Published online: 2020-03-02

SMFM Fellowship Series Article

Maternal Amino Acid Profiles to Distinguish

Constitutionally Small versus
Growth-Restricted Fetuses Defined
by Doppler Ultrasound: A Pilot Study
Anne C. Porter, MD1 Diane L. Gumina, MS1,2 Michael Armstrong, BS3 Kenneth N. Maclean, PhD4
Nichole Reisdorph, PhD3 Henry L. Galan, MD1 Sally P. Stabler, MD5 Beth A. Bailey, PhD1
John C. Hobbins, MD1 K. Joseph Hurt, MD, PhD1,2

1 Division of Maternal Fetal Medicine, Department of Obstetrics and Address for correspondence K. Joseph Hurt, MD, PhD, Room P15-
Gynecology, University of Colorado Anschutz Medical Campus, 3003, Mail Stop 8613, 12700 East 19th Avenue, Aurora, CO 80045
Aurora, Colorado (e-mail: k.joseph.hurt@cuanschutz.edu).
2 Division of Reproductive Sciences, Department of Obstetrics and
Gynecology, University of Colorado Anschutz Medical Campus,
Aurora, Colorado
3 Department of Pharmaceutical Sciences, University of Colorado

Downloaded by: Karolinska Institutet. Copyrighted material.

Anschutz Medical Campus, Aurora, Colorado
4 Department of Pediatrics, University of Colorado Anschutz Medical
Campus, Aurora, Colorado
5 Division of Hematology, Department of Medicine, University of
Colorado Anschutz Medical Campus, Aurora, Colorado

Am J Perinatol

Abstract Objective Fetuses measuring below the 10th percentile for gestational age may be
either constitutionally small for gestational age (SGA) or have pathologic fetal growth
restriction (FGR). FGR is associated with adverse outcomes; however, identification of
low-risk SGA cases is difficult. We performed a pilot study evaluating maternal markers
of pathologic FGR, hypothesizing there are distinct amino acid signatures that might be
used for diagnosis and development of new interventions.
Study Design This was a cohort study of healthy women with sonographic fetal
estimated fetal weight <5th percentile divided into two groups based upon umbilical
artery (UmA) Doppler studies or uterine artery (UtA) Doppler studies. We collected
maternal blood samples prior to delivery and used ion pair reverse phase liquid
chromatography-mass spectrometry or gas chromatography-mass spectrometry to
assess 44 amino acids.
Keywords Results Among 14 women included, five had abnormal UmA, and three had abnormal
► amino acid UtA Doppler results. Those with abnormal UmA showed elevated ornithine. Those with
► fetal Doppler abnormal UtA had lower dimethylglycine, isoleucine, methionine, phenylalanine, and
velocimetry 1-methylhistidine.
► fetal growth Conclusion We found several amino acids that might identify pregnancies affected
restriction by pathologic FGR. These findings support the feasibility of future larger studies to
► uterine artery identify maternal metabolic approaches to accurately stratify risk for small fetuses.

