Use of [13C]Bicarbonate for Metabolic Studies in Preterm
Infants: Intragastric versus Intravenous Administration
MAAIKE A. RIEDIJK, GARDI VOORTMAN, AND JOHANNES B. VAN GOUDOEVER
Department of Pediatrics [M.A.R., J.B.V.G.] and Mass Spectrometry Laboratory [G.V.], Erasmus
MC–Sophia Children’s Hospital, Rotterdam, The Netherlands.
The metabolic fate of substrates in humans can be examined
by the use of stable isotopes, one of which, [13C]bicarbonate,
may serve to estimate CO2production rate. In view of minimiz-
ing the burden of metabolic studies for preterm infants, the
authors determined whether intragastric and intravenous infu-
sions of [13C]bicarbonate would achieve the same13CO2enrich-
ment in expired air during steady state. A second aim of this
study was to determine the minimum time required to reach
steady state during intragastric infusion. Ten preterm infants
received a primed continuous [13C]bicarbonate infusion intragas-
trically, followed by an intravenous infusion the next day. Breath
samples were obtained every 30 min by the direct sampling
method.13CO2isotopic enrichment, expressed as atom percent
excess, was measured by isotopic ratio mass spectrometry. Two-
tailed t tests were used to detect statistically significant differ-
ences between the infusion routes. The isotopic enrichment at
plateau did not differ between intragastric and intravenous infu-
sion. A steady state of13CO2enrichment was achieved after 60
min of intravenous infusion and after 120 min of intragastric
infusion. In conclusion, intragastric infusion of [13C]bicarbonate
may serve to estimate the whole-body CO2production rate in
preterm infants. To reach13CO2steady state, a minimum of 120
min of bicarbonate administration is required. (Pediatr Res 58:
CI, confidence interval
AP, atom percent
APE, atom percent excess
The past two decades have seen the increased use of stable
isotopes to study amino acid metabolism in humans. These
isotopic tracer techniques have greatly enhanced our under-
standing of nutrient daily requirements and metabolism (1).
For determining the oxidation rates of specifically labeled
substrates such as amino acids or glucose, we need to quantify
substrate oxidation in each individual by measuring the13CO2
production rate during IV infusion of labeled bicarbonate (2).
The production of13CO2is made up of total CO2production
rate and13CO2enrichment in expired breath. Although total
CO2production rate is traditionally assessed by indirect calo-
rimetry,13CO2enrichment is measured by isotopic ratio mass
spectrometry. A certain amount of CO2, and thus13CO2as
well, is retained in the body. Because this amount is related to
caloric intake, a correction factor is necessary to calculate
substrate oxidation rates (3). A method that makes correction
factors and indirect calorimetry superfluous is the infusion of
NaH13CO3before the labeled substrate infusion (4).
Kien et al. (5) compared IG infusion of [13C]bicarbonate
with indirect calorimetry by the use of a correction factor. This
study showed the validation of the use of dilution stable tracer
technique to estimate CO2production. However, those authors
did not compare the13CO2enrichment during IV infusion with
IG infusion of [13C]bicarbonate.
The general purpose of this study was to determine whether
in preterm infants IG infusion of NaH13CO3yields the same
enrichment as IV infusion at steady state. To this aim, we
compared13CO2enrichment in expired breath during IG and
IV infusion of labeled bicarbonate at plateau. In addition, we
quantified the minimal tracer infusion time required to estab-
lish steady state during IG infusion.
We hypothesized that
would not differ between IG administration and IV adminis-
tration of [13C]bicarbonate.
13CO2enrichment at steady state
Subjects. We studied 10 preterm infants (8 male, 2 female) admitted to the
Neonatal Intensive Care Unit of the Erasmus MC–Sophia Children’s Hospital,
Rotterdam, The Netherlands. Their mean gestational age was 27 wks (range
Received October 29, 2004; accepted March 21, 2005.
Correspondence: J. B. van Goudoever, Erasmus MC–Sophia Children’s Hospital,
Department of Pediatrics, Division of Neonatology, Dr. Molewaterplein 60, 3015 GJ
Rotterdam, The Netherlands; e-mail: email@example.com.
Supported by the Sophia Children’s Hospital Fund, The Netherlands. This work was
also supported by Numico Research Foundation.
