5. Johnson MJ, Wootton SA, Leaf AA, Jackson AA. Preterm birth and body composition at term
equivalent age: a systematic review and meta-analysis. Pediatrics. 2012;130(3):e640-9.
6. Liu P, Ma F, Lou H, Liu Y. The utility of fat mass index vs. body mass index and percentage of
body fat in the screening of metabolic syndrome. BMC Public Health. 2013;13:629.
7. Ramirez-Velez R, Correa-Bautista JE, Sanders-Tordecilla A, Ojeda-Pardo ML, Cobo-Mejia EA,
Castellanos-Vega RDP, et al. Percentage of Body Fat and Fat Mass Index as a Screening Tool for
Metabolic Syndrome Prediction in Colombian University Students. Nutrients. 2017;9(9).
8. Wrottesley SV, Pisa PT, Micklesfield LK, Pettifor JM, Norris SA. A comparison of body composition
estimates using dual-energy X-ray absorptiometry and air-displacement plethysmography in
South African neonates. European journal of clinical nutrition. 2016;70(11):1254-8.
9. Fields DA, Demerath EW, Pietrobelli A, Chandler-Laney PC. Body composition at 6 months of
life: comparison of air displacement plethysmography and dual-energy X-ray absorptiometry.
Obesity (Silver Spring, Md). 2012;20(11):2302-6.
10. Barbour LA, Hernandez TL, Reynolds RM, Reece MS, Chartier-Logan C, Anderson MK, et al.
Striking differences in estimates of infant adiposity by new and old DXA software, PEAPOD and
skin-folds at 2 weeks and 1 year of life. Pediatric obesity. 2016;11(4):264-71.
11. Chen LW, Tint MT, Fortier MV, Aris IM, Shek LP, Tan KH, et al. Which anthropometric measures
best reflect neonatal adiposity? International journal of obesity (2005). 2018;42(3):501-6.
12. Schmelzle HR, Fusch C. Body fat in neonates and young infants: validation of skinfold
thickness versus dual-energy X-ray absorptiometry. The American journal of clinical nutrition.
2002;76(5):1096-100.
13. Deierlein AL, Thornton J, Hull H, Paley C, Gallagher D. An anthropometric model to estimate
neonatal fat mass using air displacement plethysmography. Nutrition & metabolism. 2012;9:21.
14. Koo WW, Walters JC, Hockman EM. Body composition in neonates: relationship between
measured and derived anthropometry with dual-energy X-ray absorptiometry measurements.
Pediatr Res. 2004;56(5):694-700.
15. Demerath EW, Fields DA. Body composition assessment in the infant. Am J Hum Biol.
2014;26(3):291-304.
16. Daly-Wolfe KM, Jordan KC, Slater H, Beachy JC, Moyer-Mileur LJ. Mid-arm circumference is a
reliable method to estimate adiposity in preterm and term infants. Pediatr Res. 2015;78(3):336-41.
for preterm infants. BMC Pediatr. 2013;13(1):59.
18. Zozaya C, Diaz C, Saenz de Pipaon M. How Should We Define Postnatal Growth Restriction in
Preterm Infants? Neonatology. 2018;114(2):177-80.
19. Urlando A, Dempster P, Aitkens S. A new air displacement plethysmograph for the
measurement of body composition in infants. Pediatric research. 2003;53(3):486-92.
128
20. Fomon SJ, Haschke F, Ziegler EE, Nelson SE. Body composition of reference children from birth
to age 10 years. The American journal of clinical nutrition. 1982;35(5 Suppl):1169-75.
21. Ahmad I, Nemet D, Eliakim A, Koeppel R, Grochow D, Coussens M, et al. Body composition and
its components in preterm and term newborns: A cross-sectional, multimodal investigation.
American journal of human biology : the official journal of the Human Biology Council. 2010;22(1):
69-75.
22. Dung NQ, Fusch G, Armbrust S, Jochum F, Fusch C. Body composition of preterm infants
measured during the first months of life: bioelectrical impedance provides insignificant
additional information compared to anthropometry alone. Eur J Pediatr. 2007;166(3):215-22.
23. Liotto N, Roggero P, Bracco B, Menis C, Morniroli D, Perrone M, et al. Can Basic Characteristics
Estimate Body Composition in Early Infancy? J Pediatr Gastroenterol Nutr. 2018;66(3):e76-e80.
24. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of
clinical measurement. Lancet. 1986;1(8476):307-10.
25. Rigo J, Nyamugabo K, Picaud JC, Gerard P, Pieltain C, De Curtis M. Reference values of body
composition obtained by dual energy X-ray absorptiometry in preterm and term neonates. J
Pediatr Gastroenterol Nutr. 1998;27(2):184-90.
26. Brunton JA, Weiler HA, Atkinson SA. Improvement in the accuracy of dual energy x-ray
absorptiometry for whole body and regional analysis of body composition: validation using
piglets and methodologic considerations in infants. Pediatric research. 1997;41(4 Pt 1):590-6.
27. Mazahery H, von Hurst PR, McKinlay CJD, Cormack BE, Conlon CA. Air displacement
6
plethysmography (pea pod) in full-term and pre-term infants: a comprehensive review of
accuracy, reproducibility, and practical challenges. Matern Health Neonatol Perinatol. 2018;4:12.
28. Carberry AE, Colditz PB, Lingwood BE. Body composition from birth to 4.5 months in infants
born to non-obese women. Pediatric research. 2010;68(1):84-8.
29. Cauble JS, Dewi M, Hull HR. Validity of anthropometric equations to estimate infant fat mass at
birth and in early infancy. BMC Pediatr. 2017;17(1):88.
30. Andrews ET, Beattie RM, Johnson MJ. Measuring body composition in the preterm infant:
Evidence base and practicalities. Clin Nutr. 2019.
31. Norris T, Ramel SE, Catalano P, Caoimh CN, Roggero P, Murray D, et al. New charts for the
assessment of body composition, according to air-displacement plethysmography, at birth
and across the first 6 mo of life. The American journal of clinical nutrition. 2019;109(5):1353-60.
32. de Fluiter KS, van Beijsterveldt I, Goedegebuure WJ, Breij LM, Spaans AMJ, Acton D, et al.
Longitudinal body composition assessment in healthy term-born infants until 2 years of age
using ADP and DXA with vacuum cushion. European journal of clinical nutrition. 2020.
33. Ramel SE, Zhang L, Misra S, Anderson CG, Demerath EW. Do anthropometric measures
accurately reflect body composition in preterm infants? Pediatric Obesity. 2017;12:72-7.
34. Larcade J, Pradat P, Buffin R, Leick-Courtois C, Jourdes E, Picaud JC. Estimation of fat-free
mass at discharge in preterm infants fed with optimized feeding regimen. Journal of Pediatric
Gastroenterology and Nutrition. 2017;64(1):115-8.
35. Kiger JR, Taylor SN, Wagner CL, Finch C, Katikaneni L. Preterm infant body composition cannot
be accurately determined by weight and length. J Neonatal Perinatal Med. 2016;9(3):285-90.
129
7
CHAPTER 7
Body composition in preterm infants:
a systematic review on measurement
methods
Dana FJ Yumani, Dide de Jongh, Johannes CF Ket, Harrie N Lafeber,

Mirjam M van Weissenbruch
Pediatr Res. 2022 Aug 22. doi: 10.1038/s41390-022-02262-x. Epub ahead of print.
Chapter 7
Abstract
Background
There are several methods to measure body composition in preterm infants. Yet,
there is no agreement on which method should be preferred.
Methods
PubMed, Embase.com, Wiley/Cochrane Library, and Google Scholar were searched
for studies that reported on the predictive value or validity of body composition
measurements in preterms, up to 6 months corrected age.
Results
Nineteen out of 1884 identified studies were included. Predictive equations based
on weight and length indices, body area circumferences, skinfold thickness,
bioelectrical impedance and ultrasound did not show agreement with body
composition measured with air displacement plethysmography (ADP), dual-energy
x-ray absorptiometry (DXA), magnetic resonance imaging (MRI) or isotope dilution.
ADP agreed well with fat mass density measured by isotope dilution (bias -0.002 g/
ml, limits of agreement + 0.012 g/ml, n = 14). Fat mass percentage measured with
ADP did not agree well with fat mass percentage measured by isotope dilution
(limits of agreement up to + 5.8%) and the bias between measurements was up
to 2.2%. DXA, MRI and isotope dilution were not compared to another reference
method in preterms.
Conclusions
DXA, ADP and isotope dilution methods are considered trustworthy validated
techniques. Nevertheless, this review showed that these methods may not yield
comparable results.
132
Body composition measurement methods in preterm infants
Introduction
It is becoming more evident that preterm birth increases the incidence of risk
factors for cardiovascular disease in early as well as later life. Preterm birth is
associated with a higher percentage of body fat, higher blood pressure, and an
increased risk of dysglycemia from infancy into adulthood.1-6 Furthermore, preterm
birth has been associated with an increased risk of ischemic heart disease in
adulthood.7 Therefore, it is of utmost importance to identify proper screening tools
to identify risk factors for cardiovascular disease at an early age and implement
preventive measures accordingly.
As body fat percentage and fat mass index have been shown to positively
correlate with the occurrence of metabolic syndrome components 8,9, monitoring
body composition during early life could help to implement timely preventive
measures. To gain more insight into which methods should be used to monitor
body composition in early life in preterm infants, this systematic review will assess
validation studies in preterm infants from birth up to 6 months corrected age.
Several methods to assess body composition are available - ranging from

inexpensive, bedside techniques, such as skinfold thickness (SFT) measurement
to expensive and bulky equipment, like air displacement plethysmography (ADP).
10,11
Currently, ADP and dual energy x-ray absorptiometry (DXA) are frequently used
in research settings and are considered as reliable methods. 12 Nevertheless, these
methods are not widely implemented in clinical practice. Furthermore, there is
7
no consensus on which method should preferentially be used to assess body
composition in preterm infants. This systematic review aims to determine the
validity of different methods used to measure body composition in preterm infants
and to show whether validated methods yield comparable results.
Methods
This systematic review primarily assessed the accuracy of various methods used to
measure or estimate body composition in infants born preterm.
A systematic literature review was conducted according to the Preferred Reporting

Items for Systematic Reviews and Meta-analyses (PRISMA)-statement (www.
prisma-statement.org). The review was registered and the protocol was published
on PROSPERO International prospective register of systematic reviews under ID
CRD42018107821. Searches were performed in PubMed, Embase.com, and Wiley/
Cochrane Library from inception (1809) up to 29 September 2020 by D.F.J.Y. and
J.C.F.K. The search included keywords and free-text terms for ‘premature’ and
‘body composition’ or ‘adiposity’ or ‘lean’. Animal studies, conference papers and
editorial letters were excluded. No language or publication date restrictions were
applied. The full search strategies can be found in table 1.
133
Chapter 7
In addition, the reference lists of relevant articles and Google Scholar were checked.
Where needed authors were contacted for clarification or additional information.
Table 1. Search strategies
Search strategy for PubMed (29 September 2020)

Search Query Results
#4 #3 NOT (animals[mh] NOT humans[mh]) 337
#3 #1 AND #2 339
#2 “Body Composition”[Mesh] OR fatty tissue*[ti] OR adipos*[ti] OR body 100890

composition*[ti] OR fat mass*[ti] OR fat percentage*[ti] OR lean mass[ti] OR
lean body mass[ti] OR fat free mass[ti] OR body fat[ti] OR ffm[ti] OR fat free
body mass[ti] OR lean weight*[ti] OR fatty tissue*[ot] OR adipos*[ot] OR body
composition*[ot] OR fat mass*[ot] OR fat percentage*[ot] OR lean mass[ot]
OR lean body mass[ot] OR fat free mass[ot] OR body fat[ot] OR ffm[ot] OR
fat free body mass[ot] OR lean weight*[ot]
#1 “Infant, Premature”[Mesh] OR “Premature Birth”[Mesh] OR “Intensive Care, 86314

Neonatal”[Mesh] OR “Intensive Care Units, Neonatal”[Mesh] OR nicu*[tiab]
Search strategy for Embase.com (29 September 2020)

#5 #4 NOT (‘conference abstract’/it OR ‘conference paper’/it OR ‘editorial’/it OR 589

‘letter’/it OR ‘note’/it)
#4 #3 NOT ([animals]/lim NOT [humans]/lim) 803
#3 #1 AND #2 911
#2 ‘immature and premature labor’/de OR ‘premature labor’/exp OR ‘prematurity’/ 341349

exp OR prematur*:ti,ab,kw OR preterm*:ti,ab,kw OR ‘newborn intensive care’/exp
OR nicu*:ti,ab,kw
#1 ‘body composition’/mj/exp OR ‘adipose tissue’/mj/exp OR ((fatty NEAR/3 120114

tissue*):ti,kw) OR adipos*:ti,kw OR ‘body composition*’:ti,kw OR ‘fat mass*’:ti,kw
OR ‘fat percentage*’:ti,kw OR ‘lean mass’:ti,kw OR ‘lean body mass’:ti,kw OR ‘fat
free mass’:ti,kw OR ‘body fat’:ti,kw OR ffm:ti,kw OR ‘fat free body mass’:ti,kw OR
‘lean weight*’:ti,kw
Search strategy for Wiley/Cochrane Library (29 September 2020)

#1 (prematur* OR preterm* OR (neonatal NEXT intensive NEXT care) OR 29187

nicu*):ab,ti,kw
#2 ((fatty NEAR/3 tissue*) OR adipos* OR (body NEXT composition*) OR (fat NEXT 24093
mass*) OR (fat NEXT percentage*) OR (lean NEXT mass) OR (lean NEXT body
NEXT mass) OR (fat NEXT free NEXT mass) OR (body NEXT fat) OR ffm OR (fat
NEXT free NEXT body NEXT mass) OR (lean NEXT weight*)):ti,ab,kw
#3 #1 AND #2 280
Cochrane Reviews: 17; CENTRAL: 263.
Google Scholar query

(~premature|~preterm|“neonatal intensive care”|nicu) AND (“fatty +/-193000
tissue”|adipose|”body composition”|”fat mass”|”fat percentage”|”lean mass”|”lean
body mass”|”fat free mass”|”body fat”|”fat free body mass”|”lean weight”)
The first 949 hits were retrieved and screened using Publish or Perish 6 (Harzing, A.W. (2007)
Publish or Perish, available from https://harzing.com/resources/publish-or-perish)
134
Study eligibility criteria