received Copyright © by Thieme Medical DOI https://doi.org/

December 28, 2019 Publishers, Inc., 333 Seventh Avenue, 10.1055/s-0040-1701504.
accepted New York, NY 10001, USA. ISSN 0735-1631.
December 30, 2019 Tel: +1(212) 760-0888.
Maternal Amino Acid Profiles and Fetal Dopplers Porter et al.
Fetuses with estimated fetal weight (EFW) below the 10th
percentile for gestational age have higher rates of stillbirth and
perinatal death.1 However, not all cases of EFW <10th percen-
tile have increased morbidity, as approximately 70% of these
fetuses have reached their normal genetic growth potential
inherited from the parents.2–4 The term small for gestational
age (SGA), which was originally used for neonates with birth-
weight <10th percentile, may be most appropriate for fetuses
with EFW between the 3rd and 10th percentile with normal
Doppler studies. Those small fetuses are likely “constitutionally
small” and at lower risk. Therefore, SGA can be used to describe
normal but small fetuses whereas fetal growth restriction (FGR)
is appropriate for pathophysiologic processes that cause low
EFW. The objective of this study was to investigate the feasibil-
ity of a noninvasive approach to risk stratify small fetuses as
FGR or SGA using maternal serum profiles.
Maternal, fetal, and placental pathways can all limit fetal
growth, but identification of lower risk SGA fetuses is difficult.
As a result, all fetuses with EFW <10th percentile undergo
intensive antenatal monitoring and may require early delivery.
Such monitoring includes serial umbilical artery Doppler
studies, with some practitioners performing uterine artery
(UtA) Doppler studies to further risk stratify pregnancies.
Decreased umbilical artery diastolic flow associated with
increased placental resistance may impair, or represent im-
pairment of, placental oxygen and nutrient exchange restrict-
ing fetal growth and development. Changes in placental
resistance also can be reflected in abnormal UtA Doppler flow.5
Both abnormal amino acid transport by the placenta and
decreased substrate availability resulting from severe ma-
ternal nutrient restriction can cause FGR.6–12 However, FGR
in such cases may not be entirely explained by substrate
availability alone but involve complex adaptive biology. For
example, decreased maternal protein intake has been asso-
ciated with decreased uterine blood flow and FGR in an
animal model.13 Normal fetal and placental development
requires dramatic increases in uterine blood flow, and FGR is
associated with inadequately increased UtA diameter or
flow.14,15 We conjectured that specific maternal amino acids,
such as those necessary for the hyperplasia, hypertrophy,
and vascular smooth muscle relaxation that facilitate normal
UtA flow in pregnancy, might be associated with low EFW.
Amino acid profiles can be used to reveal critical metabolic
processes, mechanisms, and pathways. To date, limited data
exist on maternal metabolomic and amino acid profiles in
FGR.16 Although two prior studies evaluated maternal metab-
olomic networks in the prediction of adverse perinatal out-
comes such as FGR, neither included amino acids.17,18
Evaluation of maternal amino acid profiles is an attractive
target in FGR research as previous work has shown that limited
amino acid availability or transport, particularly essential ami-
no acids, can affect fetal growth.19,20 Such deficiencies may then
be reflected in maternal levels of essential amino acids.
We hypothesized that specific amino acids are altered in
pathologic placental FGR pregnancies providing a unique
maternal amino acid profile in cases of pathologic FGR as
opposed to constitutional SGA. To test this, we performed a
pilot study to identify associations between maternal serum
American Journal of Perinatology
amino acids using previously collected samples from an
ongoing FGR cohort. We aimed to evaluate the feasibility
of evaluating maternal serum amino acid profiles associated
with abnormal umbilical or UtA Doppler studies.
Materials and Methods
Subjects
This study was performed with Institutional Review Board
approval (COMIRB #14–1360), and all women provided
written informed consent. We performed a cohort study of
healthy women with fetuses at less than the 5th percentile.
Women were recruited at two sites within a single tertiary
care university-affiliated center. In an attempt to include
only severe growth restriction not otherwise explained by
maternal comorbidities, inclusion and exclusion criteria
were strictly applied. Inclusion criteria were women aged
>18 years at time of delivery with an estimated fetal weight
<5th percentile identified on more than one prenatal growth
ultrasound.21 All fetuses had normal genetic screening or
Downloaded by: Karolinska Institutet. Copyrighted material.
were confirmed normal at live birth. Women were excluded
if they had severe medical disease, used medications associ-
ated with FGR, used tobacco or marijuana, had multiple
gestation, or had fetal anomalies identified on ultrasound.
Data Collection
Trained research staff collected and analyzed maternal
demographic information from detailed review of the medi-
cal record and patient interview. For each neonate, both
routine Fenton curves and individual customized growth
centiles (accounting for physiological/demographic factors
known to influence fetal growth) were calculated.22–24
Maternal serum samples were collected within 6 weeks prior
to delivery and used for analysis. For one subject, only plasma
was available for analysis; this sample was still included
given previous reports of largely consistent metabolite levels
between sera and plasma.25
Umbilical Artery Doppler Studies
At the time of diagnosis of EFW <5th percentile, we performed
both umbilical and UtA Doppler studies. Umbilical artery
Doppler studies were repeated at each return visit with fre-
quency of visits defined by clinical standard of care (►Fig. 1),
from biweekly to weekly depending on Doppler assessments.
Doppler waveforms were acquired on a GE Voluson E10 using
the RM6C matrix probe. Umbilical arteries and veins were
sampled in a free loop of cord with an angle of insonation
<30 degrees. The average peak systolic to end diastolic (S/D)
velocity ratio from both arteries was used to determine normal
and abnormal status. Abnormal was defined as S/D >95th
percentile for gestational age.26 Umbilical vein velocity (time-
averaged maximum [TAMAX]) was similarly obtained, and
umbilical vein flow (UVF) was calculated using the formula:
UVF ¼ (TAMAX cm/s)  (vein area cm2)  (60)  (0.5). Venous
vessel area was calculated using the area ellipse function in a
cross-sectional view of the umbilical cord on the ultrasound
machine; specular reflection was observed to ensure that the
image was a true cross-section. The vessel boundary for area
 Maternal Amino Acid Profiles and Fetal Dopplers
Fig. 1 Doppler waveforms examined for this study. At each visit the following parameters were evaluated: umbilical artery Doppler waveform
(A), umbilical vein Doppler waveform (B), uterine artery Doppler waveform (C). PSV, peak systolic velocity; EDV, end diastolic velocity; N,
notching.
calculation was generated within the vessel intima. Flow was
also normalized to estimated fetal weight (kg).27
Uterine Artery Doppler Studies
UtA Doppler waveforms were obtained similarly to the
approach used for umbilical arteries described above. The UtA
can be identified by color Doppler velocimetry at its origin at the
anterior division of the hypogastric artery. Pulsed Doppler is
performed after the UtA crosses the hypogastric artery prior to
its division into uterine and cervical branches.28 Velocity in the
left and right uterine arteries was acquired, and vessel diameter
was measured using the outer edge of color-fill to generate
vessel area. Flow for each artery was calculated using the
formula: (TAMAX cm/s)  (artery area cm2)  (60)  (0.5).29,30
Flow from both vessels were then averaged. A scoring system
was used to define abnormal UtA Doppler status as previously
described.31 One point was given for pulsatility index >1.2 and
one point for notching, for a total possible score of 4 from
bilateral assessment. The presence of notching was evaluated by
an expert senior author (J.H.). Scores >2 were considered
abnormal.
Ion-Pair Reverse Phase Liquid Chromatography Mass
Spectrometry or Gas Chromatography with Mass
Spectrometry Analysis
LC-MS was performed as previously described on an LC-Agilent
1200 series.32 Briefly, samples were injected into a 150-mm
 2.1-mm  1.8-μm Colum-Agilent SB-C18 column operated at
ambient temperature. Buffer A was 0.5% formic acid, 0.3%
heptafluorobutyric acid in HPLC-grade water and buffer B
was 0.5% formic acid, 0.3% heptafluorobutyric acid in acetoni-
trile. Samples were eluted at an initial flow rate of 0.3
mL/minute with a chromatographic gradient as follows: 0% B
for 5 minutes, 1 to 100% B from 5 to 11 minutes, hold at 100% B
for 1 minute, and re-equilibrate at 0% B for 8 minutes. Detection
of amino acids was accomplished using an MS Agilent 6410
triple quadruple mass spectrometer operated in positive elec-
trospray ionization mode. The MS parameters were as follows:
capillary voltage 4000 V, gas flow 9.0 L/minute, gas temperature
300C, nebulizer pressure 50 psi. MRM transitions for each
amino acid were optimized with Agilent Optimizer software
using authentic standards. Raw data from analysis were quan-
titated using Agilent Masshunter Quantitative Analysis soft-
Porter et al.
ware. A total of 44 amino acids were calibrated at levels from
15.6 to 1,000 nM/mL using LC-MS. Nine amino acids in the
methionine cycle and transsulfuration pathway were analyzed
as described previously using GC-MS after complete reduction
of the sample to yield total unoxidized amino acids.33
Downloaded by: Karolinska Institutet. Copyrighted material.
Statistical Analysis
We first compared demographics and samples in groups
defined by normal and abnormal umbilical artery Doppler
status. We then performed a similar but separate analysis
with comparing groups defined by normal and abnormal UtA
Doppler results. Given the small sample size in this pilot
study, no sample size calculation was performed. One-tailed,
independent group t-tests were performed to compare the
mean serum amino acid concentrations. In comparison of
methionine cycle metabolites, a single outlier was identified
and recoded to three standard deviations above the mean for
analysis.34 For those amino acids that were different in either
the umbilical or UtA Doppler comparisons, we performed
correlation analyses between amino acid and umbilical vein
or UtA flow. GraphPad Prism version 7 was used for all
statistical analyses (San Diego, CA). Differences were consid-
ered statistically significant at p < 0.05.
Results
Subject Selection
Subject selection is outlined in ►Fig. 2. Of 2,773 pregnant
patients screened, 73 were enrolled in the low EFW cohort at
the time of our study. Of those, only 26 had an EFW <5th
percentile, and after inclusion and exclusion were applied, 14
healthy subjects remained for our analysis. Of these 14 patients,
9 had normal umbilical artery Doppler studies and 10 had
normal UtA Doppler studies. Of the initial 73 patients enrolled,
only five had an abnormal UtA score >2 with three of those
ultimately included in our severe growth restriction cohort. Of
the three subjects with abnormal UtA Doppler studies, two also
had abnormal umbilical artery Doppler studies.
Analysis of Maternal Amino Acids by Umbilical Artery
Doppler Status
Our overall cohort is described in ►Table 1. The group was
predominantly Caucasian with an average maternal age of
American Journal of Perinatology
Maternal Amino Acid Profiles and Fetal Dopplers Porter et al.