Copyright © 2005 International Pediatric Research Foundation, Inc.
Vol. 58, No. 5, 2005
Printed in U.S.A.
26–30 wks, SD ? 1.3 wks), and they were free of gastrointestinal diseases and
were clinically stable during the 2-day study. Five of them needed artificial
ventilation, and five breathed spontaneously with O2supplementation by nasal
prong (n ? 5). Eight infants tolerated full enteral feeding, and two infants
received partial enteral and partial parenteral feeding. For all neonates, the
feeding regimen was the same on both study days. All infants were fed through
a nasogastric feeding tube because this is a standard procedure in our unit. The
study protocol was approved by the Erasmus MC Institutional Review Board,
and written and informed consent was obtained from both parents of all
Tracer protocol. For the purpose of validating this route of labeled sodium
bicarbonate the study was designed as a randomized, crossover study. The 10
infants received a primed (10 ?mol/kg/min) continuous (10 ?mol/(kg·h)
infusion of [13C]bicarbonate (sterile pyrogen free, 99% APE; Cambridge
Isotopes, Woburn, MA). The study was set up as a true crossover design: in
five infants the IV infusion was started for 6 hours on the first day, followed
by the IG infusion on the second day. The other five infants received the IG
infusion the first day and the IV infusion the second day. One hour before the
start of the study, the usual hourly feeding regimen was changed to continuous
drip feeding. Enterally infused tracer was mixed with the milk (either fortified
or nonfortified breast milk, or preterm infant formula; Nenatal, Nutricia
Nederland B.V., Zoetermeer, The Netherlands) and infused continuously via
the nasogastric tube.
Breath samples were obtained by use of the direct sampling method
described by van der Schoor et al. (6). Briefly, in mechanically ventilated
neonates, a syringe was connected to the ventilator tubing, and breath was
taken slowly during expiration with a total volume of 15 mL. When infants
were breathing spontaneously, a 6F gastric tube (6 Ch Argyle; Cherwood
Medical, Tullamore, Ireland) was placed 1 to 1.5 cm into the nasopharynx, and
end-tidal breath was taken slowly with a syringe connected at the end.
Collected air was transferred into-10 mL sterile, non–silicon-coated evacuated
glass tubes (Van Loenen Instruments, Zaandam, The Netherlands) and stored
at room temperature until analysis.
Baseline samples were obtained 15 and 5 min before tracer infusion was
started. During the experiment, duplicate13C-enriched breath samples were
collected every 30 min and every 15 min during the last 45 min of tracer
Analytic methods.13CO2isotopic enrichment in expired air was measured
by isotope ratio mass spectrometry (ABCA; Europe Scientific, Van Loenen
Instruments, Leiden, The Netherlands) and expressed as APE above baseline.
The APE was plotted relative to time. Steady state was defined as three or more
consecutive points with a slope not different from zero. Estimated body CO2
production (mmol/kg/h) was calculated for each infant with the following
Estimated body CO2production ? IE infusate * tracer infusion rate * 1000
IE breath bicarbonate
where IE infusate is the13C enrichment of the tracer (APE), IE breath
bicarbonate is the13C enrichment in the expired air (APE), and tracer infusion
rate is the rate of [13C]bicarbonate infusion (?mol/kg/h).
Statistical analysis. Descriptive data are expressed as mean ? SD. To
define the slope of the curve of the two different methods, a repeated mea-
surements linear model was used. Steady state was achieved when the linear
factor of the slope was found to be not significantly different from zero (p ?
0.05) (8). Whole-body CO2production and baseline enrichments between the
two methods were analyzed by paired t tests.
Differences in steady state between IG and IV administration were also
analyzed by paired t tests. Statistical significance was defined as p ? 0.05.
Pitman’s test (9) was used to test the null hypothesis if the variance of
two-paired measurements (IG and IV infusion) were the same. To detect
significant differences between the two-paired measurements, a paired t test
could be performed. Pearson’s correlation coefficient was performed to show
correlation between IG and IV. The analysis of Bland and Altman (10) was
performed to show accuracy between the two different infusions. All statistical
analyses were performed by the use of SPSS version 11.0 (SPSS, Chicago, IL,
The clinical characteristics of the infants are given in Table
1. The mean study weight of the infants was 1.18 ? 0.32 kg.
The postnatal age at the start of the study was 28 ? 20 d. Their
energy intakes did not differ between both study days (p ?