Studies were included if they reported on body composition measurement in
infants born before 37 weeks of gestation. The body composition measurement
had to take place between birth and 6 months corrected age. Studies needed to
evaluate methods which measure or estimate fat (-free) mass in (kilo)grams or
percentage. Studies measuring or estimating total body water (TBW) were also
included. See table 2 for a description of the included methods 12,13.
In addition to validation studies, randomized controlled trials, cohort studies and

epidemiologic studies where included if they reported the accuracy or predictive
values of body composition measurements.
Definitions
A method was deemed validated if the method showed good statistical agreement
to a reference method. Currently there is no golden standard for the measurement
of body composition. Hence, studies of all possible reference methods were
accepted.
Good agreement was defined as a maximum allowed difference of 10% of the mean
value of the body composition parameter in the study population. For example, if
the mean of the fat free mass in a study population was 2000 g, then the bias +
limits of agreement had to be smaller than + 200 g. 7
For studies where agreement was not assessed, effect size of the different methods
was determined by the r-squared value. A value below 0.5 was considered a poor
predictive value, 0.5 to 0.7 as a moderate predictive value and > 0.7 was considered
as good predictive value.14
Data extraction
Two reviewers (D.F.J.Y. and D.d.J.) separately screened the studies, initially based on
title and abstract, followed by full text review of relevant studies.
135
Chapter 7
Table 2 . In vivo techniques for measuring body composition in preterm born infants 12,13
Technique Method Application

Body proportionality calculations Ponderal index (kg/m3), Body mass These indices are generally
index (BMI) (kg/m2), Benn index included in regression
(weight/lengthp), waist, mid-arm, and models to formulate
mid-thigh circumference, mid-arm/ predictive equations that
head circumference ratio estimate fat (free) mass (g)
or fat (free) mass percentage
Skinfold thickness measurements Skinfold thickness is measured with The absolute value of the
a caliper to estimate subcutaneous skinfold at a particular site
fat or the sum of skinfolds
measured at different sites
are generally included
in regression models
to formulate predictive
equations that estimate fat
(free) mass (g) or fat (free)
mass percentage
Isotope dilution Isotope distribution is measured in Total body water is

blood or urine. calculated from the isotope
distribution. Fat free mass
(g) is calculated from total
body water.
Bioelectrical impedance analysis The flow of electrical currents Total body water is
/ Bioelectrical impedance through tissue is measured calculated from the
spectroscopy resistance to the flow of
electrical currents running
to body tissue (Total body
water = pL2/Z; p = volume
resistivity in Ω-cm, L = length
in cm, Z = impedance in Ω)
Ultrasound Measures muscle area Estimates subcutaneous fat

by subtracting the muscle
cross sectional area from the
total cross sectional area
Magnetic Resonance Imaging Measures adipose tissue volume Measures fat mass (g)
based on tissue specific energy
absorption
Dual-energy x-ray absorptiometry Measures fat mass and fat free Measures fat (free) mass (g)
mass based on tissue specific x-ray
attenuation
Air displacement Measures body volume based on air Calculates fat mass (g)
plethysmography displacement in a test chamber and fat mass percentage
through age- and sex-
specific density coefficients
Data extraction was performed by these two reviewers. The data collected included
the method and reference method, details on the study setting, methods and
results. In case of any discrepancies between the two reviewers, the two reviewers
came to agreement through discussion.
Reviewers were not blinded for authors or journal details. Where needed, authors
were contacted for clarification or additional information.
136
Risk of bias (quality) assessment

Two reviewers (D.F.J.Y. and D.d.J.) primarily assessed bias using the Critical Appraisal
Skills Program (CASP) checklist. 15 In case of any discrepancies between the two
reviewers, the two reviewers discussed and asked the expert opinion of the two
other reviewers until an agreement was made. The synthesis was based on the final
decision made under agreement of all reviewers. The quality of individual studies
was assessed with CASP checklists. In addition the Oxford Centre for Evidence-
based Medicine’s Levels of Evidence was used to grade the level of evidence of
each manuscript.16
Strategy for data synthesis

A narrative synthesis was primarily done by two researchers (D.F.J.Y. and M.M.v.W.)
and was reviewed by D.d.J., J.C.F.K. and H.N.L. before finalization.
Results
Out of 1884 identified records, 48 full-text articles were assessed for eligibility and
19 were included in this synthesis. (Figure 1) Nine studies (n=1539) reported about
the predictive value or validity of body proportionality measures. Five studies (n
= 319) investigated the validity of bioelectrical impedance analysis (BIA), three
studies (n=90) investigated the validity of SFT, two studies (n=24) investigated the
validity of ADP, one study (n=63) investigated the predictive value of ultrasound 7
and one study (n=15) investigated the validity of MRI. There were no human studies
which reported about the validation of DXA and isotope dilution studies in preterm
infants. (Table 3) Body composition measurements were performed at various
postnatal ages, ranging from 24 hours postpartum to 4 months corrected age.
Body proportionality measures

Table 3 shows our findings on the predictive value and validity of body proportionality
measures. Weight and length indices had a moderate to good predictive value for
fat-free mass (in grams). On the contrary, the predictive value of weight and length
indices for fat mass was poor to moderate and fat mass percentage was poorly
predicted by weight/length indices. 17-21
Larcade et al. and Simon et al. assessed predictive equations with clinical parameters,
such as caloric and macronutrient intake, and z-scores for weight, length and head
circumference.22,23 They found that fat-free mass (g) could be predicted by the
amount of human milk feeding, respiratory support, antenatal corticosteroid use,
growth parameters and sex. A newly modeled equation by Larcade et al showed
good agreement for fat free mass (g). However, Larcade and colleagues did not
assess fat mass percentage, while Simon and colleagues could only explain 24% of
the variance in fat mass percentage with their predictive model. 22,23 (see table 3)
137
Chapter 7
Figure 1. PRISMA flow diagram of included studies
Daly-Wolfe and colleagues reported that mid-arm circumference had a moderate

predictive value for fat mass percentage measured by ADP.24(see table 3) On
the other hand, Koo et al. measured chest, abdomen, midthigh and midarm
circumference. After including sex, race, gestation, weight and length in a predictive
equation for fat mass (g) and fat free mass (g), these body and limb circumferences
did not explain any additional part of the variance in fat (free) mass (g).17 Of
note, these findings were based on the entire study population which included
both term and preterm infants (respectively n = 68 and n = 52) who were large,
appropriate as well as small for gestational age. Nevertheless, Pereira da Silva and
colleagues investigated a group of exclusively preterm infants and reported mid-
arm circumference to have a poor predictive value arm fat area (mm2) measured
by MRI.25 (see table 3)
138
SFT measurements
Table 4 describes the predictive value of SFT. Schmelzle et al. showed that the sum
of SFT measured at four sites had a good predictive value for fat mass (g) in a study
population that included both preterm as well as term infants.26 However, only 10
out of 104 infants in this study population were born preterm.
Koo and colleagues also assessed the predictive value of SFT in a mixed population
of term and preterm infants (respectively n = 68 and n = 52). They reported that SFT,
when added to weight and length, explained an additional 13% of the variance in
fat free mass (g).17 Thus, SFT had a poor predictive value for fat free mass (g).
Schmelzle et al. as well as Koo et al. did not assess the predictive value of SFT for
fat (free) mass percentage. In contrast, Daly-Wolfe et al. did and reported a poor
predictive value of SFT for fat mass percentage (see table 4).24
Bioelectrical impedance measurements

Table 5 describes the predictive value and validity of bioelectrical impedance
measurements. The impedance index (height2 in cm2/impedance in Ω) measured
with BIA adds little to the variance in fat-free mass already explained by weight.27
(see table 5) Indeed Raghavan et al. reported that the least bias was obtained
when weight alone was used to estimate TBW.28 Table 5 shows that models used
to estimate body composition based on the impedance index alone showed poor 7
agreement. In contrast, the predictive equation by Dung et al. based on weight
and the impedance index showed good agreement.27 (see table 5)
Table 6 describes the predictive value of ultrasound measurements. Ultrasound

measurements of muscle and fat mass showed high reliability (intraclass
correlation coefficient 0.874 - 0.975; technical error of measurements 0.251 - 0.628
mm), but had a poor predictive value for fat mass percentage measured by ADP.
29
(see table 6)
Table 7 describes the predictive value of MRI. Dyke and colleagues assessed the
accuracy of body composition measured with rapid whole body MRI. Repeated
scans showed good agreement of fat mass percentage (95% limits of agreement
1.3%). 30 However, body composition measurement was not compared to a reference
method. (see table 7)
Table 8 describes the validity of ADP. Compared to isotope dilution, ADP showed
good agreement when measuring fat free mass density. 31 Forsum et al. and
Roggero et al. demonstrated that there was a small bias in the measurement of
fat mass percentage.31,32 Nevertheless, the limits of agreement were relatively wide
and thus there was poor agreement between fat mass percentage measured with
ADP compared to fat mass percentage measured by isotope dilution. (see table 8)
139
Chapter 7
Table 3. Body proportionality calculations to estimate body composition in preterm infants
Author, year, Predictor for body Reference Study setting Re

country composition method
Subjects N° preterms Time of
assessment
Koo, 2004, Circumferences of head, DXA 120 preterm 52 3 (+ 2.8) days Re
United states chest, abdomen, midarm (Hologic and term (GA 30 - 36 postnatal age
of America & midthigh QDR 1000/W infants weeks) We
densitometer, fre
Midarm & midthigh pediatric software in
muscle and fat areas version 5.64P)
Mi
Total body weight mu
Length va
Ch
no
ma
len
Daly-Wolfe, Mid-arm, mid-thigh, ADP 56 preterm and 28 Discharge Mi

2015, United abdominal circumference term infants (GA 33.3 + 0.9 home pe
States of weeks)
America
Pereira da Silva, Mid-arm circumference Regional adiposity 30 preterm 30 35 + 1 weeks Mi

2009, Portugal measured by MRI infants (GA 30.7 + 1.9 PMA arm
Upper arm muscle and weeks)
fat areas (calculated from Ar
mid-arm circumference of
and triceps skinfold
thickness) Mi
arm
Kiger, 2016, BMI ADP 239 preterm 239 PMA < 50: 51%
United States Weight-for-length infants (GA 28.1 [27.1 – median 37 (IQR ex
of America Ponderal index Body 30.1] weeks) 34 – 39) weeks an
surface area
Benn index
PMA > 50: Th
median 54 (IQR
52-65) weeks
140
Results Quality assessment
Results include preterm as well as term infants Level of evidence

4
Weight and length accounted for >97% of the variance of fat
free mass (g) in AGA and SGA infants and 46% of the variance Strengths & limitations
in LGA infants. Assessment of agreement with reference method –
Assessment of intra- and interobserver variability +
Midarm to head circumference ratio and midarm/thigh Repeated measurements +
muscle areas accounted for respectively 17% and 6% of the Coefficients of variance assessed –
variance in fat free mass (g). Sensitivity analysis –
Bootstrapping analysis –
Chest, abdomen, midarm and midthigh circumference did Cross validation group –
not explain any additional part of the variance in fat (free) External validation –
mass (g) after including gender, race, gestation, weight and Large study population –
length in the predictive equation. Exclusively preterm infants –
Mid-arm circumference was a covariate for fat mass Level of evidence

percentage and accounted for 60.4% of the variance. 2
Strengths & limitations

Assessment of agreement with reference method –
Assessment of intra- and interobserver variability –
Repeated measurements +
Coefficients of variance assessed –
Sensitivity analysis –
7
Cross validation group –
External validation –
Large study population –
Exclusively preterm infants –
Mid-arm circumference predicted 41% of the variance in mid- Level of evidence

arm circumference measured by MRI 2
Arm fat and muscle area predicted respectively 40% and 7 % Strengths & limitations
of the variance in the area measured by MRI Assessment of agreement with reference method –
Mid-arm circumference predicted 4.8% of the variance in Repeated measurements +
arm fat are (mm2) measured by MRI Coefficients of variance assessed +
Exclusively preterm infants +
51% of the variance in fat mass percentage could be Level of evidence

R explained by weight-length indices before 50 weeks PMA 4
and 16% after 50 weeks PMA
The best fit regression model most closely matched BMI Assessment of intra- and interobserver variability –
R Repeated measurements –
Large study population +
141
Chapter 7

assessment
Larcade, 2017, Weight (z-score) ADP 155 preterm 155 Hospital Bia
France Length (z-score) infants (GA 29.2 + 2.2 discharge (35 – we
Head circumference weeks) 41 weeks PMA)
(z-score) Lim
ag
Bia
we
reg
Lim
ag
Simon, 2014, Weight (z-score) ADP 141 preterms 141 Last week 82
France Length (z-score) (GA 31.5 + 2.5 before pr
Head circumference weeks) discharge. Mean we
(z-score) PMA 37.3 + 0.7
weeks 24
cli
Liotto, 2018, Weight ADP 1239 preterm 585 Term equivalent Re

Italy Length and term (GA 32.4 + 3.1 age or 3 days
Weight-for-length infants weeks) postnatal age Th
BMI inc
Ponderal index ma
po
Th
inc
ad
me
Ramel, 2017, Weight (z-score) ADP 218 preterm 218 1 – 3 days We

United States Length (z-score) infants (GA 33.8 + 1.9 postnatal age
of America Head circumference weeks) Bo
(z-score) pe
Weight-for-length
Ponderal index
BMI
142
Bias fat free mass -179 g (old predictive equation including Level of evidence
weight z-scores and clinical parameters) 2
Limits of agreement old equation + 249 g, i.e. poor Strengths & limitations
agreement (mean fat free mass study population 2500 g) Assessment of agreement with reference method
+
Bias fat free mass -12 g (new predictive equation, including Assessment of intra- and interobserver variability –
weight z-socres and taking into account new feeding Repeated measurements –
regimen) Coefficients of variance assessed +
Sensitivity analysis +
Limits of agreement new equation + 210 g, i.e. good Bootstrapping analysis +
agreement (mean fat free mass population 2500 g) Cross validation group –
82% of the variance in fat free mass (g) was explained by Level of evidence
predictive equation including clinical factors and change in 4
an weight z-score
24% of variance in fat mass percentage was explained by Assessment of agreement with reference method –
clinical factors and change in weight z-score Assessment of intra- and interobserver variability –
7
nt Results include preterm as well as term infants Level of evidence

