The Journal of Nutrition

Methodology and Mathematical Modeling

See corresponding commentary on page 2207.

Use of a ‘‘Super-child’’ Approach to Assess the

Vitamin A Equivalence of Moringa oleifera
Leaves, Develop a Compartmental Model for
Vitamin A Kinetics, and Estimate Vitamin A

Downloaded from https://academic.oup.com/jn/article-abstract/147/12/2356/4727994 by guest on 28 September 2019

Total Body Stores in Young Mexican Children
Veronica Lopez-Teros,1 Jennifer Lynn Ford,2 Michael H Green,2 Guangwen Tang,3 Michael A Grusak,4
Luis Quihui-Cota,5 Tawanda Muzhingi,6 Mariela Paz-Cassini,5 and Humberto Astiazaran-Garcia5
1
Nutritional Sciences, Department of Chemical and Biological Sciences, University of Sonora, Hermosillo, Sonora, Mexico; 2Department
of Nutritional Sciences, Pennsylvania State University, University Park, PA; 3Former Carotenoids and Health Laboratory, Jean Mayer
USDA Human Nutrition Research Center on Aging, Tufts University, Boston, MA; 4USDA Agricultural Research Service, ChildrenÕs
Nutrition Research Center, Department of Pediatrics, Baylor College of Medicine, Houston, TX; 5Department of Nutrition, Research
Center for Food and Development, Hermosillo, Sonora, Mexico; and 6International Potato Centre (CIP), International Livestock
Research Institute Campus, Nairobi, Kenya

Abstract
Background: Worldwide, an estimated 250 million children <5 y old are vitamin A (VA) deficient. In Mexico, despite
ongoing efforts to reduce VA deficiency, it remains an important public health problem; thus, food-based interventions that
increase the availability and consumption of provitamin A–rich foods should be considered.
Objective: The objectives were to assess the VA equivalence of 2H-labeled Moringa oleifera (MO) leaves and to estimate
both total body stores (TBS) of VA and plasma retinol kinetics in young Mexican children.
Methods: b-Carotene was intrinsically labeled by growing MO plants in a 2H2O nutrient solution. Fifteen well-nourished
children (17–35 mo old) consumed puréed MO leaves (1 mg b-carotene) and a reference dose of [13C10]retinyl acetate
(1 mg) in oil. Blood (2 samples/child) was collected 10 times (2 or 3 children each time) over 35 d. The bioefficacy of MO
leaves was calculated from areas under the composite ‘‘super-child’’ plasma isotope response curves, and MO VA
equivalence was estimated through the use of these values; a compartmental model was developed to predict VA TBS and
retinol kinetics through the use of composite plasma [13C10]retinol data. TBS were also estimated with isotope dilution.
Results: The relative bioefficacy of b-carotene retinol activity equivalents from MO was 28%; VA equivalence was 3.3:1 by weight
(0.56 mmol retinol:1 mmol b-carotene). Kinetics of plasma retinol indicate more rapid plasma appearance and turnover and more
extensive recycling in these children than are observed in adults. Model-predicted mean TBS (823 mmol) was similar to values
predicted using a retinol isotope dilution equation applied to data from 3 to 6 d after dosing (mean 6 SD: 832 6 176 mmol; n = 7).
Conclusions: The super-child approach can be used to estimate population carotenoid bioefficacy and VA equivalence, VA
status, and parameters of retinol metabolism from a composite data set. Our results provide initial estimates of retinol
kinetics in well-nourished young children with adequate VA stores and demonstrate that MO leaves may be an important
source of VA. J Nutr 2017;147:2356–63.

Keywords: retinol kinetics, retinol isotope dilution, b-carotene bioconversion, bioefficacy, compartmental analysis

Introduction
Vitamin A deficiency (VAD) continues to be a major nutritional
problem in many parts of the world. In 2005, the WHO esti-
mated, based on the presence of low serum retinol concentra-
tions (<0.7 mmol/L) in preschool-age children, that VAD was a
public health concern in 122 countries (1). In the case of Mexico,
results from the Mexican Nutritional and Health Survey 2012 (2)
indicate that VAD (serum retinol <0.7 mmol/L) affected 15.7% of
preschool-age children.
Although the cause of VAD is multifactorial, low dietary
intake of vitamin A (VA) plays a major role in its development.
In Mexico, efforts have been made to reduce VAD in children by

ã 2017 American Society for Nutrition.