0.75). The mean13C enrichments, expressed as AP, in breath
CO2from time point t ? 60 to t ? 360 min are shown in Fig.
1. All neonates achieved isotopic steady state in both admin-
istration routes. Baseline enrichments did not differ between IG
and IV infusion (1.0875 AP ? 0.0022 versus 1.0869 AP ?
0.0338, p ? 0.29).
The mean APE at plateau (t120–360) during IG infusion was
0.0365 ? 0.0055; during IV infusion it was 0.0371 ? 0.0067.
IG enrichment was slightly lower, though not significantly,
than IV enrichment (p ? 0.59).
The Pitman’s tests (9) showed no significant difference
between variance in IV and IG infusion (p ? 0.308), and the
Pearson’s correlation coefficient was 0.359. Agreement be-
tween the two different routes of administration was deter-
mined by the analysis of Bland and Altman (10). Figure 2
shows on the x axis the average of the IV plateau and the IG
plateau (n ? 10), whereas the y axis shows the difference
between the two measurements (n ? 10). The mean difference
is 0.0006 APE. Note that all measurements lie between the
range of the mean difference ?2 SD (0.0076 APE) and the
mean difference ?2 SD (?0.0064 APE). The 95% CI of the
mean difference is ?0.0019 to 0.0031 APE. Therefore, from
120 min onward, there was no statistically significant differ-
ence in CO2enrichment in expired air between IV or IG
infusion, nor did we find a sequence effect (no significant
difference in13CO2between infants who received NaH13CO3
Table 1. Clinical characteristics of 10 studied infants
Key: e ? enteral intake; p ? parenteral intake; v ? ventilation; np ? nasal prong.
RIEDIJK ET AL.
IV the first day or those who received NaH13CO3IG the first
The estimated CO2production did not differ between the IG
(27.68 ? 5.38 mmol/kg/h) and IV (27.67 ? 5.64 mmol/kg/h)
infusions (p ? 0.99).
Steady state was achieved from 60 min onward when the
tracer was infused IV and from 120 min onward when it was
The main purpose of this study was to validate the use of IG
administration of [13C]bicarbonate compared with IV admin-
istration for metabolic oxidation studies in preterm infants.
Clinical studies in addition to experimental research are of
great value in elucidating metabolism and nutrition in preterm
infants. Information about amino acid metabolism and protein
synthesis and oxidation is needed to provide these infants with
optimal nutrition and consequently improved growth and
A principal goal of many tracer kinetic experiments is to
determine the oxidation rate of the tracer substance by the
appearance in breath of labeled C originating from the tracer
(11). The gold standard for determining whole-body CO2
production is indirect calorimetry (3). An alternative method is
a primed continuous IV infusion of NaH13CO3. We found the
estimated body CO2production (27.67 ? 5.64 mmol/kg/h) to
be similar to that previously described (0.725 ? 0.021 mol/
kg/day) (12). Also, others have shown that NaH13CO3can be
adequately used as a method of determining CO2production
rate (4,5,13). The infusion of labeled bicarbonate before a
13C-labeled substrate carries the advantage that no correction
factor is needed to calculate substrate oxidation. In addition, IG
infusion of the tracer reduces the invasiveness of metabolic
studies. Finally, in studying the metabolic fate of an enteral
substrate, it is preferable to administer the tracer enterally as
Hoerr et al. (11) studied in adults the effects of IG and IV
infusion of labeled bicarbonate on recovery of13C in breath
and concluded that administration route did not affect recovery.
When it is considered that placing an IV catheter in preterm
infants is highly invasive, it is very important to search for
methods minimizing discomfort.
To achieve steady state during IG administration, tracer
infusion should last at least 120 min. Sample collection is
accomplished during steady state. Consequently, breath sam-
ples should be obtained from 120 min onward. To prevent
intrasubject variation, at least four breath samples should be
obtained at 10-min intervals, thus between 120 and 160 min of
We need to emphasize the small sample size of this study.
However, we presented a 95% CI (?0.0019 to 0.0031 APE) of
the mean difference to obtain an impression of a type II error.