Downloaded by: Karolinska Institutet. Copyrighted material.

Fig. 2 Subject selection flow sheet. AC, abdominal circumference; EFW, estimated fetal weight; UmA, umbilical artery; UtA, uterine artery.

Table 1 Demographics by normal and abnormal umbilical artery Doppler status

Factors Overall Normal Abnormal p-Value
n ¼ 14 n¼9 n¼5
# (%) # (%) # (%)
Maternal characteristics
Race 0.35
White 13 (93) 8 (89) 5 (100)
American Indian or Alaskan 1 (7) 1 (11) 0
Maternal age (y) 31.1 29.7  6.3 33.6  1.9 0.10
Prepregnancy BMI (kg/m2) 22.9 23.6  2.2 22.4  3.9 0.14
Gravidity 2 2 3 0.68
Parity 1 1 1 0.82
History of FGR 0.31
Yes 3 (21) 1 (11) 2 (40)
No 11 (79) 8 (89) 3 (60)
On baby aspirin 0.07
Yes 4 (29) 1 (11) 3 (60)
No 9 (64) 7 (78) 2 (40)
Not reported 1 (7) 1 (11) 0
Pregnancy and neonatal characteristics, mean  SD (#)
GA at FGR diagnosis (wk) 27.3 27.8  3.3 26.4  5.9 0.30
GA at delivery (wk) 36.2 36.4  4.1 35.9  1.8 0.40
UmV absolute flow (mL/min) 196.9 232.7  26.7 132.5  27.2 0.02
UmV (mL/min/kg) 111.4 120.8  13.3 88.2  17.8 0.08
BMI at delivery (kg/m2) 27.4 28.8  2.0 (5) 26.6  4.7 (5) 0.09
Weight gain in pregnancy (lb) 25.7 27.2  7.4 (5) 24.1  17.7 (5) 0.38
Neonatal birth weight (kg) 2.8 2.4  0.2 (9) 1.6  0.4 (4) 0.0002
Neonatal birth weight (percentile) 6.8 8.6  6.2 (9) 1.8  1.0 (4) 0.03
Customized growth (percentile) 2.4 3.4  1.1 (9) 0.1  0.1 (4) 0.02
Neonatal length (cm) 45.2 46.8  0.8 (9) 41.8  2.7 (4) 0.0003
Ponderal index (g/cm3) 2.3 2.4  0.09 (9) 2.2  0.06 (4) 0.07

Abbreviations: BMI, body mass index; FGR, fetal growth restriction; GA, gestational age; SD, standard deviation; UmV, umbilical vein.
Note: Bold denotes statistically significant items.

American Journal of Perinatology

 Maternal Amino Acid Profiles and Fetal Dopplers Porter et al.