2356 Manuscript received June 21, 2017. Initial review completed July 28, 2017. Revision accepted August 21, 2017.
First published online September 20, 2017; doi: https://doi.org/10.3945/jn.117.256974.
providing high-dose VA supplements during national immuni-
zation and health campaigns and through food fortification
programs (3, 4). Although it is logical to use preformed VA to
ameliorate VAD, the possibility of excessive intake poses risks
for toxicity, including liver damage and increased bone loss (5,
6). Because provitamin A carotenoids are a safe alternative,
researchers have been interested in identifying food-based
interventions that increase the availability and consumption of
provitamin A–rich foods (7).
To identify plant foods that may be good sources of dietary
VA (8), investigators have used stable isotope methods to estimate
the VA equivalence of intrinsically labeled carotenoids relative to a
reference dose of labeled retinyl acetate (9–12). For b-carotene, a
VA equivalence of 12:1 by weight [i.e., 12 mg b-carotene is
equivalent to 1 mg retinol or retinol activity equivalents (RAE)]
has been used (4, 13). In fact, the VA equivalence of carotenoids
depends on numerous factors including the food matrix, fat
content, and method of preparation (14, 15), and on the genotype
and VA status of the individual consuming the food (8).
In 2007, in an alternative and unconventional strategy to
increase dietary VA intake, leaves from the Moringa oleifera
(MO) plant were used as a VA supplement in several Mexican
regions (Sonora, Michoacán, and Oaxaca) (16). The MO tree is
native to western and sub-Himalayan areas of India, Pakistan,
Asia, and Africa, and it is now well distributed throughout the
Philippines, Cambodia, the Americas (from Mexico to Brazil),
and the Caribbean islands (17). The plant grows in many areas
where a high prevalence of VAD exists, according to the WHO
(1). The efficacy and effectiveness of MO as a plant-based
complementary food with good nutritional value (18) has, to
our knowledge, not yet been assessed.
In this study we used the stable isotope reference method
(19) to assess the relative bioefficacy of carotenoids and the
VA equivalence of intrinsically labeled MO leaves in young
Mexican children who ingested MO leaves intrinsically
labeled with deuterium and a reference dose of [13C10]retinyl
acetate. By applying the ‘‘super-child’’ approach described by
Haskell et al. (20), we obtained composite data sets for plasma
retinol kinetics while limiting the number of blood samples
collected from each child. In addition to assessing bioefficacy,
we used model-based compartmental analysis (21) to describe
plasma [13C10]retinol kinetics and to estimate total body stores
(TBS) of VA. Our results provide insights into the metabolism of
retinol in well-nourished children and indicate that MO leaves may
be a potentially important source of dietary VA.
Supported by the International Atomic Energy Agency research contract 16880
(to HA-G), the Nutricia Research Foundation (2012-25), and the USDA
Agricultural Research Service through agreement 58-6250-0-008 (to MAG).
MP-C received a fellowship from the National Research and Technology Council
(CONACyT) of Mexico.
Author disclosures: VL-T, JLF, MHG, GT, MAG, LQ-C, TM, MP-C, and HA-G, no
conflicts of interest.
The contents of this publication do not necessarily reflect the views or policies of
the USDA, nor does mention of trade names, commercial products, or
organizations imply endorsement by the US government.
Supplemental Table 1, Supplemental Figure 1, and Supplemental WinSAAM
Deck are available from the ‘‘Online Supporting Material’’ link in the online
posting of the article and from the same link in the online table of contents at
http://jn.nutrition.org.
Present address for MAG: USDA Agricultural Research Service, Red River Valley
Agricultural Research Center, 1605 Albrecht Boulevard North, Fargo, ND 58102.
Address correspondence to HA-G (e-mail: hastiazaran@ciad.mx).
Abbreviations used: FCRs, system fractional catabolic rate; FD, fraction of the
dose; MO, Moringa oleifera; RAE, retinol activity equivalents; SAp, retinol specific
activity in plasma; TBS, total body stores; VA, vitamin A; VAD, vitamin A
deficiency.
Methods
13
C-labeled VA. [13C10]retinyl acetate (8-, 9-, 10-, 11-, 12-, 13-, 14-, 15-,
19-, 20-[13C10]retinyl acetate) was synthesized to $98% chemical purity
by Cambridge Isotope Laboratories. Aliquot doses (1 mg or 2.96 mmol
[13C10]retinyl acetate) were placed in gelatin capsules and ;200 mL corn
oil were added to each capsule. Capsules were prepared at the Carotenoids
and Health Laboratory in Boston, Massachusetts, and then shipped on dry
ice to the Research Center for Food and Development in Hermosillo,
Sonora, Mexico. Upon arrival, capsules were stored at 270°C until they
were administered to participating children.