4
The bias + limits of agreement of the predictive equation
including GA, gender and weight was + 6.1 g for fat free Strengths & limitations
mass adjusted by length (g/cm), i.e. poor agreement (study Assessment of agreement with reference method
population mean fat free mass adjusted for length 55 g/cm) +
Repeated measurements –
The bias + limits of agreement of the predictive equation Coefficients of variance assessed –
including GA, gender and weight was + 4.9 g for fat mass Sensitivity analysis –
adjusted by length, i.e. poor agreement (study population Bootstrapping analysis –
mean fat mass adjusted by length 7.4 g/cm) Cross validation group –
Weight explained 97% of the variance of fat free mass (g) Level of evidence
4
Body mass index explained 27% of the variance in fat mass
percentage and had a prediction error of 3.5% Strengths & limitations
143
Chapter 7

assessment
Villar, 2017, Weight-for-length ADP 1019 preterm 91 ADP within Re
United BMI and term (GA 36.0 + 0.7 96h of birth,
Kingdom Ponderal index infants weeks) anthropometry We
within 12h of 81%
birth
We
50
ADP: air displacement plethysmography; AGA: appropriate for gestational age; BMI: body mass index;
DXA: dual energy x-ray absorptiometry; GA: gestational age; LGA: large for gestational age; MRI: magnetic
resonance imaging; PMA: postmenstrual age; SFT: skinfold thickness; SGA: small for gestational age
Large study population > 100 participants
Levels of evidence based on the Oxford Centre for Evidence-based Medicine’s Levels of Evidence (15)
Level 1:
Validating cohort study with good reference standards*; or clinical decision rule tested within one clinical
center.
Level 2:
Exploratory cohort study with good reference standards; clinical decision rule after derivation, or validated
only on split-sample or databases
Level 3:
Non-consecutive study; or without consistently applied reference standards
Level 4:
Case-control study, poor or non-independent reference standard
Level 5:
Expert opinion without explicit critical appraisal, or based on physiology, bench research or “first principles”
144

4
y Weight-for-length, BMI and PI, respectively explained 92%,
81% and 62% of the variance of fat free mass (g). Strengths & limitations
Weight-for-length, BMI and PI, respectively explained 54%, Assessment of intra- and interobserver variability –
50% and 35% of the variance of fat mass (g). Repeated measurements –
145
Chapter 7
Table 4. Skinfold thickness measurements to estimate body composition in preterm infants

assessment
Koo, 2004, Tricep, subscapular, DXA 120 preterm and 52 3 (+ 2.8) days Re
United states of suprailiac, and (Hologic term infants (GA 30 - 36 postnatal age
America midthigh SFT on the QDR 1000/W weeks) W
left side densitometer, ad
pediatric software
version 5.64P)
Schmelzle, Triceps, biceps, DXA (Hologic QDR 104 preterm and 10 1 – 10 days Re
2002, Germany suprailiac, and 1500, infant whole- term infants (GA 34 – 36 postnatal age,
subscapular SFTs body software weeks) and at 2 and 4 Ne
version 5.67) months of age ex
Th
es
an
10
Lim
Pr
va
Daly-Wolfe, Triceps, biceps, ADP 56 preterm and 28 Discharge home Su

2015, suprailiac, and term infants (GA 33.3 + 0.9 an
United States of subscapular SFTs weeks) for
America
ADP: air displacement plethysmography; DXA: dual energy x-ray absorptiometry; GA: gestational age; SFT:
skinfold thickness
Level 1:
center.
Level 2:
Level 3:
146

4
When added to weight and length, SFT explained an
additional 13% of the variance in fat free mass (g) Strengths & limitations

2
Newly modeled predictive equation based on SFT
explained 94% of the variance in fat mass (g). Strengths & limitations
Assessment of agreement with reference method +
The mean error of the predictive equation for the Assessment of intra- and interobserver variability –
estimation of fat mass (g) was ± 75, ± 170, ± 300, Repeated measurements +
and ± 380 g for infants with an FM ≤ 500, 501–1000, Coefficients of variance assessed –
1001–2000, or > 2000 g, respectively Sensitivity analysis –
Bootstrapping analysis +
7
Limits of agreement not reported Cross validation group +
Previously modeled predictive equations could not be Large study population –
validated Exclusively preterm infants –
me Supra-iliac SFT was a covariate for fat mass percentage Level of evidence
and together with mid-arm circumference accounted 2
for 49% of the variance.
Level 4:
Level 5:
147
Chapter 7
Table 5. Bioelectrical impedance to estimate body composition in preterm infants

assessment
Dung, 2007, Impedance index DXA 118 preterm 118 38.6±3.8 weeks W
Germany derived from mono (Hologic QDR 1500, infants (GA 30.1 + 3.1 PMA th
frequency BIA infant software weeks)
version 5.67) Pr
m
Bia
-4
Lim
+2
po
Raghavan, 1998, Impedance index Isotope dilution 42 preterm 42 Day 1 – 7 W

United States of derived from mono infants (GA 24 – 34 postnatally ex
America frequency BIA weeks)
Bia
Lim
(m
Kushner, 1992, Impendance index Isotope dilution 133 preterm 32 Day 1 and day 4-7 Re
Unites States derived from infants, (GA 33 + 1.6 postnatally
of America and monofrequency BIA prepubertal weeks)
Peru children and
adults To
99
Tang, 1997, Impendance index Isotope dilution 28 preterm Unknown Day 1 -7 Re

United derived from and term GA median postnatally
Kingdom monofrequency BIA infants 30.5, range 24 W
– 38 weeks pr
th
ex
148
Weight and impedance index together explained 94% of Level of evidence

the variance in fat free mass (g) 2
Precision in prediction gained by impedance index was Strengths & limitations

max 3.03% Assessment of agreement with reference method +
Bias fat free mass: Repeated measurements –
-40 g (estimated with weight and impedance index) Coefficients of variance assessed –
Limits of agreement Bootstrapping analysis +
+ 20g, i.e. good agreement (mean fat free mass study External validation –
population 2260 g) Large study population +
Weight, length and the impedance index together Level of evidence

explained 90% of the variance in TBW. 4
Bias TBW 25.7 g (estimated with impedance alone) Strengths & limitations
Limits of agreement TBW + 95 g, i.e. poor agreement Assessment of intra- and interobserver variability –
(mean TBW study population 794 g) Repeated measurements +
Coefficients of variance assessed +
Cross validation group +
7
-7 Results include preterm infants, children and adults Level of evidence

2

Together with weight the impedance index explained Assessment of agreement with reference method –
99.5% of the variance in TBW Assessment of intra- and interobserver variability –

4
Weight explained 99.2 % of the variance in TBW. The
prediction model was significantly improved by adding Strengths & limitations
the impedance index to the model: 99.5% of the variance Assessment of agreement with reference method –
explained and a smaller 95% CI of 165 vs 200 ml. Assessment of intra- and interobserver variability –
149
Chapter 7

assessment
Collins, 2013, Bioelectrical Isotope dilution 99 preterm 99 Mean PMA 35 Bia
Australia impedance infants (GA 33 + 2 weeks Bia
spectroscopy weeks)
Lim
(m
Lim
ag
102
BIA: bioelectrical impedance; CI: confidence interval; DXA: dual energy x-ray absorptiometry; GA:
gestational age; PMA: postmenstrual age; SFT: skinfold thickness; TBW: total body water
Level 1:
center.
Level 2:
Level 3:
Level 4:
Level 5:
150
Bias TBW 10 ml Level of evidence

Bias extracellular water 40 ml 2
Limits of agreement TBW ±650 ml, i.e poor agreement Strengths & limitations
(mean TBW study population 1760 ml) Assessment of agreement with reference method +
Limits of agreement extracellular water ±360 ml, i.e. poor Repeated measurements +
agreement (mean extracellular water study population Coefficients of variance assessed –
1025 ml) Sensitivity analysis –
151
152
Table 6. Ultrasound to estimate body composition in preterm infants
Author, year, Measurement method Reference Study setting Results Quality assessment
country method
Chapter 7
Subjects N° Time of
preterms assessment
Nagel, 2020, Ultrasound images of ADP 63 preterm 63 PMA 35.1 + 1.2 Biceps adipose Level of evidence
United States the biceps (brachii and infants (GA 32.0 + weeks tissue and total 2
of America brachialis), abdomen 2.2 weeks) sum of adipose
(rectus abdominis), and tissue thickness Strengths & limitations
quadriceps (rectus femoris explained Assessment of agreement with
and vastus intermedius) respectively reference method –
using a portable B-mode 17% and 16% Assessment of intra- and
US device of fat mass interobserver variability +
percentage Repeated measurements +
Coefficients of variance
assessed +
Sensitivity analysis +
ADP: air displacement plethysmography; GA: gestational agePMA: postmenstrual age

Level 1:
Validating cohort study with good reference standards*; or clinical decision rule tested within one clinical center.
Level 2:
Exploratory cohort study with good reference standards; clinical decision rule after derivation, or validated only on split-sample or databases
Level 3:
Level 4:
Level 5:
Table 7. Magnetic Resonance Imaging to estimate body composition in preterm infants
Author, Measurement Reference Study setting Results Quality assessment

year, method method
country
Subjects N° Time of
preterms assessment
Dyke, Rapid whole No 25 preterm 15 Term Results include preterm as well as Level of evidence
2018, body imaging reference and term (GA 25.9 + (equivalent) term infants 2
United method for infants 1.3 weeks) age
States of fat (free) MRI calculated whole-body mass Strengths & limitations
America mass. correlated closely with measured Assessment of agreement
A weighing body weight (R2= 0.87;P < 0.001) with reference method –
scale used Assessment of intra- and
as reference Bias fat mass percentage 0.37% for interobserver variability –
for weight repeated measurements Repeated measurements +
Coefficients of variance
Limit of agreement + 1.3%, i.e. good assessed –
agreement (study population Sensitivity analysis –
mean fat mass percentage 25.5%) Bootstrapping analysis –
GA: gestational age; MRI: magnetic resonance imaging

Level 1:
Validating cohort study with good reference standards*; or clinical decision rule tested within one clinical center.
Level 2:
Exploratory cohort study with good reference standards; clinical decision rule after derivation, or validated only on split-sample or databases
Level 3:
Level 4:
Level 5:
153
7
Chapter 7
Table 8. Air displacement plethysmography to estimate body composition in preterm infants
Author, year, Measurement Reference Study setting Re

country method method
assessment
Forsum, 2016, ADP Isotope dilution 14 preterm 14 3 – 7 days postnatal Bia
Sweden infants (GA 34.1 + 1.3 age
weeks) Lim
(st
Bia
Lim
(st
Roggero, 2012, ADP Isotope dilution 79 preterm 70 Mean age 6 – 32 Pre

Italy and term (GA 24-36 weeks) days postnatal age Bia
infants AD
Lim
rep
ma
Ac
Bia
Lim
po
Int
pre
Bia
Lim
po
ADP: air displacement plethysmography; GA: gestational age

Level 1:
center.
Level 2:
Level 3:
Level 4:
Level 5:
154
al Bias fat mass percentage -1 % Level of evidence

2
Limits of agreement + 5.8 %, i.e. poor agreement
(study population mean fat mass percentage 3.2%) Strengths & limitations
Bias fat free mass density g/ml -0.002 g/ml Assessment of intra- and interobserver variability –
Limits of agreement + 0.012 g/ml, i.e. good agreement Coefficients of variance assessed –
(study population mean fat free mass density 1.06 g/ml) Sensitivity analysis –
Precision of ADP was assessed in 57 preterm infants Level of evidence

e Bias for fat mass percentage 0.15% between repeated 2
ADP measurements
Limits of agreement + 2.2%, i.e. poor agreement between Assessment of agreement with reference method +
repeated ADP measurements (study population mean fat Assessment of intra- and interobserver variability –
mass percentage 9.2%) Repeated measurements +
Accuracy was assessed in subgroup of 10 preterm infants Sensitivity analysis –
Bias fat mass percentage 0.32 % Bootstrapping analysis –
7
Limits of agreement + 3.1 %, i.e. poor agreement (study External validation –
population mean fat mass percentage 6.0 %) Large study population –
Interdevice reliability was assessed in a subgroup of 12
preterm and term infants
Bias fat mass percentage 0.42 %
Limits of agreement + 2.29 %, i.e. poor agreement (study

population mean fat mass percentage 8.8%)
155
Chapter 7
Discussion
Numerous studies have addressed how to measure body composition in preterm
infants.17-35 Indeed there is an urgent need to monitor body composition in this
population given the less favorable body composition in infancy and childhood
as well as the increased risk of adverse cardiometabolic outcomes in later life.1-7
Nevertheless, till date, there is no consensus on which method should preferentially
be used to assess body composition in preterm infants.
Reference methods
The studies included in this review used ADP, DXA, isotope dilution and MRI as
reference methods. In our opinion, ADP, DXA and isotope dilution are acceptable
reference methods.
ADP, DXA and isotope dilution have been validated against chemical carcass analysis
in piglets. 36-40 The body composition of piglets are considered to be comparable to
the body composition of preterm infants. 37 Therefore, in practice, these methods
are accepted as accurate measures. Nevertheless, we should take into consideration
that there is variation within and between these methods. For example different
types of software are used to analyze DXA. It has been reported that pediatric and
infant software rely on different assumptions and yield varying results.41 Furthermore,
as discussed below (under the accuracy of ADP), the statistical agreement between
ADP and isotope dilution may be interpreted as poor.31,32 Therefore, we believe that
some reservation is needed when comparing different reference methods.
In contrast to ADP, DXA and isotope dilution, MRI has not been validated against
chemical carcass analysis in subjects comparable to the neonatal population, but
it has been found accurate in adult human cadaver and animal studies.42,43 Over
the years MRI has been increasingly used to measure body composition in the
neonatal population as well.30,44-46 However, it is yet to be widely implemented and
we find it preliminary to use MRI as a reference method in comparative studies
with preterm infants.
In our opinion, due to the lack of a gold standard and the difference between
reference methods, some reservation is needed when drawing conclusions on the
validation studies included in this review.
Assessment of validity
To assess whether two methods agree, Bland-Altman analyses are an accepted
and widely implemented method. 47 Agreement should be based on a maximum
difference between the two methods that is clinically acceptable.47-50 Nevertheless,
this so-called predefined clinical agreement limit was omitted in all the studies
included in this review. Hence, the interpretation of these studies is limited.
156
Accuracy of body proportionality measures