31.1 years. The average gestational age at diagnosis of EFW Table 2 Serum concentration of maternal metabolites by
<5th percentile was 27 weeks. When grouped by umbilical umbilical artery Doppler status
artery Doppler status, there were no significant differences
in maternal or birth characteristics. However, umbilical vein Metabolite Normal Abnormal p-Value
UmA (9) UmA (5)
absolute flow was significantly lower in the abnormal group.
In addition, neonates were significantly smaller in the 1-methylhistidine 3.3  0.8 3.2  0.6 0.45
abnormal group with lower birthweight, birth length, birth- 3-methylhistidine 9.5  7.2 9.8  14.6 0.48
weight percentile, and customized growth percentile. α-amino-n-butyric 10.3  3.9 8.9  2.4 0.13
Of the 44 maternal amino acids analyzed by LC-MS, 13 were acida
below the detection limit (phosphoserine, phosphoethanol- Alanine 436.4  85.3 407.3  66 0.26
amine, B-alanine, g-aminobutyric acid, L-aminoadipic acid, Arginine 72.5  21.1 85.6  35.7 0.20
L-cystine, hydroxylysine, cystathionine, reduced glutathione, Asparagine 41.9  13.5 45.0  13.6 0.35
oxidized glutathione, carnosine, anserine, and homocysteine).
Aspartic acid 16.7  8.3 16.4  9.3 0.48
For the additional nine amino acids analyzed by GC-MS, six had
β-aminoisobutyric 1.0  0.8 0.4  0.3 0.10
also been analyzed by LC-MS. Given the increased detection
acid
using GC-MS, values from GC-MS were used for comparison of
Citrulline 19.0  7.0 18.9  4.9 0.49
those amino acids and adding three additional analytes. In total,
33 amino acids were evaluated. Of these, only ornithine was Creatinine 55.8  12.0 60.8  13.3 0.24
significantly different between the two groups. Ornithine levels Cystathioninea 228.9  94.9 210.8  71.9 0.35
in mothers of fetuses with abnormal umbilical artery Doppler Cysteinea 176.1  38.8 170.2  20.5 0.17

Downloaded by: Karolinska Institutet. Copyrighted material.

status were approximately 30% higher than fetuses with nor- Dimethylglycine a
7.0  1.5 6.5  0.4 0.07
mal umbilical artery Doppler status (►Table 2; ►Fig. 3). Glutamic acid 167.7  119.7 166.9  33.7 0.49
Glutamine 441.7  155.4 342.3  88.6 0.11
Analysis of Maternal Amino Acids by Uterine Artery
Glycinea 126.2  21.5 131.4  31.4 0.27
Doppler Status
When grouped by normal versus abnormal UtA Doppler status, Histidine 143.2  42.3 119.2  17.5 0.13
there were no significant differences in maternal or birth Homocysteinea 3.6  0.8 4.2  0.8 0.35
characteristics; however, those with abnormal UtA Doppler Hydroxyproline 12.5  16.4 4.3  4.5 0.15
studies were diagnosed with EFW <5th percentile significantly Isoleucine 52.0  13.5 51.3  20.5 0.47
earlier in pregnancy (22 vs. 28 weeks). Neonates were not
Leucine 81.5  16.1 86.9  34.6 0.35
different in terms of birthweight or birth length (►Table 3).
Lysine 209.8  56.5 227.1  69.4 0.31
Of the 33 maternal amino acids that were above the
detection threshold on either LC-MS or GC-MS, maternal Methioninea 17.7  4.0 12.8  4.3 0.47
dimethylglycine, isoleucine, methionine, phenylalanine, Ornithine 35.7  3.2 45.7  9.8 0.007
and 1-methylhistidine were all significantly lower in the Phenylalanine 58.8  9.7 60.7  14.3 0.38
group with abnormal UtA Doppler studies. Most prominent- Proline 153.9  30.6 171.5  44.3 0.20
ly, isoleucine was approximately 40% lower in the abnormal Sarcosinea 2.3  2.3 1.8  0.8 0.17
UtA Doppler group, and phenylalanine was approximately a
Serine 85.7  20.4 84.2  11.0 0.13
20% lower (►Table 4; ►Fig. 4).
Taurine 73.4  35.5 76.6  26.9 0.43

Amino Acid and Flow Correlations Threonine 346.4  160.3 270.4  84.5 0.17
For those amino acids identified as significantly different in Tryptophan 47.4  9.8 46.3  9.0 0.42
either the umbilical artery or UtA comparisons, we performed Tyrosine 56.3  16.3 65.8  37.0 0.26
correlation analyses with umbilical vein or UtA flow, respec- Valine 209.1  43.6 205.5  85.2 0.46
tively. Ornithine was significantly correlated with absolute
UVF (r ¼ 0.63, p ¼ 0.02) but not UVF/EFW in kg (r ¼ 0.49, Abbreviation: GC-MS, gas chromatography–mass spectrometry; UmA,
umbilical artery.
p ¼ 0.08). There were no significant correlations between UtA
Note: Mean values  standard deviation are shown in nM/mL. Bold
average flow and maternal levels of dimethylglycine, isoleu- denotes statistically significant items.
a
cine, methionine, phenylalanine, or 1-methylhistidine. Denotes analysis by GC-MS.

only severely small fetuses, hoping to identify the high-risk

Discussion
This pilot study confirms the feasibility of our approach. We
identified several preliminary associations between specific
amino acids and abnormal Doppler studies in fetuses with
growth <5th percentile. The ability to distinguish between
constitutional SGA and pathologic FGR is a primary concern for
perinatologists who manage such pregnancies. We examined
(i.e., pathologic or FGR) group by maternal amino acid assay.
Maternal ornithine levels were significantly higher in preg-
nancies with abnormal umbilical artery Doppler studies.
When compared by umbilical artery Doppler status, we found
no differences in essential and conditionally essential amino
acids that were previously shown to influence fetal
growth.11,35–37 On the other hand, abnormal UtA blood flow

American Journal of Perinatology

Maternal Amino Acid Profiles and Fetal Dopplers Porter et al.