Intrinsically labeled MO. Moringa plants were grown hydroponically
at the ChildrenÕs Nutrition Research Center (Houston, Texas) through
the use of a nutrient solution enriched with 23 at.% deuterium oxide, as
Downloaded from https://academic.oup.com/jn/article-abstract/147/12/2356/4727994 by guest on 28 September 2019
previously described (22). Plants were started from seeds and grown to
5 wk of age, at which time leaves were harvested. Leaves were shipped to
the Carotenoids and Health Laboratory in Boston and analyzed by
HPLC to assess the carotenoid profile and concentrations (23, 24);
through the use of LC-MS, the percentage enrichment of [2H]b-carotene
was confirmed to be sufficient to carry out the proposed human studies.
Leaves were then shipped on dry ice to the Research Center for Food and
Development in Hermosillo; they were stored at 270°C for subsequent
dose preparation.
On the day of dosing, 7-g portions of MO leaves containing 1 mg
[2H]b-carotene were weighed, individually steamed in small bags for
2.5 min, and then soaked in cold purified water for 2 min. Leaves were
drained and homogenized using an industrial homogenizer (Ultra-Turrax
T25 Basic; IKA Works) at 24,000 3 g in 100 g applesauce (Gerber;
Nestlé), which, according to the label, contained 6.9 mg VA (0% daily
value). Homogenates were placed in an ice chest with cool packs and
delivered to day care centers for immediate consumption by the study
subjects.
Subjects and study design. Children (17–35 mo old) from middle-
income socioeconomic areas in Northwest Mexico were eligible to
participate in this experimental, longitudinal, prospective study. Selec-
tion of participants was intentional and nonprobabilistic. During
recruitment, the 35-d study protocol (Figure 1) was explained in detail
to parents at selected day care centers, and signed informed consent was
obtained from those willing to have their children evaluated for
inclusion. Fifteen children were screened and enrolled. As part of the
Mexican National Immunization campaign, all 15 children received
a 60-mg VA supplement 3 mo before the study. Inclusion criteria
were absence of subclinical inflammation as assessed by C-reactive
protein (>6 mg/L; catalog no. CP2488; Randox Laboratories Ltd) (25)
and absence of iron deficiency anemia (hemoglobin <110 g/L) (26).
Exclusion criteria included malnutrition (any anthropometric index z
score <22 SD), signs and symptoms of xerophthalmia, and presence of
pathogenic parasites in stool samples (27, 28). Based on the selection
criteria described above, no children had to be excluded. The study was
approved by the Research Center for Food and Development Bioethics
Committee (CE/002/2013) and by the School Health and Safety
Department of the Sonoran Education Secretary.
After an overnight fast, the dose of intrinsically labeled MO leaves
in applesauce was given to each child at their day care center. The
amount of [2H]b-carotene ingested was determined by weighing the
container that held the homogenate before and after the child ate. After
consumption of the homogenate, a capsule containing 1 mg [13C10]
retinyl acetate in corn oil was given to each participant. Blood samples
(2 samples/child, except for 1 child from whom only 1 sample was
obtained) were collected at 10 sampling times (7, 9.5, and 12 h, and 1, 3,
5, 6, 14, 21, and 35 d; 2 or 3 children/time point) after the labeled doses
were consumed. At each sampling time, 3–5 mL venous blood was
drawn from the subjects using either a Vacutainer Safety Lok blood
collection set or a syringe plus a clot activator gel Vacutainer tube
(Becton Dickinson), which was covered with aluminum foil to prevent
photodegradation of VA. Blood samples were placed on cool packs for
transport. Serum was obtained and stored at 270°C, then later ana-
lyzed for serum retinol and carotenoids or shipped on dry ice to the
Super-child model for vitamin A value and kinetics 2357

FIGURE 1 Study protocol and sampling times. Isotopes (2H
carotenoids in intrinsically labeled Moringa oleifera leaves and [13C10]
retinyl acetate in oil) were administered on day 0 of the 35-d
experiment; blood samples were obtained from 2 or 3 children at each
time point (7, 9.5, and 12 h, or 1, 3, 5, 6, 14, 21, and 35 d).
Carotenoids and Health Laboratory, where it was stored at 270°C until
isotopic enrichment was analyzed.