Studies conducted so far, have mainly assessed the predictive value of body
proportionality measures.17,18,20,21 Two studies assessed the agreement between
ADP and a predictive equation including weight and clinical parameters.19,22 Liotto
and colleagues found poor agreement between ADP and their predictive equation
which estimated fat (free) mass adjusted by length (g/cm).19 However, they
included both preterm and term infants in their analysis, which makes it difficult
to extrapolate their findings to only preterm infants - the target population of
this review. Larcade et al.22 investigated a study population of exclusively preterm
infants and could not validate the predictive equation for fat free mass (g) made by
Simon and colleagues23. A difference in nutritional care and ensuing better growth
in Larcade’s population may have been the cause of an underestimation of fat free
mass by the previously modeled equation.
In our opinion, it is difficult to develop a predictive equation that can be validated

externally. Just as Larcade et al. found nutritional practices to influence the
prediction model, changes in neonatal care over time and across neonatal intensive
care units (NICUs) influence the predictive equations and limit their universal
application. Moreover, investigators have used mixed study populations which
include small, appropriate as well as large for gestational age infants.17 Meanwhile,
Koo et al. demonstrated that associations between weight/length indices and
body composition differ for those born large for gestational age, which makes the
use of a mixed study population inappropriate.17
7
In addition, it is important to note that predictive equations generally found that
a large proportion of the variance in fat (free) mass (g) could be explained by
weight or BMI. This logically follows the fact that fat mass (g) and fat free mass (g)
together make up total body weight. However, fat mass percentage was poorly
explained by weight or length indices. Meanwhile, in our opinion, fat (free) mass
percentage, may be a more relevant parameter when it comes to comparing the
body composition of an individual or groups because it takes the subject’s weight
into account.
All-in-all, we conclude that the predictive equations based on weight and length
indices currently cannot be implemented in clinical practice, because of the lack
of external validation and a poor predictive value for fat (free) mass percentage.
Daly-Wolfe and colleagues found that mid-arm circumference had a moderate

predictive value for fat mass percentage measured with ADP. 24 Koo et al., on the
other hand, found that midarm circumference together with chest, abdomen
and midthigh circumference, added less than 5 % to the variance in fat mass
percentage already explained by weight and length. In their study, however, fat
mass was determined by DXA. Pereira – Da Silva and colleagues compared upper
arm anthropometry to regional fat mass measured with MRI and found it to be an
157
Chapter 7
inaccurate predictor of regional body composition.25 In our view, currently there

is inconclusive evidence on the predictive value of body area circumferences and
more research is needed to assess the potential of mid-arm circumference as a
predictor of whole body or regional fat mass.
Accuracy of skinfold measurements

Several studies assess SFT in preterm infants. 24,51-56 For example, Daly-Wolfe and
colleagues investigated the predictive value of SFT and found that, together with the
mid-arm circumference, SFT explained 49% of the variance in fat mass percentage.
24
However, only one study included in the review assessed the validity of SFT in
preterm infants 26. Unfortunately, only 10 late preterm infants were included in this
study and analysis included term infants as well, so no robust conclusions can be
drawn from their findings.26 Recently, we found poor agreement between SFT and
body composition in a study with exclusively preterm infants.56
Moreover, SFT is influenced by the fluid status and there is a high inter-observer
variability.13 In addition, it could be deemed controversial to use SFT calipers
in extremely preterm infants in light of their vulnerable skin. Therefore, there is
insufficient evidence to support SFT as an clinically useful measure of fat mass in
preterm infants at this time.
Accuracy of bioelectrical impedance

Both BIA, as well as bioelectrical impedance spectroscopy, have a poor predictive
value for TBW measured with isotope dilution analyses 28,35,57 and body composition
measured with DXA 27. BIA did not seem to provide an additive predictive effect for
fat-free mass or TBW, compared to body weight alone.27,28 Though, both Kushner
and colleagues as well as Tang and colleagues found the impedance index (cm2/
Ω) to significantly improve the prediction of TBW, the majority of variance in
TBW measured with isotope dilution was still explained by weight. 33,57 Kushner
and colleagues concluded that the impedance index (cm2/ Ω) explained 99% of
the variance in TBW. However, in their subgroup of preterm infants there was a
significant bias: a higher variance was found for higher values of the impedance
index. This bias was only eliminated by the addition of weight to the prediction
equation 33 – implying that bioelectrical impedance on its own is not an adequate
predictor of body composition in preterm infants.
Accuracy of ultrasound
Ahmad et al. previously demonstrated that ultrasound measurements correlate with
fat mass in preterm infants.58 Depending on the site of ultrasound measurement
the intra-observer variability was reported to be up to 14.7% in preterm infants59. To
our knowledge the inter-observer variability has not been investigated in preterm
infants, but has been reported to show high inter-observer agreement (0.89-0.95)
158
in 1- and 2-year-olds.60 Nevertheless, we only found one study which assessed the
predictive value of ultrasound measurement for body composition in preterm
infants and they found a poor predictive value.29 Others did report on ultrasound
measurements as a means to estimate body composition in preterm infants, but
they did not assess the predictive value or validity.58,59 Hence, in our opinion, there
seems to be insufficient evidence for the use of ultrasound as a reference method
for body composition. However, since ultrasound measurements showed high
reliability, it may be of interest to investigate whether other body sites are a better
representation of body composition.
Accuracy of MRI
Though several authors suggested the use of MRI to measure body composition
in preterm infants44,61, only one study was found which assessed the validity of MRI
for the assessment of body composition in a small study population.30 Despite a
sound assessment of repeatability, the actual fat mass (g) measurement was not
compared with other techniques. Therefore, it should be concluded that more
studies are necessary to draw conclusions on the use of MRI in determining fat
mass in preterm infants.
Accuracy of ADP
Roggero and colleagues found a small bias when comparing ADP to isotope
dilution.32 However, the limits of agreement were relatively wide, resulting in poor
7
agreement between the two methods. Of note, the accuracy of ADP was only
assessed in a small subgroup of 10 preterm infants. In line with their findings
Forsum et al. also found a small bias when comparing ADP with deuterium dilution
in 14 preterm infants.31 Fat mass percentage, however, had relatively wide limits
of agreement and thus poor agreement. Fat free mass density (g/ml), however,
agreed well. Precision was studied by Roggero in a larger group of 57 preterm
infants and also showed a small bias for fat mass percentage between repeated
ADP measurements. 32 Despite wide limits of agreement the authors concluded
that ADP shows good agreement with isotope dilution for fat mass percentage
as well fat free mass density.31,32 Likewise, carcass analyses showed small bias, with
relatively wide limits of agreement.36 Nevertheless it is generally accepted that
ADP is a reliable method in infants. Taking into account the relatively small study
populations and small bias, it is to be expected that larger studies would yield
better agreement. Therefore, we conclude that ADP is a reliable method to assess
fat mass in preterm infants.
Accuracy of isotope dilution

Isotope dilution is a well-established method for the measurement of TBW
from which the fat free mass can be derived.62 As a result there were no studies
validating isotope dilution against another reference method, such as DXA, in
159
Chapter 7
preterm infants. Hartnoll and colleagues, however, did show similar results as late
nineteenth century cadaver studies, even though no comparative analysis was
done.63 A conclusion thus cannot be drawn based on human studies in preterm
infants, but is deemed reliable based on carcass analysis of piglets.37
Accuracy of DXA
There are no comparative studies with preterm infants where DXA was compared
with other methods, such as isotope dilution. DXA has been validated in piglets
38,40
and in practice is accepted as an accurate measure. Nevertheless, in human as
well as animal studies DXA has been reported to overestimate fat mass, especially
in lower weights. 40,64 Moreover, different software algorithms yield varying results
in body composition.41
It would be insightful to investigate the agreement between isotope dilution, ADP

and DXA in preterm infants. This would help to give us guidance on which method
should be preferentially used to assess body composition in preterm infants.
Patient-friendliness, ease of use and costs

Methods are chosen based on local experience and available resources. In a clinical
setting there is a preference for quick, easy-to-use, but accurate methods, which
could be used at the bedside. In contrast, in a research setting there is more room
for less flexible methods.
Body proportionality measures are quick, low-cost and minimally invasive methods.
They are ideal in the intensive care setting as well as in outpatient department for
follow-up. Unfortunately, studies so far have not confirmed these methods have
sufficient accuracy. Likewise, SFT and bioelectrical impedance techniques are easy
bedside methods, but robust evidence supporting their use is lacking. Moreover,
there is a high inter-observer variability and in extremely preterm infants SFT
measurements should be taken with caution to prevent injury of their vulnerable
skin. Nevertheless, when used with caution SFT is a safe, non-invasive method.10
DXA, ADP, and isotope dilution are accurate methods, but have some practical
downsides. For DXA infants need to be clinically stable and free from respiratory
support and monitoring, making DXA more appropriate from term age onwards.
Taking into account that movement is not allowed, it could be used in infants up
to 6 months corrected age who can be swaddled or nursed to sleep during the
procedure.
ADP could be used in the NICU if the infant is clinically stable and not on respiratory
support. Nonetheless, the machinery is bulky and recalibration is needed after
movement, making it less suitable for such use.
160
On the contrary, isotope dilution can be used in small infants who aren’t stable yet,
making it very suitable for use in the NICU - were it not that it has a relatively high
workload. In addition, oral administration of the isotope solution is challenging in
older infants.
MRI is yet to become a widely used method to measure fat mass in preterm infants,
but seems promising with high precision. Furthermore, MRI is a safe, radiation-
free method. Nevertheless, infants on respiratory support and monitoring cannot
easily undergo an MRI procedure. All-in-all, at this time there is no prospect of an
accurate, easy, low-cost point of care instrument which could be used during the
NICU stay or the follow-up period in a clinical setting. In a research setting ADP
may be the most practical, yet reliable, method to use in infants up to 6 months
corrected age.
Limitations
This review included all potential reference methods, which made it challenging
to come to a concise conclusion. Furthermore, it is important to note that, overall,
validation studies were conducted in a limited number of study subjects with a wide
range of gestational ages and varying postnatal ages at the time of assessment.
Hence, it was even more difficult to draw definitive conclusions on the assessment
of body composition at different gestational and postnatal ages. Moreover, the
reference methods used in the various studies have not been validated in humans,
or were only validated in a small number of subjects. This lack of a solid golden
7
standard further undermines any conclusions drawn from these studies. In
addition, only formally published data was included, leaving potential publication
bias unassessed. There is a need for larger cross-sectional studies comparing these
instruments at different time points as well as longitudinal studies investigating
the accuracy of the use of the instruments over time.
Conclusions
Monitoring body composition remains important in the light of the increased
cardiometabolic disease risk in adults born prematurely.1-7 Therefore the quest for
accurate but also practical methods to assess body composition should continue.
This review reaffirmed that weight and length indices, body area circumferences,
SFT, BIA and ultrasound do not adequately reflect body composition. MRI
looks promising for the use in preterm infants, but hasn’t been validated for
the measurement of body composition. On the other hand, DXA, ADP and
isotope dilution methods are considered trustworthy and validated techniques.
Nevertheless, this review showed that these methods may not yield comparable
results. Therefore, caution should be taken when comparing body composition
measured with different methods. Moreover, to facilitate future studies and support
161
Chapter 7
clinical practice it would be valuable for researchers and physicians to come to