was associated with several essential amino acids. Specifically,

maternal levels of dimethylglycine, isoleucine, methionine,
phenylalanine, and 1-methylhistidine were all lower in cases
of abnormal UtA blood flow. In this pilot study, we identified an
amino acid profile in the readily available maternal compart-
ment that if replicated on a larger scale might be used to
identify at-risk pregnancies in pathologic FGR.
Our findings highlight a possible role of ornithine in altered
fetal growth, given its demonstrated increase in cases of abnor-
mal umbilical artery Doppler studies. Ornithine is a precursor for
synthesis of polyamines via ornithine decarboxylase. Poly-
amines such as putrescine, spermidine, and spermine have
key roles in cell proliferation and differentiation, angiogenesis,
embryogenesis and fetal, and placental growth.38 Putrescine in
particular is involved in angiogenesis via activation of mTOR
signaling.39 Prior research in a sheep model showed that
maternal protein restriction reduced ornithine in fetal plasma
and amniotic fluid.40 The higher maternal ornithine levels seen
in our study may reflect impaired transport by the placenta,
resulting in decreased fetal ornithine and thus impaired growth

Downloaded by: Karolinska Institutet. Copyrighted material.

Fig. 3 Serum concentration of significantly different maternal me- potential due to limited polyamines. Alternatively, FGR may
tabolite by umbilical artery Doppler status.

Table 3 Demographics by normal and abnormal uterine artery Doppler status

Factors Overall Normal Abnormal p-Value

n ¼ 13 n ¼ 10 n¼3
# (%) # (%) # (%)
Maternal characteristics
Race 0.70
White 12 (92) 9 (90) 3 (100)
American Indian or Alaskan 1 (8) 1 (10) 0
Maternal age (y) 31.2 31  6.5 31.7  1.2 0.43
2
Prepregnancy BMI (kg/m ) 23.1 23.7  2.3 21.7  4.7 0.16
Gravidity 2 2 2 0.36
Parity 1 1 1 0.28
History of FGR 0.78
Yes 3 (23) 2 (20) 1 (33)
No 10 (77) 8 (80) 2 (67)
On baby aspirin 1.00
Yes 4 (31) 3 (30) 1 (33)
No 8 (62) 6 (60) 2 (67)
Not reported 1 (8) 1 (10) 0
Pregnancy and neonatal characteristics, mean  SD (#)
GA at FGR diagnosis (wk) 27.1 28.5  3.8 22.5  2.7 0.02
GA at delivery (wk) 36.1 36.0  3.7 36.3  2.9 0.46
Average UtA flow (mL/min) 600.2 735.5  168.5 (9) 594.5  94.8 (3) 0.33
BMI at delivery (kg/m2) 27.4 26.8  2.5 (7) 26.3  6.1 (3) 0.43
Weight gain in pregnancy (lb) 25.7 25.5  8.2 (7) 26.0  23.1 (3) 0.48
Neonatal birth weight (kg) 2.1 2.3  0.4 (9) 1.9  0.7 (3) 0.11
Neonatal birth weight (percentile) 6.8 7.9  7.0 (9) 4.0  3.6 (3) 0.17
Customized growth (percentile) 2.6 3.1  1.2 (9) 1.0  1.0 (3) 0.18
Neonatal length (cm) 45.0 45.7  1.6 (9) 42.9  4.8 (3) 0.07
Ponderal index (g/cm3) 2.3 2.4  0.09 (9) 2.3  0.06 (3) 0.32

Abbreviations: BMI, body mass index; FGR, fetal growth restriction; GA, gestational age; SD, standard deviation; UtA, uterine artery.
Note: Bold denotes statistically significant items.

American Journal of Perinatology

 Maternal Amino Acid Profiles and Fetal Dopplers Porter et al.

Table 4 Serum concentration of maternal metabolites by fetus, alterations in flow may result in decreased substrate
uterine artery Doppler status availability. Because we only examined maternal blood from a
single point in later pregnancy, we are unable to distinguish the
Metabolite Normal UtA Abnormal p-Value cause or effect and timing of maternal ornithine abnormalities.
(10) UtA (3)
Although we identified no differences in essential amino
1-methylhistidine 3.5  0.7 2.6  0.3 0.04 acids by umbilical artery Doppler status, when grouped by UtA
3-methylhistidine 10.7  10.7 4.4  7.7 0.19 status several amino acids were significantly lower in mothers
α-amino-n-butyric 10.1  3.8 7.9  1.7 0.09 with abnormal Doppler studies. Of these, phenylalanine, iso-
acida leucine, and methionine are all essential amino acids. Previous
Alanine 428.7  84.8 410.3  78.7 0.37 work has primarily focused on essential amino acids in the fetal
Arginine 73.5  19.9 89.2  50.4 0.21 compartment, with multiple studies showing alterations in
Asparagine 41.3  13.7 43.6  11.4 0.40 cases of FGR.43–45 Methionine is of particular interest in FGR
given the key role of the methionine cycle in one carbon
Aspartic acid 15.7  8.5 18.5  10.6 0.32
metabolism, which is crucial for cell proliferation and growth.
β-aminoisobutyric 0.8  0.8 0.5  0.4 0.27
Depletion of methionine leads to decreased S-adenosylmethio-
acid
nine (SAM) required for synthesis of several polyamines as
Citrulline 20.7  6.1 14.1  4.8 0.06
well.46 Furthermore, SAM provides the methyl group for meth-
Creatinine 60.5  11.6 50.2  14.4 0.11 ylation-dependent epigenetic effects which may be altered in
Cystathioninea 235.9  93.2 178.3  63.1 0.14 FGR.47
Cysteinea 175.7  37.8 162.7  12.1 0.19 Our finding of decreased dimethylglycine is notable given

Downloaded by: Karolinska Institutet. Copyrighted material.