Dietary information. Because breakfast and lunch were provided at day
care centers, dietary information was obtained from the menus at each
institution. To estimate daily dietary VA intake, we weighed foods pro-
vided at participating day care centers. In addition, parents completed an
FFQ to ascertain what foods were frequently consumed outside of the
day care center. Nutrient intakes were calculated using the USDA food
composition tables (29). VA adequacy was assessed based on the RDA of
300 mg RAE/d for 1- to 3-y-old children (13).
Analyses. Retinol and carotenoids were extracted from serum and
analyzed by reverse-phase HPLC, as previously described (24), with
retinyl acetate and echinenone as internal standards. Analyte concen-
trations were calculated through the use of external calibration curves
derived from pure standards obtained from Sigma-Aldrich. In addition,
we used pooled serum to assess interrun variation (<9% for both retinol
and provitamin A carotenoids).
Isotopic enrichment of [13C10]- and [2H]retinol in serum was
determined by GC–electron capture negative chemical ionization MS
(30). Total enrichment of labeled retinol was determined by integrating
the peaks under the reconstructed mass chromatograms of the negative
ions at 268–270 m/z for [H] and 278–280 m/z for [13C]retinol. Retinol
derived from b-carotene from MO (i.e., [2H]retinol) was calculated as 2
times the sum of 271–274 m/z, based on the assumed symmetric
distribution of the labeled b-carotene in MO leaves; that is, we calculated
percentage enrichment (30):
½Labeled retinol O ðunlabeled retinol þ labeled retinolÞ 
ð1Þ
3  100
All procedures were carried out under red light. To reduce
interassay variability, sample analyses were performed simultaneously,
and personnel who analyzed the samples were blinded to time-point
assignments.
‘‘Super-child’’ data sets. Serum enrichment data for [13C]- and [2H]
retinol were converted to plasma fraction of dose (FD):
Plasma FD ¼ ½enrichment 3 retinol concentration ðmmol=LÞ
3 estimated plasma volume ðLÞ
O dose ðmmol RAEÞ ð2Þ
We used plasma volume to analyze bioefficacy and kinetics; we calculated
plasma volume for each child using estimated blood volume (liters) and
an average hematocrit value for children in this age group, where blood
volume was estimated using age (years), height (centimeters), and body
weight (kilograms) (Figure 2 and Table 1 in reference 31). Because 1 mol
b-carotene can produce 2 mol retinol, we converted the b-carotene dose
(micromoles) to molar RAE (micromoles) by multiplying the b-carotene
dose (micromoles) by 2. Values for the geometric mean plasma FD at
each time were plotted versus time, providing composite (‘‘super-child’’)
2358 Lopez-Teros et al.
response curves for the 2 isotopes. We used geometric means because
they are an appropriate indicator of central tendency.
Bioefficacy and VA equivalence. To estimate the VA equivalence of
b-carotene from MO leaves, we first estimated relative bioefficacy using
the isotope reference method described by Tang et al. (19), for which
areas under the isotope response curves were calculated using NCSS11
statistical software (version 11.0.6) and then used in Equation 3:
Relative bioefficacy ð%Þ ¼ ðAUCFD-2H =AUCFD-13C Þ 3 100 ð3Þ
where AUCFD-2H and AUCFD-13C are the areas under the FD-versus-time
curves for [2H]- and [13C]retinol, respectively, integrated from 0 to 21 d.
Note that ‘‘units’’ for AUC are thus FD 3 days. We then used bioefficacy
(Equation 3) to calculate the relative VA yield, or amount of [2H]retinol
derived from the [2H]b-carotene in MO leaves, using Equation 4:
Downloaded from https://academic.oup.com/jn/article-abstract/147/12/2356/4727994 by guest on 28 September 2019
2 
H retinol ðmmol RAEÞ ¼ bioefficacy ðfractionÞ
  ð4Þ
3 2 H b-carotene dose ðmmol RAEÞ
Next, to determine the VA equivalence (micrograms of b-carotene to
micrograms of retinol) of labeled b-carotene in MO, we compared the
amount of b-carotene in the oral dose with the amount of retinol derived
from it using Equation 5:
VA equivalence 5 ½ b-carotene in MO leaves ðmmolÞ 3 536
=½2 Hretinol derived from
½2 Hb-carotene doseðmmolÞ 3 286 ð5Þ
where b-carotene in MO leaves (micromoles) was calculated from the
1-mg [2H]b-carotene dose and [2H]retinol (micromoles) derived from
[2H]b-carotene was calculated using Equation 4.