an agreement on which reference should preferentially be used to measure body
composition in preterm infants.
162
References
1 Kerkhof, G. F., Breukhoven, P. E., Leunissen, R. W., Willemsen, R. H. & Hokken-Koelega, A. C.
Does Preterm Birth Influence Cardiovascular Risk in Early Adulthood? J Pediatr 161, 390-396
e391 (2012).
2 Sipola-Leppanen, M. et al. Cardiometabolic Risk Factors in Young Adults Who Were Born
Preterm. Am J Epidemiol 181, 861-873 (2015).
3 Morrison, K. M. et al. Cardiometabolic Health in Adults Born Premature with Extremely Low
Birth Weight. Pediatrics 138 (2016).
4 Johnson, M. J., Wootton, S. A., Leaf, A. A. & Jackson, A. A. Preterm Birth and Body Composition
at Term Equivalent Age: A Systematic Review and Meta-Analysis. Pediatrics 130, e640-e649
(2012).
5 Stigson, L. et al. Bone and Fat Mass in Relation to Postnatal Levels of Insulin-Like Growth
Factors in Prematurely Born Children at 4 Y of Age. Pediatric Research 75, 544-550 (2014).
6 Darlow, B. A., Martin, J. & Horwood, L. J. Metabolic Syndrome in Very Low Birth Weight Young
Adults and Controls: The New Zealand 1986 Vlbw Study. J Pediatr 206, 128-133 e125 (2019).
7 Crump, C. et al. Association of Preterm Birth with Risk of Ischemic Heart Disease in Adulthood.
JAMA Pediatr 173, 736-743 (2019).
8 Liu, P., Ma, F., Lou, H. & Liu, Y. The Utility of Fat Mass Index Vs. Body Mass Index and Percentage
of Body Fat in the Screening of Metabolic Syndrome. BMC Public Health 13, 629 (2013).
9 Ramirez-Velez, R. et al. Percentage of Body Fat and Fat Mass Index as a Screening Tool for
Metabolic Syndrome Prediction in Colombian University Students. Nutrients 9 (2017).
10 Strydom, K., Van Niekerk, E. & Dhansay, M. A. Factors Affecting Body Composition in Preterm
Infants: Assessment Techniques and Nutritional Interventions. Pediatr Neonatol 60, 121-128
11
(2019).
Mazahery, H., von Hurst, P. R., McKinlay, C. J. D., Cormack, B. E. & Conlon, C. A. Air Displacement
7
Plethysmography (Pea Pod) in Full-Term and Pre-Term Infants: A Comprehensive Review of
Accuracy, Reproducibility, and Practical Challenges. Matern Health Neonatol Perinatol 4, 12
(2018).
12 Andrews, E. T., Beattie, R. M. & Johnson, M. J. Measuring Body Composition in the Preterm
Infant: Evidence Base and Practicalities. Clin Nutr (2019).
13 Demerath, E. W. & Fields, D. A. Body Composition Assessment in the Infant. Am J Hum Biol 26,
291-304 (2014).
14 Moore, D. S., Notz, W. I. & Flinger, M. A. 138 (W. H. Freeman and Company, 2013).
15 Programme, C. A. S. Casp Checklist, <https://casp-uk.net/casp-tools-checklists/> (2018).
16 Philips B., B. C., Sackett D., Badenoch D., Straus S., Haynes B., Dawes M., Howick J. . Oxford
Centre for Evidence-Based Medicine: Levels of Evidence (March 2009), <https://www.cebm.
ox.ac.uk/resources/levels-of-evidence/oxford-centre-for-evidence-based-medicine-levels-of-
evidence-march-2009> (2009).
17 Koo, W. W., Walters, J. C. & Hockman, E. M. Body Composition in Neonates: Relationship
between Measured and Derived Anthropometry with Dual-Energy X-Ray Absorptiometry
Measurements. Pediatr Res 56, 694-700 (2004).
18 Kiger, J. R., Taylor, S. N., Wagner, C. L., Finch, C. & Katikaneni, L. Preterm Infant Body Composition
Cannot Be Accurately Determined by Weight and Length. J Neonatal Perinatal Med 9, 285-
290 (2016).
19 Liotto, N. et al. Can Basic Characteristics Estimate Body Composition in Early Infancy? Journal
of Pediatric Gastroenterology and Nutrition 66, e76-e80 (2018).
20 Ramel, S. E., Zhang, L., Misra, S., Anderson, C. G. & Demerath, E. W. Do Anthropometric Measures
Accurately Reflect Body Composition in Preterm Infants? Pediatric Obesity 12, 72-77 (2017).
163
Chapter 7
21 Villar, J. et al. Body Composition at Birth and Its Relationship with Neonatal Anthropometric
Ratios: The Newborn Body Composition Study of the Intergrowth-21 St Project. Pediatric
Research 82, 305-316 (2017).
22 Larcade, J. et al. Estimation of Fat-Free Mass at Discharge in Preterm Infants Fed with Optimized
Feeding Regimen. Journal of Pediatric Gastroenterology and Nutrition 64, 115-118 (2017).
23 Simon, L. et al. Determinants of Body Composition in Preterm Infants at the Time of Hospital
Discharge. American Journal of Clinical Nutrition 100, 98-104 (2014).
24 Daly-Wolfe, K. M., Jordan, K. C., Slater, H., Beachy, J. C. & Moyer-Mileur, L. J. Mid-Arm Circumference
Is a Reliable Method to Estimate Adiposity in Preterm and Term Infants. Pediatr Res 78, 336-341
(2015).
25 Pereira-Da-Silva, L., Abecasis, F., Virella, D. & Videira-Amaral, J. M. Upper Arm Anthropometry Is
Not a Valid Predictor of Regional Body Composition in Preterm Infants. Neonatology 95, 74-79
(2009).
26 Schmelzle, H. R. & Fusch, C. Body Fat in Neonates and Young Infants: Validation of Skinfold
Thickness Versus Dual-Energy X-Ray Absorptiometry. American Journal of Clinical Nutrition
76, 1096-1100 (2002).
27 Dung, N. Q., Fusch, G., Armbrust, S., Jochum, F. & Fusch, C. Body Composition of Preterm Infants
Measured During the First Months of Life: Bioelectrical Impedance Provides Insignificant
Additional Information Compared to Anthropometry Alone. Eur J Pediatr 166, 215-222 (2007).
28 Raghavan, C. V. et al. Estimation of Total Body Water in Very-Low-Birth-Weight Infants by Using
Anthropometry with and without Bioelectrical Impedance and H2[18o]. American Journal of
Clinical Nutrition 68, 668-674 (1998).
29 Nagel, E. et al. Can Ultrasound Measures of Muscle and Adipose Tissue Thickness Predict Body
Composition of Premature Infants in the Neonatal Intensive Care Unit? Journal of Parenteral
and Enteral Nutrition (2020).
30 Dyke, J. P., Garfinkel, A. C., Groves, A. M. & Kovanlikaya, A. High-Resolution Rapid Neonatal
Whole-Body Composition Using 3.0 Tesla Chemical Shift Magnetic Resonance Imaging.
Pediatr Res 83, 638-644 (2018).
31 Forsum, E., Olhager, E. & Tornqvist, C. An Evaluation of the Pea Pod System for Assessing Body
Composition of Moderately Premature Infants. Nutrients 8, 238 (2016).
32 Roggero, P. et al. Evaluation of Air-Displacement Plethysmography for Body Composition
Assessment in Preterm Infants. Pediatric Research 72, 316-320 (2012).
33 Kushner, R. F., Schoeller, D. A., Fjeld, C. R. & Danford, L. Is the Impedance Index (Ht2/R) Significant
in Predicting Total Body Water? American Journal of Clinical Nutrition 56, 835-839 (1992).
34 Tang, W., Ridout, D. & Modi, N. Influence of Respiratory Distress Syndrome on Body Composition
after Preterm Birth. Archives of Disease in Childhood: Fetal and Neonatal Edition 77, F28-F31
(1997).
35 Collins, C. T. et al. Prediction of Body Water Compartments in Preterm Infants by Bioelectrical
Impedance Spectroscopy. Eur J Clin Nutr 67 Suppl 1, S47-53 (2013).
36 Frondas-Chauty, A., Louveau, I., Le Huerou-Luron, I., Roze, J. C. & Darmaun, D. Air-Displacement
Plethysmography for Determining Body Composition in Neonates: Validation Using Live
Piglets. Pediatr Res 72, 26-31 (2012).
37 Rudolph, B. C., Stahly, T. S. & Cromwell, G. L. Estimation of Body Composition of Neonatal Pigs
Via Deuterium Oxide Dilution: Validation of Technique. J Anim Sci 66, 53-61 (1988).
38 Brunton, J. A., Bayley, H. S. & Atkinson, S. A. Validation and Application of Dual-Energy X-Ray
Absorptiometry to Measure Bone Mass and Body Composition in Small Infants. Am J Clin Nutr
58, 839-845 (1993).
39 Picaud, J. C., Rigo, J., Nyamugabo, K., Milet, J. & Senterre, J. Evaluation of Dual-Energy X-Ray
Absorptiometry for Body-Composition Assessment in Piglets and Term Human Neonates. Am
J Clin Nutr 63, 157-163 (1996).
164
40 Rigo, J. et al. Reference Values of Body Composition Obtained by Dual Energy X-Ray
Absorptiometry in Preterm and Term Neonates. Journal of Pediatric Gastroenterology and
Nutrition 27, 184-190 (1998).
41 Picaud, J. C. et al. Software and Scan Acquisition Technique-Related Discrepancies in Bone
Mineral Assessment Using Dual-Energy X-Ray Absorptiometry in Neonates. Acta Paediatr 91,
1189-1193 (2002).
42 Fowler, P. A., Fuller, M. F., Glasbey, C. A., Cameron, G. G. & Foster, M. A. Validation of the in Vivo
Measurement of Adipose Tissue by Magnetic Resonance Imaging of Lean and Obese Pigs. Am
J Clin Nutr 56, 7-13 (1992).
43 Abate, N., Burns, D., Peshock, R. M., Garg, A. & Grundy, S. M. Estimation of Adipose Tissue Mass
by Magnetic Resonance Imaging: Validation against Dissection in Human Cadavers. J Lipid
Res 35, 1490-1496 (1994).
44 Uthaya, S., Bell, J. & Modi, N. Adipose Tissue Magnetic Resonance Imaging in the Newborn.
Horm Res 62 Suppl 3, 143-148 (2004).
45 Harrington, T. A. et al. Fast and Reproducible Method for the Direct Quantitation of Adipose
Tissue in Newborn Infants. Lipids 37, 95-100 (2002).
46 Stokes, T. A. et al. The Clinical Utility of Anthropometric Measures to Assess Adiposity in a Cohort
of Prematurely Born Infants: Correlations with Mri Fat Quantification. J Neonatal Perinatal
Med 10, 133-138 (2017).
47 Lu, M. J. et al. Sample Size for Assessing Agreement between Two Methods of Measurement by
Bland-Altman Method. Int J Biostat 12 (2016).
48 Bland, J. M. & Altman, D. G. Comparing Methods of Measurement: Why Plotting Difference
against Standard Method Is Misleading. Lancet 346, 1085-1087 (1995).
49 Bland, J. M. & Altman, D. G. Statistical Methods for Assessing Agreement between Two Methods
50
of Clinical Measurement. Lancet 1, 307-310 (1986).
Giavarina, D. Understanding Bland Altman Analysis. Biochem Med (Zagreb) 25, 141-151 (2015).
7
51 Bustamante, S. A., Jacobs, P. & Gaines, J. A. Body Weight, Static and Dynamic Skinfold Thickness
in Small Premature Infants During the First Month of Life. Early human development 8, 217-
224 (1983).
52 Dauncey, M. J., Gandy, G. & Gairdner, D. Assessment of Total Body Fat in Infancy from Skinfold
Thickness Measurements. Arch Dis Child 52, 223-227 (1977).
53 Olhager, E. & Forsum, E. Assessment of Total Body Fat Using the Skinfold Technique in Full-
Term and Preterm Infants. Acta Paediatrica, International Journal of Paediatrics 95, 21-28
(2006).
54 Petersen, S., Gotfredsen, A. & Ursin Knudsen, F. Lean Body Mass in Small for Gestational Age
and Appropriate for Gestational Age Infants. Journal of Pediatrics 113, 886-889 (1988).
55 Sheng, H. P., Muthappa, P. B., Wong, W. W. & Schanler, R. J. Pitfalls of Body Fat Assessments in
Premature Infants by Anthropometry. Biol Neonate 64, 279-286 (1993).
56 Yumani, D. F. J., de Jongh, D., Lafeber, H. N. & van Weissenbruch, M. M. A Comparative Study
Using Dual-Energy X-Ray Absorptiometry, Air Displacement Plethysmography, and Skinfolds
to Assess Fat Mass in Preterms at Term Equivalent Age. Eur J Pediatr 180, 919-927 (2021).
57 Tang, W., Ridout, D. & Modi, N. Assessment of Total Body Water Using Bioelectrical Impedance
Analysis in Neonates Receiving Intensive Care. Archives of disease in childhood Fetal and
neonatal edition 77, F123-126 (1997).
58 Ahmad, I. et al. Body Composition and Its Components in Preterm and Term Newborns: A
Cross-Sectional, Multimodal Investigation. American journal of human biology : the official
journal of the Human Biology Council 22, 69-75 (2010).
59 McLeod, G. et al. Feasibility of Using Ultrasound to Measure Preterm Body Composition and
to Assess Macronutrient Influences on Tissue Accretion Rates. Early human development 89,
577-582 (2013).
165
Chapter 7
60 Holzhauer, S. et al. Sonographic Assessment of Abdominal Fat Distribution in Infancy. Eur J

Epidemiol 24, 521-529 (2009).
61 Ward, L. C., Poston, L., Godfrey, K. M. & Koletzko, B. Assessing Early Growth and Adiposity:
Report from an Earlynutrition Academy Workshop. Ann Nutr Metab 63, 120-130 (2013).
62 Agency, I. A. E. Body Composition Assessment from Birth to Two Years of Age. (Vienna, 2013).
63 Hartnoll, G., Bétrémieux, P. & Modi, N. Body Water Content of Extremely Preterm Infants at
Birth. Archives of Disease in Childhood: Fetal and Neonatal Edition 83, F56-F59 (2000).
64 Brunton, J. A., Weiler, H. A. & Atkinson, S. A. Improvement in the Accuracy of Dual Energy X-Ray
Absorptiometry for Whole Body and Regional Analysis of Body Composition: Validation Using
Piglets and Methodologic Considerations in Infants. Pediatric Research 41, 590-596 (1997).
166
167
DISCUSSION &
SUMMARY
8
CHAPTER 8
General discussion
Chapter 8
The “Developmental Origins of Health and Disease” hypothesis states that exposure
to so called insults in critical periods of development influences later health
outcomes. (1) For preterm infants the extra-uterine third trimester is such a critical
period and postnatal growth restriction is an insult which has been associated with
potentially severe adverse health outcomes. Limiting postnatal growth restriction
could therefore be a key factor in improving long-term outcomes of preterm
infants.
Nutrition in relation to the endocrine regulation of

preterm growth and body composition
It is known that nutrition potentiates growth and has a direct effect on the
endocrine axis.(2) In preterm infants, however, there seems to be a certain time
frame in which nutrition can exert a significant impact on the endocrine axis. In
line with previous studies, our findings as described in chapter 4 suggest that
the influence of nutrition on the endocrine axis increases, in particular IGF-I, with
increasing postmenstrual age. (3) This could imply that less nutrient intake is
required to potentiate an effect on the endocrine axis and subsequently on growth
at higher postmenstrual ages. Hypothesizing, this could be due to a rise in growth
factors which exert more influence on growth as their levels rise with increasing
postmenstrual age. Indeed, initially serum IGF-I levels in preterm infants are low
and start showing an increase between 30 and 32 weeks postmenstrual age. (2)
As described in chapter 5 in our study population as well as other extremely and
very preterm study cohorts (4), an increase in growth SD scores is seen from 30 to
32 weeks postmenstrual age. This gain in growth SD scores coincides with the rise
in IGF-I levels. Therefore, it could be speculated that at this point IGF-I levels pass a
certain threshold and are now able to direct growth. Hypothesizing from this point
onwards nutrient enrichment could be decreased. Since IGF-I seems to potentiate
growth from this point an excess of nutrients could lead to fat deposition and an
unfavourable body composition. Or it could cause rapid growth which has been
associated with adverse cardiometabolic outcomes in later life.(5)
Originally the NUTRIE study was designed as a randomized controlled trial