Dimethylglycine a
7.1  1.3 6.3  0.1 0.04 the decreased methionine levels. Homocysteine is metabolized
Glutamic acid 177.1  109.4 164.9  36.0 0.43 to methionine by betaine-homocysteine methyltransferase,
which also generates dimethylglycine. Elevated homocysteine
Glutamine 402.5  153 369.8  100.3 0.37
is associated with cardiovascular and thrombotic morbidity as
Glycinea 126.0  24.0 121.3  18.7 0.37
well as adverse pregnancy outcomes such as preeclampsia and
Histidine 139.3  40.9 113.4  14.0 0.16 growth restriction.48,49 Although we did not detect elevated
Homocysteinea 3.8  0.9 3.8  0.6 0.46 homocysteine, we speculate that decreased dimethylglycine
Hydroxyproline 12.2  15.6 3.8  3.5 0.19 and methionine may represent an interruption in homocyste-
Isoleucine 57.1  14.0 34.2  9.1 0.01 ine metabolism, thus increasing homocysteine levels; however,
our pilot study may simply be underpowered to detect a
Leucine 88.7  23.6 66.9  20.4 0.09
difference. Taken together, the decreased maternal essential
Lysine 220.0  69.9 196.6  10.4 0.29
amino acids in our study highlight the possibility that specific
Methioninea 16.6  4.9 12.4  2.2 0.03 decreased amino acid substrate availability precedes restricted
Ornithine 38.7  8.7 42.2  6.3 0.26 UtA flow and decreased fetal and placental growth.
Phenylalanine 63.5  10.4 49.0  5.1 0.02 This study has several strengths. We used a novel approach
Proline 168.6  35.6 147.6  28.0 0.19 to identify pathophysiologic factors in FGR, and enrolled
Sarcosinea 2.4  2.1 1.4  0.1 0.09 rare patients with extremely small fetuses. Our patient data
a were prospectively collected, and the amino acid analyses were
Serine 84.3  18.2 82.3  16.2 0.44
performed with state-of-the-art techniques. We defined our
Taurine 68.1  22.7 82.5  55.5 0.25
groups by clinically relevant Doppler assessment rather than
Threonine 294.6  142.7 324.8  71.9 0.37 normal and abnormal fetal growth percentiles, making the
Tryptophan 48.3  10.0 44.4  8.0 0.28 comparisons most relevant to the practicing perinatologist.
Tyrosine 65.1  26.4 46.0  17.5 0.14 Furthermore, these comparisons discriminate associations
Valine 223.8  57.3 160.7  50.9 0.06 with constitutional and pathologically small fetuses, which
could potentially alter management, allowing constitutionally
Abbreviations: GC–MS, gas chromatography–mass spectrometry; UtA, small SGA fetuses to forego some intensive surveillance.
uterine artery.
Although previous studies showed that global maternal nutri-
Note: Mean values  standard deviation are shown in nM/mL. Bold
denotes statistically significant items. ent restriction lowers amino acid levels and can cause FGR, the
a
Denotes analysis by GC-MS. role of specific amino acid changes leading to altered uterine
blood flow has not been studied. We identified several sub-
precede altered ornithine uptake through a yet undefined strates that could support the increased uterine blood flow
mechanism. Our finding that ornithine is correlated with UVF necessary for normal pregnancy growth. In addition, most
may support ornithine’s role in angiogenesis via polyamine prior work on metabolic profiles in pregnancy has focused
availability. Decreased UVF in cases of FGR with abnormal on the fetal compartment via umbilical vessel or fetal sampling
umbilical artery Doppler studies, as seen in our study, is consis- which is not as useful in routine clinical practice. Instead, we
tent with prior research showing that reduced umbilical vein focused on the more readily accessible maternal compartment,
velocity resulted in decreased flow in FGR.41,42 As the umbilical which could provide convenient opportunities for diagnosis
vein is the conduit of oxygen and nutrients to the developing and intervention during pregnancy. It is possible, of course, that