VA kinetics and the super-child model. Data for geometric mean
plasma FD for [13C10]retinol were plotted versus time (7 h to 35 d) to
provide a composite (‘‘super-child’’) isotope response curve over the 35-d
study. Then we applied model-based compartmental analysis (21) using
the Windows version of the Simulation, Analysis and Modeling Software
[WinSAAM version 3.3.0 (32)] to develop a multicompartment model
describing whole-body VA kinetics (see Results). Once a satisfactory fit
was obtained between the observed data and the model predictions,
weighted nonlinear regression analysis (with a fractional SD of 0.05 as
the weighting factor) was used to estimate final values for the model
FIGURE 2 Composite plasma isotope response curve and model-
simulated data for the 2 extravascular pools. The graphs show
geometric mean plasma FD data (circles) for [13C10]retinol vs. time for
14 children (2 or 3 children/time) after ingestion of a labeled dose of
retinyl acetate and the model-calculated fit to the data (solid line). Also
shown are FD data vs. time simulated for the extravascular pools with
faster (dotted line) and slower (dashed line) turnover (compartments
4 and 5, respectively; Figure 3). FD, fraction of dose.
TABLE 1 General characteristics of participating children
(n = 15)
Mean 6 SD
Age, y 2.2 6 0.6
Weight, kg 13.1 6 1.5
Height, cm 88.6 6 5.7
z Score
Weight-for-age 0.4 6 1 children 1–3 y old [300 mg RAE/d (13)], results indicate that VA
Height-for-age 0.3 6 1.2
Weight-for-height 0.8 6 0.8
Serum retinol, μmol/L 1.33 6 0.3
parameters [fractional transfer coefficients; L(I,J)s, or the fraction of
retinol transferred to compartment I from compartment J each day] and
their statistical uncertainty.
To derive other kinetic parameters, we estimated the size of the
plasma retinol pool for each child:
Retinol pool size ¼ retinol concentration ðmmol=LÞ
ð6Þ
3 estimated plasma volume ðLÞ
Then, the geometric mean plasma pool size (micromoles) was used in a
steady-state solution in WinSAAM to predict the other compartment
pool sizes [M(I); micromoles], including VA TBS and the disposal rate
(micromoles per day), defined as the rate of VA loss from compartment 5.
Other kinetic parameters were calculated using a steady-state solution in
WinSAAM (see Results). For additional details related to compartmental
modeling and kinetic parameters, see Cifelli et al. (21).
Estimation of VA TBS by retinol isotope dilution. For comparison
with the modelÕs prediction of VA stores, we also calculated TBS for
selected individual subjects using the retinol isotope dilution equation
presented by Green (33), shown here as Equation 7:


TBS ðmmolÞ ¼ Fa 3 S 3 1 SAp ð7Þ
where Fa is the FD absorbed and found in the VA storage compartment
at time t; S is the ratio of the specific activity of retinol in plasma to that
in the storage compartment at time t; and SAp is retinol specific activity
(tracer/total tracee) in plasma, calculated as {([13C]retinol/dose)/([13C]
retinol + [12C]retinol)} at time t; the dose is micromoles. We simulated
values for the time-variant coefficients Fa and S using the super-child model
at 3, 5, and 6 d; 1/SAp was calculated for each child from observed data at
those times. Then, population estimates for Fa and S, along with each childÕs
1/SAp value, were applied to Equation 7 to predict individual VA TBS.
Statistical analyses. Descriptive statistics for general characteristics of
the study population were determined using NCSS statistical soft-
ware. Values presented for subject characteristics, dietary analyses,
and TBS predictions by isotope dilution are presented as means 6 SDs.
For modeling, we evaluated model complexity (1 compared with 2
extravascular compartments) using an F statistic (34); model com-
plexity was increased only when it resulted in a significant improve-
ment in the weighted sum of squares. P < 0.05 was considered
significant except when testing goodness-of-fit of data to the model
through the use of the Wald-Wolfowitz runs test, for which P > 0.05
indicates a good fit.
Results
Subjects. General characteristics of the 15 enrolled children are
shown in Table 1; none were anemic or undernourished, based on
hemoglobin concentration and anthropometric assessment, respec-
tively. Serum retinol concentrations were normal (1.3 mmol/L), and
we found no significant differences in serum retinol concentrations
between the 2 samples obtained from each child (P = 0.20; data
not shown). Serum concentrations of carotenoids are shown in
Supplemental Table 1. The mean estimated plasma volume for
participating children, which we used to analyze bioefficacy and
kinetics, was 5.6% 6 0.09% of body weight.