comparing preterm infants fed limited energy- and nutrient-enriched formula
from 32 weeks postmenstrual age onwards with preterm infants fed standard fully
enriched preterm starters formula.(6) This would have enabled us to investigate
whether in the light of growth, body composition and neurodevelopment indeed
less nutrient enrichment is required once IGF-I passes the threshold concentration.
Due to the establishment of a human milk bank in our centre, there was a large
increase in human milk use. Therefore it was no longer possible to complete the
original trial design. Instead, we carried out an observational study. This resulted
in a detailed collection of data on nutritional intake and growth which enabled
a good exploration of potential associations with health outcomes. However,
to reduce total blood sampling volumes, we sampled IGF-I at a low frequency
172
General discussion
(every other week). This led to relatively small sample sizes and thus limited
statistical power. With techniques which use smaller sample volumes, such as
IGF-I measurements in dried blood spots(7), future studies could increase IGF-I
sampling frequency and increase the statistical power. In the future sufficiently
powered interventional studies are required to investigate whether growth and
body composition can be optimized further with nutritional interventions. In light
of the growing availability of donor human milk and exclusive human milk diets it
would be valuable to assess whether human milk fortification should be adapted to
IGF-I levels. Currently, in most practices human milk for preterm infants is fortified
with a standard amount of protein. However, Rochow and colleagues showed
that analysing human milk samples and individually adjusting the macronutrient
content to meet recommended intakes, improves growth.(8) On the contrary, in
a study by Agakidou et al. this so-called protein-targeting fortification of human
milk, compared to fixed-fortification, led to lower daily protein and energy intake.
In their study population these lower macronutrient intakes were associated with
transiently lower IGF-I levels. (9) It would be insightful to investigate whether a
more sustainable effect on IGF-I and growth could be obtained if IGF-I levels were
considered in determining the amount of human milk fortification. Speculatively,
protein fortification could be reduced, once IGF-I levels increase and pass the
theoretical threshold value of IGF-I between 30 and 32 weeks postmenstrual age.
Moreover, the effect of donor human milk on growth, body composition and
comorbidities needs to be explored further. An exclusive human milk diet is
associated with improved health outcomes in preterm infants.(10) Studies show
that, compared to preterm formula, donor human milk reduces the risk of
necrotizing enterocolitis.(11) Nevertheless, several controlled trials have failed
to show that supplementation of donor human milk compared to preterm
8
formula improves short term outcomes or mortality. (11-13) During hospitalization
predominant donor human milk feeding has been associated with less weight
gain and linear growth compared to preterm formula feeding (11, 14, 15) and also
compared to own mother’s milk feeding (15). This is in line with our findings
described in chapter 3, suggesting donor human milk use is associated with lower
IGF-I levels. Speculatively, this could be associated with less catch-up growth and
improved cardiometabolic outcomes, but poorer neurodevelopmental outcomes.
However, on the long term no differences in growth have been reported (11, 16) and
5 year olds born very low birth weight showed comparable body composition when
donor human milk feeding was compared to formula feeding. (16) Furthermore,
despite lower protein content in donor human milk, donor human milk has not
been shown to worsen neurodevelopmental outcome when compared to formula
feeding. (11) Nevertheless, while own mother’s milk improves neurodevelopmental
outcome when compared to formula feeding, donor human milk has not been
shown to have that same beneficial effect when compared to formula feeding.
(17) It would be valuable to investigate how donor human milk could be optimized
further to retain the beneficial effects of own mother’s milk.
173
Chapter 8
Pasteurization may play a role in the effect of donor human milk on health outcomes.
Essential factors such as IGF-I and micronutrients are significantly reduced during
pasteurization. (18) In a meta-analysis Villamore et al. found that pasteurization
increased the risk of bronchopulmonary dysplasia. This would be in line with
the negative association between donor human milk use and the occurrence of
bronchopulmonary dysplasia we described in our study population in chapter 3.
However, in our study population the proportion of donor human milk used was
relatively small and infants were fed own mother’s milk for the majority. Therefore
we cannot draw definitive conclusions from our findings. In fact, in another meta-
analysis by Miller and colleagues the use of unpasteurised human milk was not
found to reduce morbidity when compared to pasteurised human milk. (19) Based
on these inconclusive results further investigation of the pasteurization process of
donor human milk may be warranted, as optimizing of this process may increase
the beneficial effect of donor human milk.
The developing endocrine axis in relation to

comorbidities in preterm infants
Major comorbidities in preterm infants, such as sepsis, necrotizing enterocolitis,
retinopathy of prematurity and bronchopulmonary dysplasia, are associated
with low IGF-I levels.(20) Though causality could not be confirmed due to the
observational design of our study, indeed in our study population the odds of BPD
decreased with higher IGF-I levels. (Chapter 3) Moreover, these major comorbidities
are associated with a pro-inflammatory state. Indeed, higher levels of pro-
inflammatory proteins such as IL-6 and CRP, have been associated with lower IGF-I
levels in preterm infants. (21, 22) Speculating, this suggests that either low IGF-I
elicits a pro-inflammatory state or a pro-inflammatory state suppresses IGF-I. In
vitro studies showed that IGF-I can prevent apoptosis induced by TNF-alpha by
triggering the signalling pathway that activates the XIAP protein which inhibits
apoptosis. (23) Moreover, in mice, the administration of IGF-I before the inoculation
with Pseudomonas aeruginosa, improved bacterial clearance and reduced death
from severe sepsis. (23) These findings suggest that a lack of IGF-I results in less
inhibition of inflammation and thus elicits a pro-inflammatory state. On the other
hand, it could very well be that a pro-inflammatory state suppresses IGF-I. In the
case of inflammation energy stores may be redirected to ward off pathogens while
anabolic processes involving IGF-I, are suppressed. In line with that it has been
hypothesized in previous studies that parenteral nutrition leads to lower IGF-I levels
through the enhancement of an inflammatory state. (24) Nevertheless, certain
microRNA’s, small non-coding RNA’s which regulate gene expression, have been
found to be associated with the occurrence of inflammatory conditions in children
and the down-regulation of the IGF-I receptor. (25) Hypothesizing this could imply
that there is a common factor causing both inflammation and low IGF-I levels and
subsequent impaired growth.
174
General discussion
Even so, Leviton and colleagues reported positive associations between certain
pro-inflammatory proteins and IGF-I. For example, vascular cell adhesion molecule
1, which mediates the adhesion of leukocytes to vascular endothelium, increases
as IGF-I levels increase. Also, matrix metallopeptidase 9, a regulatory factor in
neutrophil migration, is positively associated with IGF-I. (22) Hypothesizing IGF-I
may have pro-inflammatory as well as anti-inflammatory properties, which has also
been implied by older studies. (26) This warrants for further research. Especially, in
the light of recent studies on the administration of IGF-I in preterm infants which
failed to show a reduction in the occurrence of retinopathy of prematurity, but did
find a significant reduction in the occurrence of bronchopulmonary dysplasia. (27)
Speculatively, the balance between the pro- and anti-inflammatory properties of
IGF-I may have played a role in this.
Determinants and assessment of body composition in

preterm infants
Body composition in infancy and childhood is influenced by early life events. For
example, disease severity after preterm birth is negatively associated with fat
free mass in infancy. (28) Also nutrition and growth patterns in early life influence
body composition. Both increased energy and protein intakes between preterm
birth and hospital discharge have been shown to be associated with higher
fat free mass at hospital discharge. (29) In line with these findings weight gain
between preterm birth and term equivalent age has shown positive correlations
with fat free mass at term equivalent age.(30, 31) After preterm birth, however, it
is a challenge to administer the recommended amount of nutrients, resulting in
relative malnutrition in the first week of life. Moreover, preterm birth is associated 8
with potentially severe comorbidities, especially with decreasing gestational age at
birth. Therefore unsurprisingly, postnatal growth restriction is often seen in preterm
infants.(32) In line with the previously mentioned association between growth and
lean mass, in the case of preterm infants who frequently suffer postnatal growth
restriction, a relative increase in fat mass is expected to be seen. Indeed, increased
fat mass at term equivalent age has been described when comparing preterm
infants with those born at term and when comparing extremely preterm to very
preterm infants. (33) Thus, body composition is a reflection of the condition of the
infant in early life. It would be insightful to investigate the influence of optimizing
IGF-I levels, either through dietary interventions or administration of IGF-I in
the first weeks of life, on later body composition. To the best of our knowledge
there are no reports on the effect of IGF-I administration on body composition
in preterm infants or animal models resembling preterm infants. Interestingly in
children with growth hormone insensitivity, higher doses of IGF-I administration,
led to an increase in fat mass percentage. (34) Hypothesizing, there may be a dose-
response relationship in preterm infants as well, where a moderate IGF-I levels
would stimulate lean mass, while higher levels would increase adiposity.
175
Chapter 8
Besides, a higher fat mass percentage and a higher fat mass index have been
associated with an increased occurrence of metabolic syndrome components
(35, 36). Therefore, monitoring body composition in early life would be a valuable
asset to help to implement timely measures to prevent adverse outcomes on
the long term. As discussed in chapter 6 and 7, diverse methods are available to
estimate or measure body composition in preterm infants. Nevertheless, there is
no consensus on which reference method should preferentially be used to assess
body composition in infants. Meanwhile, the few comparative studies conducted
till date show poor agreement. (37) . As reported in chapter 6 the measurement
of skinfolds and anthropometry such as waist circumference, is easy to conduct
without expensive equipment, but unfortunately estimates of fat mass based on
these methods show poor agreement with air displacement plethysmography and
dual-energy x-ray absorptiometry. Nevertheless our study had a small sample size
and did not compare the different methods before term corrected age. Therefore
at this point it remains to be concluded whether there should remain a place for
skinfold measurements and alternative anthropometry during hospitalisation
before term age. Furthermore, as described in our systematic review in chapter 7,
overall validation studies have been conducted in small study cohorts and there
is no method that qualifies as the gold standard in living infants. In conclusion, at
this time, there is inconclusive evidence to advise one particular method for the
assessment for body composition in preterm infants. In clinically stable subjects
we would suggest using air displacement plethysmography which is minimally
invasive, has a moderate workload and yields accurate results. Nevertheless,
currently there is no air displacement devices available for infants weighing
between 8 and 10 kg, which complicates long-term follow up. Alternatively, isotope
dilution is an accurate method which could be used in infants who aren’t clinically
stable and is not restricted to certain weight limits. However, it has a high workload.
Currently it seems there are no easy-to-use, low-cost, bedside techniques available
that yield accurate results. All-in-all further research is warranted, which should
investigate instruments at different time points and consensus should be reach on
which method should be used as a reference.
Future research directions

This thesis touched on the influence of nutrition, growth and the developing
endocrine axis on outcomes of preterm infants in early infancy. Additional outcomes
of the NUTRIE study, such as neurodevelopment and risk factors for the metabolic
syndrome, e.g. blood pressure and lipid profile, are yet to be reported. This will
shed more light on the complex interplay between IGF-I and the development of
preterm infants. Nevertheless, it can be concluded that adequate levels of IGF-I are
required to ensure sufficient growth, a favourable body composition and to prevent
major comorbidities. Using intra-uterine IGF-I levels as a reference, interventions
are needed to increase IGF-I levels after preterm birth. This should preferably be
done through non-invasive methods. Nevertheless, to date non-invasive methods,
176
General discussion
such as enteral IGF-I supplementation, have failed to yield the expected results
in preterm infants.(38) Future research should aim to explore how more IGF-I
bioactivity can be retained when IGF-I is supplemented enterally. Furthermore, it
would be of interest to investigate whether adapting nutrient intake to IGF-I levels
improves health outcomes. At the theoretical threshold value of IGF-I between
30 and 32 weeks postmenstrual age, human milk fortification could be limited
to prevent rapid growth and possible associated adverse events. On the other
hand, a phase 2 trial showed that continuous intravenous supplementation of
IGF-I reduces the occurrence of major comorbidities such as BPD. (27) With these
promising results, it is important for future research to aim to investigate the effect
of IGF-I administration on both growth and body composition, in addition to the
occurrence of morbidities.
Lastly, cross-sectional longitudinal studies are warranted to investigate the best