American Journal of Perinatology

Maternal Amino Acid Profiles and Fetal Dopplers Porter et al.
Fig. 4 Serum concentration of significantly different maternal metabolites by uterine artery Doppler status.
fetal or placental abnormalities only occur after the inciting
event that initially causes FGR, after which intervention would
be less beneficial. Nonetheless, identifying those fetuses, which
are small but at their maximum growth potential, would be
clinically useful.
As this is only a pilot study, we recognize that our work has
several limitations. Given strict inclusion and exclusion criteria
and a hypothesis-generating approach, our sample size is small.
In addition, we only determined levels of amino acids as
opposed to entire metabolic networks. We analyzed only one
maternal blood draw from the third trimester, obtained as a
convenience sample under nonfasting conditions, which may
not reflect the fetal compartment. We did not correct for
gestational age given the small sample size. The current find-
ings cannot distinguish between association and causation,
which was further limited by our lack of normal controls.
However, addition of controls would result in three groups
for analysis, further decreasing the power to detect differences.
In addition, amino acid differences may result from several
pathways including changes in fetal production, alterations in
placental uptake or efflux, or variable maternal availability.
However, although therapies for FGR are certainly of interest,
our focus currently is on the accurate identification of patho-
logic, nonconstitutional FGR. Therefore, the etiology of signifi-
cant changes in the maternal amino acid profile is a secondary
concern for now.
In conclusion, we have highlighted the possible association of
specific amino acids in the maternal compartment with abnor-
mal Doppler studies and pathologic FGR; however, additional
research is needed, and this study identifies several targets for
additional investigation. The role of essential branched chain
American Journal of Perinatology
Downloaded by: Karolinska Institutet. Copyrighted material.
amino acids and sulfur-containing amino acids in UtA flow
modulation is of special interest to our group. Future research
will plan to assess broader metabolomic networks to hopefully
identify additional targets for FGR diagnosis and therapy.
Note
This work was presented in part at the Society for Repro-
ductive Investigation meeting 2018 in San Diego, CA.
Funding
This work was supported by an SMFM/AAOGF Scholar
Award (K.J.H.), the Perelman Study for Fetal Growth
Restriction (D.L.G., H.L.G., and J.H.), the William R Hum-
mel Homocystinuria Research Fund (K.N.M.), and the
Ehst-Hummel-Kaufmann Family Endowed Chair in Inher-
ited Metabolic Disease (K.N.M.).
Conflict of Interest
None declared.
References
1 Mayer C, Joseph KS. Fetal growth: a review of terms, concepts and
issues relevant to obstetrics. Ultrasound Obstet Gynecol 2013;41
(02):136–145
2 Ott WJ. The diagnosis of altered fetal growth. Obstet Gynecol Clin
North Am 1988;15(02):237–263
3 Group GS; GRIT Study Group. A randomised trial of timed delivery
for the compromised preterm fetus: short term outcomes and
Bayesian interpretation. BJOG 2003;110(01):27–32
4 Unterscheider J, Daly S, Geary MP, et al. Optimizing the definition
of intrauterine growth restriction: the multicenter prospective
PORTO Study. Am J Obstet Gynecol 2013;208(04):290.e1–290.e6
 Maternal Amino Acid Profiles and Fetal Dopplers
5 Sciscione AC, Hayes EJ; Society for Maternal-Fetal Medicine.
Uterine artery Doppler flow studies in obstetric practice. Am J
Obstet Gynecol 2009;201(02):121–126
6 Pantham P, Rosario FJ, Nijland M, et al. Reduced placental amino
acid transport in response to maternal nutrient restriction in the
baboon. Am J Physiol Regul Integr Comp Physiol 2015;309(07):
R740–R746
7 Paolini CL, Marconi AM, Ronzoni S, et al. Placental transport of
leucine, phenylalanine, glycine, and proline in intrauterine
growth-restricted pregnancies. J Clin Endocrinol Metab 2001;
86(11):5427–5432
8 Marconi AM, Paolini CL, Stramare L, et al. Steady state maternal-fetal
leucine enrichments in normal and intrauterine growth-restricted
pregnancies. Pediatr Res 1999;46(01):114–119
9 Brett KE, Ferraro ZM, Yockell-Lelievre J, Gruslin A, Adamo KB.
Maternal-fetal nutrient transport in pregnancy pathologies: the
role of the placenta. Int J Mol Sci 2014;15(09):16153–16185
10 Sibley CP, Turner MA, Cetin I, et al. Placental phenotypes of
intrauterine growth. Pediatr Res 2005;58(05):827–832
11 Pantham P, Rosario FJ, Weintraub ST, et al. Down-regulation of
placental transport of amino acids precedes the development of
intrauterine growth restriction in maternal nutrient restricted
baboons. Biol Reprod 2016;95(05):98
12 Jansson N, Pettersson J, Haafiz A, et al. Down-regulation of placental
transport of amino acids precedes the development of intrauterine
growth restriction in rats fed a low protein diet. J Physiol 2006;576
(Pt 3):935–946
13 Lekatz LA, Swanson TJ, Camacho LE, et al. Maternal metabolizable
protein restriction during late gestation on uterine and umbilical
blood flows and maternal and fetal amino acid concentrations
near term in sheep. Anim Reprod Sci 2015;158:115–125
14 Konje JC, Howarth ES, Kaufmann P, Taylor DJ. Longitudinal quan-
tification of uterine artery blood volume flow changes during
gestation in pregnancies complicated by intrauterine growth
restriction. BJOG 2003;110(03):301–305
15 Velauthar L, Plana MN, Kalidindi M, et al. First-trimester uterine
artery Doppler and adverse pregnancy outcome: a meta-analysis
involving 55,974 women. Ultrasound Obstet Gynecol 2014;43
(05):500–507
16 Conde-Agudelo A, Papageorghiou AT, Kennedy SH, Villar J. Novel
biomarkers for predicting intrauterine growth restriction: a sys-
tematic review and meta-analysis. BJOG 2013;120(06):681–694
17 Heazell AE, Bernatavicius G, Warrander L, Brown MC, Dunn WB. A