When we assessed dietary VA intake based on records of
foods consumed at the day care centers plus reports from the
FFQs and compared those estimates to the RDA for VA for
intakes were adequate. Institutional foods, which generally
accounted for two-thirds of the childrenÕs food intake 5 d/wk,
provided 62% of the RDA (125 6 116 mg RAE/d). Using results
from the FFQs completed by the childrenÕs parents, we estimated
that a median 665 mg RAE/d were provided outside of the day
Downloaded from https://academic.oup.com/jn/article-abstract/147/12/2356/4727994 by guest on 28 September 2019
care centers for the 12 children whose reported intake was
<1000 mg RAE/d. Taken together, and using a weighted scale
based on the number of meals consumed at the day care centers
compared with those consumed outside, we estimate that the
mean daily VA intake in this group of children was 400 mg
RAE/d (i.e., >100% of the RDA).
Intrinsically labeled MO. Analysis of the intrinsically labeled
MO leaves showed that lutein (186.6 6 22.4 mg/g), all-trans-
b-carotene (147.4 6 8.8 mg/g), and 9-cis-b-carotene (22.5 6
1.5 mg/g) were the main carotenoids present. However, because
lutein lacks VA activity, it was not included in estimates of
b-carotene equivalence. Analysis of the leaves by LC-MS at
an m/z of 536–558 showed that unlabeled b-carotene was
detected at an m/z of 537 (M + H+) (minimum amount: <1%).
The highest abundance of enrichment was detected at an m/z of
547 (M + H+ + 10) for labeled [2H10]b-carotene (Supplemental
Figure 1), which also represented the mean of the distribution of
[2H]b-carotene isotopomers. Using a mean molecular weight of
546 (M + 10 for b-carotene), we calculated that 3.66 mmol RAE
were provided by the 1-mg dose of [2H]b-carotene.
Bioefficacy and VA equivalence. A qualitative examination
of our results showed that [2H]b-carotene from the MO leaves
was converted to [2H]retinol and that the oral reference dose of
[13C10]retinyl acetate was absorbed and converted to [13C10]
retinol in all subjects. Both forms of labeled retinol peaked in the
samples collected 9 h (0.375 d) after administration of the doses.
The relative bioefficacy of b-carotene from MO leaves,
calculated using Equation 3, was 28%. When bioefficacy was
then used to estimate the relative VA yield (Equation 4), we
found that 1.025 mmol retinol were derived from the 3.66 mmol
RAE dose of [2H]b-carotene provided in MO leaves. Using Equa-
tion 5, the VA equivalence for intrinsically labeled b-carotene
(1 mg, or 1.83 mmol) from MO leaves was 3.3:1 by weight
(i.e., 293 mg retinol were derived from 1 mg b-carotene).
Kinetic data and the super-child compartmental model for
[13C]retinol. The composite data set for geometric mean plasma
[13C]retinol FD is plotted on a semi-log plot versus time in
Figure 2. The plasma response data indicate early and rapid
appearance of retinol in plasma, peaking 9 h after dosing,
followed by rapid disappearance as retinol was distributed to
extravascular tissues. The initial rapid decline transitions after
9 h into a slower decline representing the recycling of retinol
from tissues to plasma; a final bend occurs at ;5 d. Finally, after
6 d, the data appear to reach a terminal slope, reflecting the
apparent system fractional catabolic rate (FCRs) for VA.
Based on the composite data set and in light of previously
published compartmental models of retinol metabolism in adults
Super-child model for vitamin A value and kinetics 2359
(35, 36), we developed a 5-compartment model for VA kinetics
in children (Figure 3). Note that absorption efficiency of labeled
VA was fixed at 80% (4). The model includes a delay element
(component 1) to accommodate retinoid digestion, absorption,
and chylomicron metabolism; a compartment for initial metab-
olism of chylomicron remnant VA in hepatocytes (compartment
2); the plasma retinol pool (compartment 3); and 2 extravascular
pools (compartments 4 and 5). One of these extravascular pools
(compartment 4) turns over quickly, and the other (compart-
ment 5) turns over more slowly, presumably representing VA
TBS; irreversible loss occurred from compartment 5. In contrast
to models developed for adults (35, 36), a second extravascular com-
partment (compartment 4) was needed to fit the observed plasma
data; adding this compartment significantly improved the weighted
sum of squares as determined by an F statistic (F2,5 = 5.8; P < 0.05),
and was therefore justified for this data set.