method for the follow-up of body composition in preterm infants.
177
Chapter 8
References
1. Hoffman DJ, Reynolds RM, Hardy DB. Developmental origins of health and disease: current
knowledge and potential mechanisms. Nutr Rev. 2017;75(12):951-70.
2015;77(1-2):156-63.
in Extremely and Very Preterm Infants During Hospitalisation. Nutrients. 2020;12(3).
5. Euser AM, Finken MJ, Keijzer-Veen MG, Hille ET, Wit JM, Dekker FW. Associations between
prenatal and infancy weight gain and BMI, fat mass, and fat distribution in young adulthood:
a prospective cohort study in males and females born very preterm. Am J Clin Nutr.
2005;81(2):480-7.
6. Yumani DF, Lafeber HN, Van Weissenbruch M. NUTRIE Study trial registration 2015 [Available
from: https://www.trialregister.nl/trial/5171.
7. Cox HD, Rampton J, Eichner D. Quantification of insulin-like growth factor-1 in dried blood
spots for detection of growth hormone abuse in sport. Anal Bioanal Chem. 2013;405(6):1949-58.
8. Rochow N, Fusch G, Ali A, Bhatia A, So HY, Iskander R, et al. Individualized target fortification of
breast milk with protein, carbohydrates, and fat for preterm infants: A double-blind randomized
controlled trial. Clin Nutr. 2021;40(1):54-63.
9. Agakidou E, Karagiozoglou-Lampoudi T, Parlapani E, Fletouris DJ, Sarafidis K, Tzimouli V, et al.
Modifications of Own Mothers’ Milk Fortification Protocol Affect Early Plasma IGF-I and Ghrelin
Levels in Preterm Infants. A Randomized Clinical Trial. Nutrients. 2019;11(12).
10. Taylor SN. Solely human milk diets for preterm infants. Semin Perinatol. 2019;43(7):151158.
11. Quigley M, Embleton ND, McGuire W. Formula versus donor breast milk for feeding preterm or
low birth weight infants. Cochrane Database Syst Rev. 2018;6:CD002971.
12. Corpeleijn WE, de Waard M, Christmann V, van Goudoever JB, Jansen-van der Weide MC, Kooi
EM, et al. Effect of Donor Milk on Severe Infections and Mortality in Very Low-Birth-Weight
Infants: The Early Nutrition Study Randomized Clinical Trial. JAMA Pediatr. 2016;170(7):654-61.
13. Villamor-Martinez E, Pierro M, Cavallaro G, Mosca F, Kramer BW, Villamor E. Donor Human
Milk Protects against Bronchopulmonary Dysplasia: A Systematic Review and Meta-Analysis.
Nutrients. 2018;10(2).
14. Madore LS, Bora S, Erdei C, Jumani T, Dengos AR, Sen S. Effects of Donor Breastmilk Feeding
on Growth and Early Neurodevelopmental Outcomes in Preterm Infants: An Observational
Study. Clin Ther. 2017;39(6):1210-20.
15. Brownell EA, Matson AP, Smith KC, Moore JE, Esposito PA, Lussier MM, et al. Dose-response
Relationship Between Donor Human Milk, Mother’s Own Milk, Preterm Formula, and Neonatal
Growth Outcomes. J Pediatr Gastroenterol Nutr. 2018;67(1):90-6.
16. McGee M, Unger S, Hamilton J, Birken CS, Pausova Z, Kiss A, et al. Adiposity and Fat-Free Mass
of Children Born with Very Low Birth Weight Do Not Differ in Children Fed Supplemental
Donor Milk Compared with Those Fed Preterm Formula. J Nutr. 2020;150(2):331-9.
17. O’Connor DL, Gibbins S, Kiss A, Bando N, Brennan-Donnan J, Ng E, et al. Effect of Supplemental
Donor Human Milk Compared With Preterm Formula on Neurodevelopment of Very Low-
Birth-Weight Infants at 18 Months: A Randomized Clinical Trial. JAMA. 2016;316(18):1897-905.
18. Peila C, Moro GE, Bertino E, Cavallarin L, Giribaldi M, Giuliani F, et al. The Effect of Holder
Pasteurization on Nutrients and Biologically-Active Components in Donor Human Milk: A
Review. Nutrients. 2016;8(8).
178
General discussion
19. Miller J, Tonkin E, Damarell RA, McPhee AJ, Suganuma M, Suganuma H, et al. A Systematic
Review and Meta-Analysis of Human Milk Feeding and Morbidity in Very Low Birth Weight
Infants. Nutrients. 2018;10(6).
20. Hellstrom A, Ley D, Hansen-Pupp I, Hallberg B, Lofqvist C, van Marter L, et al. Insulin-like growth
factor 1 has multisystem effects on foetal and preterm infant development. Acta Paediatr.
2016;105(6):576-86.
21. Hellgren G, Lofqvist C, Hansen-Pupp I, Gram M, Smith LE, Ley D, et al. Increased postnatal
concentrations of pro-inflammatory cytokines are associated with reduced IGF-I levels and
retinopathy of prematurity. Growth Horm IGF Res. 2018;39:19-24.
22. Leviton A, Allred EN, Fichorova RN, VanderVeen DK, O’Shea TM, Kuban K, et al. Early Postnatal
IGF-1 and IGFBP-1 Blood Levels in Extremely Preterm Infants: Relationships with Indicators of
Placental Insufficiency and with Systemic Inflammation. American journal of perinatology.
2019;36(14):1442-52.
23. Ashare A, Nymon AB, Doerschug KC, Morrison JM, Monick MM, Hunninghake GW. Insulin-like
growth factor-1 improves survival in sepsis via enhanced hepatic bacterial clearance. Am J
Respir Crit Care Med. 2008;178(2):149-57.
25. Cirillo F, Lazzeroni P, Catellani C, Sartori C, Amarri S, Street ME. MicroRNAs link chronic
inflammation in childhood to growth impairment and insulin-resistance. Cytokine Growth
Factor Rev. 2018;39:1-18.
26. Clark R. The Somatogenic Hormones and Insulin-Like Growth Factor-1: Stimulators of
Lymphopoiesis and Immune Function. Endocrine Reviews. 1997;18(2):157-79.
27. Ley D, Hallberg B, Hansen-Pupp I, Dani C, Ramenghi LA, Marlow N, et al. rhIGF-1/rhIGFBP-3 in
Preterm Infants: A Phase 2 Randomized Controlled Trial. J Pediatr. 2019;206:56-65 e8.
28. Ramel SE, Gray HL, Ode KL, Younge N, Georgieff MK, Demerath EW. Body composition
changes in preterm infants following hospital discharge: comparison with term infants. J
Pediatr Gastroenterol Nutr. 2011;53(3):333-8.
29. Ramel SE, Haapala J, Super J, Boys C, Demerath EW. Nutrition, Illness and Body Composition
in Very Low Birth Weight Preterm Infants: Implications for Nutritional Management and 8
Neurocognitive Outcomes. Nutrients. 2020;12(1).
30. Tremblay G, Boudreau C, Belanger S, St-Onge O, Pronovost E, Simonyan D, et al. Body
Composition in Very Preterm Infants: Role of Neonatal Characteristics and Nutrition in
Achieving Growth Similar to Term Infants. Neonatology. 2017;111(3):214-21.
31. Simon L, Frondas-Chauty A, Senterre T, Flamant C, Darmaun D, Rozé JC. Determinants of body
composition in preterm infants at the time of hospital discharge. American Journal of Clinical
Nutrition. 2014;100(1):98-104.
33. Bruckner M, Khan Z, Binder C, Morris N, Windisch B, Holasek S, et al. Extremely Preterm Infants
Have a Higher Fat Mass Percentage in Comparison to Very Preterm Infants at Term-Equivalent
Age. Front Pediatr. 2020;8:61.
34. Guevara-Aguirre J, Rosenbloom AL, Guevara-Aguirre M, Saavedra J, Procel P. Recommended
IGF-I dosage causes greater fat accumulation and osseous maturation than lower dosage
and may compromise long-term growth effects. The Journal of clinical endocrinology and
metabolism. 2013;98(2):839-45.
35. Liu P, Ma F, Lou H, Liu Y. The utility of fat mass index vs. body mass index and percentage of
body fat in the screening of metabolic syndrome. BMC Public Health. 2013;13:629.
36. Ramirez-Velez R, Correa-Bautista JE, Sanders-Tordecilla A, Ojeda-Pardo ML, Cobo-Mejia EA,
Castellanos-Vega RDP, et al. Percentage of Body Fat and Fat Mass Index as a Screening Tool for
Metabolic Syndrome Prediction in Colombian University Students. Nutrients. 2017;9(9).
179
Chapter 8
37. Yumani DFJ, de Jongh D, Lafeber HN, van Weissenbruch MM. A comparative study using dual-
energy X-ray absorptiometry, air displacement plethysmography, and skinfolds to assess fat
mass in preterms at term equivalent age. Eur J Pediatr. 2020.
38. Mank E, Naninck EFG, Limpens J, van Toledo L, van Goudoever JB, van den Akker CHP.
Enteral Bioactive Factor Supplementation in Preterm Infants: A Systematic Review. Nutrients.
2020;12(10).
180
General discussion
181
9
CHAPTER 9
Summary
Chapter 9
Preterm birth disrupts a key phase of human development. Immature organ systems
have to guide the growth and development of the preterm infant which results in
an increased risk of several comorbidities, including postnatal growth failure. This
thesis focused on the postnatal modulation of growth and body composition in very
preterm and extremely preterm infants. Furthermore, the influence of nutritional
intake and the immature endocrine axis on health outcomes was investigated. The
background of this thesis is discussed further in Chapter 1.
Part I. IGF-I and nutrition in relation to growth, body

composition and health outcomes in preterm infants
In chapter 2 a review of the literature demonstrated that preterm infants are at risk
of impaired growth and a suboptimal body composition. It was shown that IGF-I
levels in the period between preterm birth and term age play a pivotal role in the
regulation of growth from preterm birth through to infancy and seem to influence
body composition up to childhood. Low IGF-I levels between preterm birth and
term age were associated with impaired growth up to term age and possibly an
unfavourable body composition at term equivalent age.
IGF-I levels were also associated with nutritional intake: higher caloric and protein
intake were associated with higher IGF-I levels from a gestational age from 30
weeks onwards. Indeed in chapter 4, nutritional intake was associated with
IGF-I levels. In particular between 30 and 33 weeks postmenstrual age, higher
macronutrient and total caloric intake were associated with higher IGF-I levels.
Furthermore, parenteral nutrition in the second week of life, which is the major
source of nutrition at that time, was found to be associated with lower IGF-I levels.
The review of literature in Chapter 2 also demonstrated that both higher IGF-I
levels as well as increased protein intake improved neurodevelopmental outcome
in preterm infants. In addition, Chapter 3 showed that low IGF-I levels between
preterm birth and 36 weeks postmenstrual age increased the odds of BPD.
Furthermore, a higher intake of donor human milk increased the odds of BPD.
Hence there may be a window of opportunity to increase IGF-I levels through
nutritional interventions and thus improve growth, body composition and health
outcomes in infants born preterm.
Part II. Determinants and assessment of body

composition in preterm infants
There are concerns that infants born preterm have a less favourable body
composition in infancy, i.e. lower fat free mass, compared to infants born at term.
This in turn may be associated with adverse cardiometabolic outcomes in later
life. Despite these concerns, there are no guidelines on which methods should
184
Summary
be used to assess body composition in preterm infants. Nevertheless, chapter

6, demonstrated that commonly accepted methods show poor agreement.
Fat mass percentage measured with ADP showed a mean difference of 4.5 ±
4.7% when compared to fat mass percentage measured with DXA. Moreover,
there was a proportional bias: as the mean fat mass percentage increased, the
absolute difference in fat mass percentage between the two methods increased.
Furthermore, it was shown that the measurement of skinfolds could not accurately
predict fat mass measured by ADP or DXA. A systematic review of the literature in
chapter 7, showed that there are actually very few studies validating commonly
accepted methods to measure body composition in preterm infants against one
another. In line with chapter 6 body proportionality measures, skinfold thickness,
bioelectrical impedance and ultrasound were found to be a poor representation of
body composition in preterm infants. In current practice, DXA, ADP and isotope
dilution, are accepted as accurate measures. Nevertheless, there is variation within
and between these methods and our systematic review showed that the statistical
agreement between ADP and isotope dilution may be interpreted as poor.
Therefore, caution should be taken when comparing body composition measured
with different methods. To facilitate future studies and support clinical practice
it would be valuable for researchers and physicians to come to an agreement on
which reference should preferentially be used to measure body composition in
preterm infants. In our studies, we primarily used ADP and suggest the use of ADP
or DXA or isotope dilution depending on local availability and expertise.
Monitoring body composition remains important in the light of the potential

increased cardiometabolic disease risk in adults born prematurely. Early life events
and possible interventions in this period may have the potential of improving
long term outcomes of preterm infants. Indeed, in chapter 5 it was shown that
higher IGF-I levels in the first month of life are associated with increased fat free
mass at term equivalent age. Likewise, increased growth in the first month of life
was associated with an increased fat free mass percentage at term equivalent
9
age. However, a greater increase in weight, length or head circumference SDS
after this window of the first month of life was associated with a less favourable
body composition, i.e. a lower fat free mass percentage at term equivalent age.
These findings suggest that the window of opportunity to improve the body
composition of preterm infants may be limited to the early postnatal phase. In
this phase interventions to increase IGF-I levels may have the potential to improve
health outcomes of infants born preterm. Meanwhile after this phase IGF-I levels
would have reached a threshold value where IGF-I determines the growth rate and
further enrichment of nutrition could lead to an unfavourable body composition
due to increased fat mass deposition.
185
APPENDIX
10
CHAPTER 10
PhD portfolio
Chapter 10
1. PhD training
Year ECTS
General courses
• BROK (basiscursus regelgeving klinisch onderzoek) 2015 1.5
• E-BROK hercertificering 2019 0.5
Specific courses
• Open Clinica Training 2014 0.5
• Reference Manager 2014 0.5
Seminars, workshops and master classes
• JOK Pitching your research in English 2015 0.11
• TULIPS Child Health Young Researchers Day 2015 0.29
• Danone scholing “150 jaar zorg voor kinderen en het 2015 0.07
belang van voeding in de eerste 1000 dagen”
• ICaR-VU Theatre Skills Workshop 2016 0.11
• JOK time management workshop 2016 0.07
• JOK networking workshop 2017 0.07
• 25th LNF symposium 2017 0.29
190
PhD portfolio
Presentations
• Poster Presentation: Catch-up growth in preterm 2016 0.5
infants | Science Exchange Day VUmc
• Poster presentation: bone development in preterm 2016 0.5
infants | Nutrition & Growth Conference 2016
• Oral presentation: Early onset metabolic syndrome | 2017 0.5
Amsterdam Symposium 2017
• Poster presentation: Retinopathy of prematurity in 2017 0.5
relation to early life nutrition & growth | Nutrition &
Growth Conference 2017
• Oral presentation: early life nutrition & retinopathy 2017 1.0
of prematurity | Chiesi Dutch Neonatal Fellow
Meeting 2017 & Amsterdam Kindersymposium 2017
• Poster presentation: Human milk, growth & body 2017 0.5
composition in preterm infants | jENS 2017
• Oral presentation: Juggling on a high wire | 2018 0.5
Symposium Securing Future Health after Preterm
Birth
• Oral presentation: Insulin-like growth factor 1 in 2019 1.0
bronchopulmonary dysplasia, late-onset sepsis and
necrotizing enterocolitis | Chiesi Dutch Neonatal
Fellow Meeting 2019 & jENS 2019
• Poster presentation: The role of skin fold 2019 0.5
measurements, air displacement plethysmography
and dual-energy x-ray absorptiometry in the
assessment of adiposity in preterm infants at term
equivalent age | jENS 2019
• Poster presentation: A systematic review on 2020 0.5
assessing fat mass in preterm infants in early life |
Nutrition & Growth 2020
(Inter)national conferences
• Amsterdam Kindersymposium (participant) 2015
2017
0.29
-
10
• Amsterdam Kindersymposium (presenter)
• Nutrition & Growth Conference (presenter) 2016, 2017 & 2020 -
• Chiesie Dutch Neonatal Fellow Meeting (presenter) 2017 & 2019 -
• Joint European Neonatal Societies congress 2017 & 2019 -
(presenter)
Other
• Kindergeneeskunde Journal Club/Watch 2014 – 2018 2.0
• Nutrisociety research meetings 2014 – 2018 0.36
• Presentations on study progress for general 2014 – 2020 0.5
paediatrics and neonatology department
191
Chapter 10
2. Teaching
Year ECTS
Lecturing
• Keuzeonderwijs wetenschap Neonatologie 2014 & 2015 0.11
(Bachelor medicine)
• Keuzeonderwijs wetenschap Schrijfvaardigheid 2015 & 2016 0.86
(Bachelor Medicine)
• Klinisch trainingsonderwijs Kindergeneeskunde 2017 – 2018 53.57
(Master Medicine)
• Nutrition in early life (Master Health Sciences) 2018 0.07
Tutoring, Mentoring
• Tutorgroep (Bachelor medicine) 2014 – 2015 1.61
• Co-referaat 2015 – 2017 0.18
Supervising
• Marvin Eyra - Nutritional intake in association 2016 1.0
with complications in preterm infants
(wetenschappeljke stage geneeskunde)
• Khanh Vu - Human milk composition and risk 2016 – 2017 1.0
factors for metabolic syndrome at term age
(honours programme geneeskunde)
• Susan Verboon - Ophthalmologic follow up in 2017 1.0
extremely premature infants (wetenschappelijke
stage geneeskunde)
• Dide de Jong - Which measuring instrument 2017 1.0
best reflects body composition by preterm
infants? (wetenschappelijke stage
gezondheidswetenschappen)
• Isabelle Koster - Investigating the relationships 2018 -2019 1.0
between postnatal anthropometry, bone
mineralization and IGF-I in very and extremely
preterm infants up to 6 months corrected
age. (verlengde wetenschappelijke stage
geneeskunde)
• Floor Walschot - The interaction between IGF- 2018 – 2019 1.0
1 and the occurrence of NEC, LOS and BPD in
very and extremely preterm infants (verlengde
wetenschappelijke stage geneeskunde)
Other
• Leerhuisdocent Kindergeneeskunde (Teaching 2017 -2018 See under
& organising preparatory classes for medical lecturing
students’ clinical rotation in paediatrics)
192
PhD portfolio
3. Publications
Year
Peer reviewed
Publications in thesis
• Yumani DF, Lafeber HN, van Weissenbruch MM. Dietary proteins 2014
and IGF I levels in preterm infants: determinants of growth, body
composition, and neurodevelopment. Pediatr Res. 2015 Jan;77(1-2)
:156-63. doi: 10.1038/pr.2014.172. Epub 2014 Oct 21.
• Yumani DFJ, Calor AK, van Weissenbruch MM. The Course Of IGF-1 Levels 2020
and Nutrient Intake in Extremely and Very Preterm Infants During
Hospitalisation. Nutrients. 2020 Mar 2;12(3):675. doi: 10.3390/nu12030675.
• Yumani DFJ, de Jongh D, Lafeber HN, van Weissenbruch MM. A 2020
comparative study using dual-energy X-ray absorptiometry, air
displacement plethysmography, and skinfolds to assess fat mass in
preterms at term equivalent age. Eur J Pediatr. 2021 Mar;180(3):919-927.
doi: 10.1007/s00431-020-03812-3. Epub 2020 Oct 1.
• Yumani DFJ, Lafeber HN, van Weissenbruch MM. IGF-I, Growth, and 2021
Body Composition in Preterm Infants up to Term Equivalent Age. J
Endocr Soc. 2021 Jun 18;5(7):bvab089. doi: 10.1210/jendso/bvab089.
• Yumani DFJ, de Jongh D, Ket JCF, Lafeber HN, van Weissenbruch 2022
MM. Body composition in preterm infants: a systematic review on
measurement methods. Pediatr Res. 2022 Aug 22. doi: 10.1038/s41390-
022-02262-x. Epub ahead of print.
Other peer reviewed publications