metabolomic approach identifies differences in maternal serum
in third trimester pregnancies that end in poor perinatal outcome.
Reprod Sci 2012;19(08):863–875
18 Horgan RP, Broadhurst DI, Walsh SK, et al. Metabolic profiling
uncovers a phenotypic signature of small for gestational age in
early pregnancy. J Proteome Res 2011;10(08):3660–3673
19 Jansson T, Scholtbach V, Powell TL. Placental transport of leucine
and lysine is reduced in intrauterine growth restriction. Pediatr
Res 1998;44(04):532–537
20 Lager S, Powell TL. Regulation of nutrient transport across the
placenta. J Pregnancy 2012;2012:179827
21 Hadlock FP, Harrist RB, Martinez-Poyer J. In utero analysis of fetal
growth: a sonographic weight standard. Radiology 1991;181(01):
129–133
22 Gardosi J, Chang A, Kalyan B, Sahota D, Symonds EM. Customised
antenatal growth charts. Lancet 1992;339(8788):283–287
23 Customised Centile Calculator JG. In: A F, ed.: Gestation Network.
Available at: www.gestation.net. Accessed January 14, 2018
24 Fenton TR, Kim JH. A systematic review and meta-analysis to
revise the Fenton growth chart for preterm infants. BMC Pediatr
2013;13:59
25 Yu Z, Kastenmüller G, He Y, et al. Differences between human plasma
and serum metabolite profiles. PLoS ONE 2011;6(07):e21230
26 Acharya G, Wilsgaard T, Berntsen GK, Maltau JM, Kiserud T.
Reference ranges for serial measurements of umbilical artery
Porter et al.
Doppler indices in the second half of pregnancy. Am J Obstet
Gynecol 2005;192(03):937–944
27 Barbera A, Galan HL, Ferrazzi E, et al. Relationship of umbilical
vein blood flow to growth parameters in the human fetus. Am J
Obstet Gynecol 1999;181(01):174–179
28 Berkley E, Chauhan SP, Abuhamad A; Society for Maternal-Fetal
Medicine Publications Committee. Doppler assessment of the
fetus with intrauterine growth restriction. Am J Obstet Gynecol
2012;206(04):300–308
29 Wilson MJ, Lopez M, Vargas M, et al. Greater uterine artery blood
flow during pregnancy in multigenerational (Andean) than
shorter-term (European) high-altitude residents. Am J Physiol
Regul Integr Comp Physiol 2007;293(03):R1313–R1324
30 Flo K, Wilsgaard T, Acharya G. Longitudinal reference ranges for
umbilical vein blood flow at a free loop of the umbilical cord.
Ultrasound Obstet Gynecol 2010;36(05):567–572
31 Hernandez-Andrade E, Brodszki J, Lingman G, Gudmundsson S,
Molin J, Marsál K. Uterine artery score and perinatal outcome.
Ultrasound Obstet Gynecol 2002;19(05):438–442
32 Armstrong M, Jonscher K, Reisdorph NA. Analysis of 25 under-
ivatized amino acids in human plasma using ion-pairing reversed-
phase liquid chromatography/time-of-flight mass spectrometry.
Rapid Commun Mass Spectrom 2007;21(16):2717–2726
Downloaded by: Karolinska Institutet. Copyrighted material.
33 Allen RH, Stabler SP, Lindenbaum J. Serum betaine, N,N-dimethyl-
glycine and N-methylglycine levels in patients with cobalamin and
folate deficiency and related inborn errors of metabolism. Metabo-
lism 1993;42(11):1448–1460
34 Barnett V, Lewis T. Outliers in statistical data. 3rd ed. Chichester;
New York: Wiley; 1994
35 Lin G, Wang X, Wu G, et al. Improving amino acid nutrition to
prevent intrauterine growth restriction in mammals. Amino Acids
2014;46(07):1605–1623
36 Brown LD, Green AS, Limesand SW, Rozance PJ. Maternal amino
acid supplementation for intrauterine growth restriction. Front
Biosci (Schol Ed) 2011;3:428–444
37 Bhasin KK, van Nas A, Martin LJ, Davis RC, Devaskar SU, Lusis AJ.
Maternal low-protein diet or hypercholesterolemia reduces cir-
culating essential amino acids and leads to intrauterine growth
restriction. Diabetes 2009;58(03):559–566
38 Wu G, Bazer FW, Cudd TA, Meininger CJ, Spencer TE. Maternal
nutrition and fetal development. J Nutr 2004;134(09):2169–2172
39 Kong X, Wang X, Yin Y, et al. Putrescine stimulates the mTOR
signaling pathway and protein synthesis in porcine trophecto-
derm cells. Biol Reprod 2014;91(05):106
40 Dunford LJ, Sinclair KD, Kwong WY, et al. Maternal protein-energy
malnutrition during early pregnancy in sheep impacts the fetal
ornithine cycle to reduce fetal kidney microvascular develop-
ment. FASEB J 2014;28(11):4880–4892
41 Ferrazzi E, Rigano S, Bozzo M, et al. Umbilical vein blood flow in
growth-restricted fetuses. Ultrasound Obstet Gynecol 2000;16
(05):432–438
42 Rigano S, Bozzo M, Ferrazzi E, Bellotti M, Battaglia FC, Galan HL.
Early and persistent reduction in umbilical vein blood flow in the
growth-restricted fetus: a longitudinal study. Am J Obstet Gyne-
col 2001;185(04):834–838
43 Cosmi E, Visentin S, Favretto D, et al. Selective intrauterine growth
restriction in monochorionic twin pregnancies: markers of endo-
thelial damage and metabolomic profile. Twin Res Hum Genet
2013;16(04):816–826
44 Lin G, Liu C, Feng C, et al. Metabolomic analysis reveals differences
in umbilical vein plasma metabolites between normal and
growth-restricted fetal pigs during late gestation. J Nutr 2012;
142(06):990–998
45 Sanz-Cortés M, Carbajo RJ, Crispi F, Figueras F, Pineda-Lucena A,
Gratacós E. Metabolomic profile of umbilical cord blood plasma
from early and late intrauterine growth restricted (IUGR) neo-
nates with and without signs of brain vasodilation. PLoS One
2013;8(12):e80121
American Journal of Perinatology
Maternal Amino Acid Profiles and Fetal Dopplers Porter et al.

46 Kalhan SC. One carbon metabolism in pregnancy: impact on
maternal, fetal and neonatal health. Mol Cell Endocrinol 2016;
435:48–60
47 Kalhan SC, Marczewski SE. Methionine, homocysteine, one car-
bon metabolism and fetal growth. Rev Endocr Metab Disord 2012;
13(02):109–119
48 Vollset SE, Refsum H, Irgens LM, et al. Plasma total homocysteine,
pregnancy complications, and adverse pregnancy outcomes: the
Hordaland Homocysteine study. Am J Clin Nutr 2000;71(04):
962–968
49 Loscalzo J. Homocysteine-mediated thrombosis and angiostasis
in vascular pathobiology. J Clin Invest 2009;119(11):3203–3205

Downloaded by: Karolinska Institutet. Copyrighted material.

American Journal of Perinatology