When we used nonlinear regression analysis in WinSAAM to
estimate final parameter values after initial data fitting, the value
for L(10,5) (i.e., irreversible loss from the system or FCRs; Figure
3) converged to zero because of the shallow terminal slope of the
isotope response curve between 14 and 35 d (Figure 2). To
constrain L(10,5) to a physiologically reasonable (nonzero)
value, we used the estimated dietary VA intake [U(1) = 400 mg/d,
or 1.39 mmol/d] and assumed a steady state. Specifically, because
input equals output in a steady state, VA input [dietary VA intake
(1.39 mmol/d) 3 fractional retinol absorption efficiency (0.8)]
equalled the VA disposal rate (1.11 mmol/d). We manually
adjusted L(10,5) until the model predicted this value for the
disposal rate, and we fixed that value for the final fitting process.
Overall, the model provided an adequate fit to the geometric
mean plasma data (runs test; P > 0.05), as shown by comparing
the data points with those of the model-simulated curve (solid
line) in Figure 2. Also shown in the figure are model-simulated
FIGURE 3 Proposed ‘‘super-child’’ compartmental model of retinol
kinetics in children. The model was developed for a composite data
set based on plasma [13C10]retinol fraction of dose versus time data.
Circles represent compartments, the square is a delay component,
and interconnectivities between compartments [L(I,J)s] are fractional
transfer coefficients or the fraction of retinol transferred to compart-
ment I from compartment J each day. *The site of input of the orally
administered isotope-labeled dose. DT(1), delay time in component 1;
EV, extravascular; TBS, total body stores; U(1), dietary vitamin A
intake (mmol/d).
2360 Lopez-Teros et al.
data for FD versus time in the 2 extravascular compartments. FD
in the pool that turned over faster (compartment 4; dotted line)
paralleled FD in plasma after 1 d. By contrast, FD in storage
compartment 5 (dashed line) rose rapidly as tracer was trans-
ferred from plasma to stores, plateaued after ;5 d once retinol in
plasma and stores had mixed, and then paralleled plasma after 10 d.
The final deck is presented as Supplemental WinSAAM Deck.
Model-predicted fractional transfer coefficients and the delay
time in component 1 are shown in Table 2; the mean FSD for the
5 adjustable parameters was 0.055. We used the geometric mean
value for the plasma retinol pool size [M(3); Table 3] and
calculated a steady-state solution in WinSAAM to obtain the
kinetic parameters shown in Table 3.
Downloaded from https://academic.oup.com/jn/article-abstract/147/12/2356/4727994 by guest on 28 September 2019
Estimation of VA TBS by isotope dilution. In addition to the
model prediction of TBS, stores were also estimated through the
use of retinol isotope dilution (Equation 7), where factor Fa is
equivalent to the simulated FD in storage compartment 5
(Figures 2 and 3), and S was calculated as (FD in plasma com-
partment 3/mass of retinol in plasma)/(FD in storage compart-
ment 5/mass of retinol in compartment 5). Shown in Table 4 are
the time-variant values for Fa and S, calculated with the use of
the super-child model at blood sampling times from 3 to 6 d; the
corresponding individual subject values for SAp; and predictions
of VA TBS. The composite factor Fa 3 S was 2.9 at 3 d; it was
lower at 5 and 6 d (1.2 and 0.98, respectively). Compared with
the modelÕs prediction of TBS for the group (823 mmol), TBS
estimated with the use of Equation 7 was 832 6 176 mmol for
the 7 children sampled at 3, 5, and 6 d.
Discussion
Here we use a composite super-child approach to determine the
VA equivalence of MO leaves—a potentially useful source of
provitamin A that these children readily consumed—and to add
to the limited knowledge about plasma retinol kinetics in
children. The super-child approach allowed us to obtain popu-
lation estimates for bioefficacy, VA equivalence, and retinol
kinetic parameters, including VA TBS, while minimizing the
sampling burden on each subject. Haskell et al. (20) previously
applied a super-child approach to describe plasma retinol
kinetics and assess VA TBS in Peruvian children; the advantages
of the method led us to use and extend it here.