• Krikke M, Yumani D, Rustenburg C, Cranendonk A, Twisk J, Lafeber H,
van Weissenbruch M. Assessing bone development in preterm infants 2017
using quantitative ultrasonography showed a decline in the early
postnatal period. Acta Paediatr. 2018 Feb;107(2):227-233. doi: 10.1111/
apa.14088. Epub 2017 Oct 25.
• Calor AK, Yumani DFJ, van Weissenbruch MM. Early Nutrition during
Hospitalization in Relation to Bone Health in Preterm Infants at Term
Age and Six Months Corrected Age. Nutrients. 2021 Apr 5;13(4):1192. doi:
2021
10
10.3390/nu13041192.
Under review
Manuscripts submitted to peer reviewed journals
• Yumani DFJ, Walschot FH, Lafeber HN, van Weissenbruch MM. 2022
Associations between bronchopulmonary dysplasia, Insulin-like
growth factor I and nutrition. Submitted to Archives of Disease in
Childhood Fetal and Neonatal Edition
Other
• Van Weissenbruch MM, Lafeber HN, Yumani DFJ. Impact van gezondheid 2015
en voeding in het vroege leven. Praktische Pediatrie. 2015 (1): 10 -16.
193
11
CHAPTER 11
Acknowledgements
Chapter 11
“Thank you God, for always leading me in your triumph in Christ and for
manifesting through me a sweet aroma of your knowledge in every place.“
This statement is on the first page of every notebook I used during this PhD
project. Indeed, it feels like a tremendous triumph, now that I finally get to write
the acknowledgement section. I’m grateful to God and all the wonderful people he
blessed me with to make this PhD project happen.
First of all, I would like to acknowledge my supervisors prof. dr. Harrie Lafeber and
prof. dr. Mirjam van Weissenbruch.
On the day I started my PhD project, Harrie asked me whether I was sure I would finish
the project. Honestly, I was a bit startled, thinking “have you ever seen me quit anything
before…”. In retrospect however, Harrie had a foresight of the rocky road ahead. We
went from a fire delaying the production of study formula, to a complete change of
the study design and extending my research time with teaching & psychiatry. Painfully,
we also had to say our last goodbye to Harrie. How I would have loved to see Harrie’s
face on the day of my defense… I’m sure he would have been so proud. Harrie was
one of the most passionate and encouraging professors I’ve ever met. I’m more than
grateful that I had the privilege of working with him and learning from him.
Lieve Mirjam, dankjewel. Een van mijn huisgenootjes noemde je mijn “work mom”
en dat is misschien wel de beste beschrijving van onze relatie. Ik heb ontzettend veel
van je geleerd als onderzoeker, maar ook persoonlijk. Ik heb eindeloos veel uren op
je kamer mogen zitten. We hebben ontzettend veel gelachen, en af en toe een traan
moeten wegpinken. Dankjewel voor je geduld en steun door alle ups & downs.
I would like to express my gratitude to the doctorate committee: prof. dr. J.B. van
Goudoever, prof. dr. R.M. van Elburg, prof. dr. A.C. Heijboer, prof dr. A.C.S. Hokken-
Koelega, dr. W. Onland and dr. J. Rotteveel, I am honored and grateful to have you
all on the committee.
This project could not have succeeded without the study participants of the NUTRIE
study and their parents. While their babies were still small and vulnerable, they
were willing to contribute to this work. Thank you so much for your willingness to
participate. I am truly grateful.
My appreciation goes to Nutricia Research and the Nutrisociety for facilitating this
research and giving me a platform to grow as a researcher.
A big thank you also goes out to the Amsterdam UMC neonatal intensive care
unit, location VUmc: all the doctors, nurses and supporting staff thank you for
contributing to the NUTRIE study and making me part of the team. Joyce, Shirley,
Meredith, Jacob, Kofi en Samson, het is altijd lachen met jullie, bedankt!
196
Acknowledgements
Chantal and your colleagues at Asito, thank you for your words of encouragement!
Annemieke en Sandra, als student mocht ik al bij jullie op de kamer zitten en

hebben jullie het onderzoek op de neonatologie aan mij geïntroduceerd. Dank
voor de gezellige tijd samen!
De vakgroep kindergeneeskunde en opleiders in het Bronovo, OLVG en Amsterdam

UMC, dankjulliewel voor de fijne tijd in de kliniek en jullie steun om door te zetten
in onderzoek. Brigitte, terwijl ik het gevoel had dat ik bij niemand opviel tijdens
mijn onderzoeksperiode, ben jij me niet vergeten. Dankjewel daarvoor en bedankt
dat je mij in de laatste fase hebt geholpen om mijn onderzoek toch af te krijgen,
ondanks de drukke kliniek.
Dide en Floor, super goede coauteurs, dankjulliewel! Marieke, Isabelle, Khan, Marvin,
Susan, Sabrina, Rajshri, Nienke, Kimberly, Juliette, Fleur, Bo, Kirsten, Roos, Janice en
Seher, dank voor jullie harde werk. Jullie hebben mooie verslagen geschreven, mij
helpen groeien als begeleider en super goed geholpen met de dataverzameling.
Marita, Kim, Jolice, Mirjam, Charlotte, Jonneke, Lindsay en Stefanie, collega’s van
kamer 033, dankjulliewel voor de gezelligheid, het foute uur, en zelfs de thee en
Tony’s (die ik bijna altijd afsloeg).
Omaima, zusje, we go way back. Ik zie je nog zitten in de bieb voordat je de eerste
keer naar Marceline ging. We hebben veel van elkaar kunnen leren in al deze jaren.
Dankjewel voor je support en dat je mijn paranimf wilde zijn.
Bouchra, geen geneeskunde zonder jou. Vanaf jaar 1 zijn we altijd steady samen
geweest. Ik heb zoveel met je kunnen delen. Dankjewel voor je luisterend oor en
motivatie als het even niet meer ging. Dankjewel dat je mijn paranimf wilde zijn.
Alexandra, topper! Dankjewel voor je support, gedeelde leed en al het comfort food.
Zondagavonden met openclinica, snacks en muziek zal ik niet meer vergeten. Je
bent een blessing!
Mindy, altijd als we samen zijn, is het weer als vanouds. Dankjewel voor je steun en
de leuke city tripjes! 11
Benie, lieverd, mijn kleine zusje. Je hebt de hele rollercoaster van dit onderzoek
meegemaakt en ik heb er heerlijk met je over kunnen lachen en klagen. Thanks
voor je support!
My Shepherd’s Place family thank you for your love and prayers. All the teens and youth
who were in explorer club with me, my people, just know that you made my week every
single time! I appreciate and love each one of you so much. Church family, ministers,
197
Chapter 11
deacons, pastor Rufaro, thank you for supporting me while I tried to navigate work &
ministry and thank you for sowing the Word into my life. Pastor Rachel, thank you for
your patience, understanding, love, prayers, correction and support when I needed it. I
have learned so much from you. Thank you for helping me through this season.
Peju, my personal person, always there when I need it. Jemimah & Uche, prayer
warriors. Thank you all for sowing into my life with our daily prayers. Ladies, you are
awesome friends and I appreciate you so much. Thank you especially for pulling
me out of my 24/7 work mode.
Tiffany, Fiyin, Obinna, Itua, and Derek thanks for your friendship and care. Derek,
thank you for your 5 am calls, so I could meet my deadlines.
Timi, words cannot express my gratitude towards you enough. You have been such
an awesome friend and mentor. Thank you for all your advice and not getting (or
showing that you were) tired of all my text messages. Thank you!
Team extra, zonder jullie was ik hier niet doorheen gekomen. Maddy, onze “9 uur
momenten” en eindeloze uren aan de telefoon zijn goud waard. Menig keer heb
je me door het schrijven heen geholpen, wanneer je vriendin weer eens iets had
gedaan en ieder ander life event dat voorbij kwam. Dankjewel, bestie!
Kiki, ba yaya bazalaka mais oyo ya ngai nde aleki. Zonder jou was ik zo ie zo niet
de persoon geworden die ik nu ben. Thanks dat je altijd naar mijn halve verhalen
wilt luisteren en dat je letterlijk altijd, in alles, achter me staat. God knew I needed
a sister like you. Dankjewel!
Jayden, jij staat aan het begin van mijn onderzoekscarrière. Terwijl ik aan het
oppassen was, heb je de eerste versies van mijn artikelen aan moeten horen. Jay,
je bent mijn pride & joy. Eliyah, tata mwasi alingaka yo mingi! Dankjewel voor de
glimlach die jij altijd op mijn gezicht brengt.
Yaya ya ba leki, grote broer, dankjewel! Je checkt altijd of het nog goed met me
gaat, of ik wel heb gegeten en niet te veel werk. Dankjewel dat je altijd achter me
staat; ook als ik weer eens de wijsneus uithang.
Bae, thank you for rooting for me and praying when I’m too tired to pray. Thank you
for bearing with the mood swings when the lack of sleep and the workload get the
best of me. Thank you for simply being in my corner.
Mama & Tijmen, dankjulliewel voor jullie steun en enthousiasme. Mama, jij bent
mijn rolmodel. Ik weet dat al jouw gebeden op de achtergrond mij hebben
gebracht tot waar ik nu ben. Ik kan jou niet genoeg bedanken. Ik houd ontzettend
veel van jou.
198
Acknowledgements
I can’t close out these acknowledgements without acknowledging my father. He

called me doctor from the age of 4. He signed every one of my crafts with dr. and
would address me as dr. DFJ Yumani in his messages. If he was still with us today,
I’m sure he would have been so overwhelmed with joy, that he wouldn’t have been
able to speak. He also would have claimed this PhD as his own. So, this PhD is
dedicated to him: mijn papa. I thank God for blessing me with a father who was
able to nurture my gifts.
Blessed be the name of the Lord from this time forth and forevermore!
Psalm 113:2 NKJV
11
199

D F J Yumani - Thesis

Uploaded by

Copyright:

Available Formats

D F J Yumani - Thesis

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

D F J Yumani - Thesis

Uploaded by

Copyright:

Available Formats

VU Research Portal

NUTRITION IN RELATION TO THE ENDOCRINE REGULATION OF PRETERM GROWTH

Link to publication in VU Research Portal

citation for published version (APA)

Take down policy

Download date: 23. May. 2024

Dana F.J. Yumani

© Copyright 2022: Dana F.J. Yumani, Amsterdam, The Netherlands

NUTRITION IN RELATION TO THE ENDOCRINE

ter verkrijging van de graad Doctor aan

de Vrije Universiteit Amsterdam,

op gezag van de rector magnificus

prof.dr. J.J.G. Geurts,

in het openbaar te verdedigen

ten overstaan van de promotiecommissie

van de Faculteit der Geneeskunde

op woensdag 12 oktober 2022 om 13.45 uur

in een bijeenkomst van de universiteit,

Dana Foekina Johanna Yumani

promotiecommissie: prof.dr. J.B. van Goudoever

PART I. THE ROLE OF NUTRITION AND IGF-I ON GROWTH, BODY 25

CHAPTER 3: Associations between bronchopulmonary dysplasia, 47

PART II. THE DETERMINANTS OF BODY COMPOSITION AND METHODS 89

CHAPTER 6: A comparative study using Dual-energy X-ray 115

DISCUSSION & SUMMARY 169

CHAPTER 9: Summary 183

The regulation of fetal and postnatal growth and

The endocrine regulation of growth and body composition

IGF-I is a small polypeptide with a wide range of function. It is mainly synthesized

Nutrition in relation to growth and body composition

intake needs to be combined with sufficient fatty acids and carbohydrates,

Aims and objectives

To investigate these hypotheses a longitudinal cohort study was designed and

Study participants were followed-up until 2 years corrected age. Anthropometric

In addition cord blood analysis of infants with a gestational age of 24 to 42 weeks

Figure 1. NUTRIE study recruitment and inclusion

Figure 2. NUTRIE Study: procedures during hospitalization

Figure 3. NUTRIE Study: procedures during follow-up

Part II of this thesis focuses on the determinants of body composition and

Dana FJ Yumani, Harrie N Lafeber and Mirjam M van Weissenbruch.

The role of IGF I in growth and body composition

Table 1. Associations between IGF I and growth

Van de Van de Giapros Chellakooty Ong (24) Socha (26)

Timing Term age 3 months 6 weeks, 3 & 3 months 3 months 6 months

IGF I & ↗ Δ weight ↗ Δ weight ↗ weightc ↗ Δ weight Not ↗ ΔWFL

↗ Δ length ↗ Δ length ↗ lengthc ↗ Δ length

IGF I & ↘ Δ weight → Δ weight → (growth ↘Δ weight →Δ weight →Δ WFL

↘ Δ length → Δ length →Δ length ↗ Δ length

↗ = positive correlation; ↘ = negative correlation; → = no correlation; Δ = gain; a = corrected age; b = chronological

Concluding these diverse associations may possibly illustrate a regulatory effect

Yet it remains to be clarified whether IGF I is primarily associated with change in

The role of IGF I in neurodevelopment

In premature infants a higher rate of increase of IGF I until 35 weeks postmenstrual

In analogy with the development of retinopathy of prematurity the sudden decrease

The role of dietary proteins in growth and body

Figure 2. Protein intake and change in weight z-sore.

Because preterm infants have limited ability to synthesize certain non-essential

The role of dietary proteins in neurodevelopment

So far pathways explaining the association between neurodevelopment and protein