We considered a difference of ?10% between IG plateau and
IV plateau to be acceptable. We calculated the difference of the
minimal (?5%) and maximal (8%) of the 95% CI limit of the
average plateau of IG and IV (0.0368 APE). As we assumed,
the plateau of IV and IG infusion can vary from ?5% to 8% in
the general population.
Additionally, we wish to stress that in metabolic studies in
parenterally fed infants, [13C]bicarbonate should preferably be
In conclusion, our findings are consistent with the absence of
significant differences in13CO2enrichment between IG and IV
infusion after 120 min of infusion, and therefore it would be
valid to infuse [13C]bicarbonate IG for the determination of
whole-body CO2production rate in preterm infants.
Acknowledgments. The authors thank Chris van den Akker,
Frans te Braake, and Ineke van Vliet for their support; Paul
Mulder for statistical help; and Ko Hagoort for critical review
of the manuscript.
Figure 2. Bland-Altman analysis showing the difference between IG and IV
enrichments of13CO2in breath of 10 infants. The mean difference is 0.0006
APE (dotted line). All measurements are within two standard deviations: ? 2
SD (0.0076 APE) and ?2 SD (?0.0064 APE). The 95% CI of the mean
difference is ?0.0019 to 0.0031 APE.
Figure 1. Breath13CO2enrichments of 10 infants, expressed as mean AP ?
SD. IV (●) vs IG (°) infusion of [13C]-bicarbonate. At plateau (t120–t360)13C
enrichment IV is not significantly different from IG (p ? 0.59).
ENTERAL NaH13CO3INFUSION IN PRETERM INFANTS
1. Kleinman RE, Barness LA, Finberg L 2003 History of pediatric nutrition and fluid
therapy. Pediatr Res 54:762–772
2. Kingdon CC, Mitchell F, Bodamer OA, Williams AF 2000 Measurement of carbon
dioxide production in very low birth weight babies. Arch Dis Child Fetal Neonatal Ed
3. Van Aerde JE, Sauer PJ, Pencharz PB, Canagarayar U, Beesley J, Smith JM, Swyer
PR 1985 The effect of energy intake and expenditure on the recovery of 13CO2 in the
parenterally fed neonate during a 4-hour primed constant infusion of NAH13CO3.
Pediatr Res 19:806–810
4. van Goudoever JB, Sulkers EJ, Chapman TE, Carnielli VP, Efstatopoulos T,
Degenhart HJ, Sauer PJ 1993 Glucose kinetics and glucoregulatory hormone
levels in ventilated preterm infants on the first day of life. Pediatr Res 33:583–
5. Kien CL, McClead RE 1996 Estimation of CO2 production in enterally fed preterm
infants using an isotope dilution stable tracer technique. JPEN J Parenter Enteral Nutr
6. van der Schoor SR, de Koning BA, Wattimena DL, Tibboel D, van Goudoever JB
2004 Validation of the direct nasopharyngeal sampling method for collection of
expired air in preterm neonates. Pediatr Res 55:50–54
7. van Goudoever JB, Stoll B, Henry JF, Burrin DG, Reeds PJ 2000 Adaptive regulation
of intestinal lysine metabolism. Proc Natl Acad Sci U S A 97:11620–11625
8. Hoerr RA, Matthews DE, Bier DM, Young VR 1991 Leucine kinetics from [2H3]-
and [13C]leucine infused simultaneously by gut and vein. Am J Physiol 260:E111–E117
9. Pitman EJG 1939 A note on the normal correlation. Biometrica 31:9–12
10. Bland JM, Altman DG 1986 Statistical methods for assessing agreement between two
methods of clinical measurement. Lancet 1:307–310
11. Hoerr RA, Yu YM, Wagner DA, Burke JF, Young VR 1989 Recovery of 13C in
breath from NaH13CO3 infused by gut and vein: effect of feeding. Am J Physiol
12. Shew SB, Beckett PR, Keshen TH, Jahoor F, Jaksic T 2000 Validation of a
[13C]bicarbonate tracer technique to measure neonatal energy expenditure. Pediatr
13. Spear ML, Darmaun D, Sager BK, Parsons WR, Haymond MW 1995 Use of
[13C]bicarbonate infusion for measurement of CO2 production. Am J Physiol
RIEDIJK ET AL.