TABLE 2 Retinol kinetic parameters for preschool-aged
Mexican children1
Parameter Value FSD
L(0,1) 0.2
L(2,1) 0.8
DT(1) 0.326
L(3,2) 10.3 0.0604
L(4,3) 59.3 0.0697
L(3,4) 2.40 0.0726
L(5,3) 25.4 0.0269
L(3,5) 0.0255 0.0473
L(10,5) 0.00135
1
Shown here are model-predicted fractional transfer coefficients (the fraction of
retinol transferred to compartment I from compartment J each day) and the DT(1) for
the composite (‘‘super-child’’) model shown in Figure 3. FSDs are shown for the
adjustable parameters. DT(1), delay time in component 1; FSD, fractional SD; L(I,J),
fractional transfer coefficient.
TABLE 3 Model-derived kinetic parameters1
Parameter
Compartment mass, μmol
M(3)
M(4)
M(5)
Disposal rate, μmol/d
Mean transit time, d
t(3)
t(4)
t(5)
Mean residence time, d
T(3,3)
T(4,3)
T(5,3)
T(SYS)
Plasma recycling number
Plasma recycling time, d
1
Values are vitamin A masses in model compartments 3, 4, and 5 (see Figure 3); mean
transit (turnover) times (the mean time that a molecule of retinol that enters
compartment I spends there during a single transit before leaving that compartment
reversibly or irreversibly); mean residence times [the mean time that a molecule of
retinol spends in compartment I, after entering the system via plasma (compartment
3), before irreversibly leaving compartment I]; plasma recycling number (the mean
number of times a molecule of retinol recycles through plasma before leaving
irreversibly); and recycling time (the time required for an average molecule of retinol
that leaves plasma to cycle back). Vitamin A disposal rate was calculated as L(10,5) 3 M
(5). For more details on calculating these kinetic parameters, see Cifelli et al. (21). M(I),
vitamin A mass in compartment I; t(I), mean transit time; T(I,J), mean residence time.
Unlike the group studied by Haskell et al. (20), our study
population was well nourished, with no evidence of VAD (based
on serum retinol concentrations >0.7 mmol/L and estimated
stores), anemia, or subclinical inflammation. Serum concentra-
tions of provitamin A carotenoids in our group were similar to
those observed in Kenyan school-age children (37) and Chinese
preschoolers (38). Our data on VA intake from meals consumed
at the day care centers, combined with reports from FFQs, show
that these children consume adequate concentrations of dietary
VA (400 mg/d, which is higher than the RDA for children of this
age). In contrast to results from the National Health and
Nutrition Survey 2012 (39), in which the prevalence of in-
adequate VA intakes among Mexican children aged 1–4 y was
8–13%, with the highest prevalence among the poorest popu-
lations, the children in our study, who were from middle-income
socioeconomic areas in Northwest Mexico, had adequate
intakes of VA.
The relative VA equivalence of provitamin A carotenoids
from various food sources has been estimated using the isotope
reference method (19) in several studies (9, 10, 40). In each case,
numerous blood samples (#20) were collected from each subject
in order to calculate AUCs (9); such frequent sampling is
generally not feasible in children. Using a super-child approach,
we collected just 2 samples from each child and pooled data
from 2 or 3 subjects at each time. When we calculated relative
bioefficacy (Equation 3) and the VA equivalence (Equation 5)
at 21 d for intrinsically labeled b-carotene from MO leaves,
bioefficacy was 28% and the VA equivalence was 3.3:1 by
weight. This means that a 7-g serving of steamed MO leaves can
provide the daily VA requirement for 1- to 3-y-old children (300
mg RAE). Our results for MO are similar to the VA equivalence
estimated for yellow maize (3.2:1) (40), golden rice (3.8:1) (11),
and spirulina (4.5:1) (10) in adults, and are higher than values
reported for spinach (21:1) or carrots (15:1) in adults (9). If our
results are confirmed in future studies, MO leaves should be
Value considered a potentially valuable source of available VA, espe-
cially because the plant grows well in many areas that have
0.870
problems with VAD.
21.5
Since this experiment was done, a new method—the plasma
823
retinol isotope ratio—has been proposed to estimate b-carotene
1.11
bioefficacy, based on the ratio of labeled b-carotene-derived
retinol to labeled retinyl acetate–derived retinol in a single blood
0.0118
sample obtained at an appropriate time after dosing (41). When
0.416
we applied that method to our data from day 21, predicted
37.3
bioefficacy was 28%, equivalent to the estimate obtained by the
AUC method. This result provides, to our knowledge, the first
0.783
confirmation of the validity of the retinol isotope ratio method
Downloaded from https://academic.oup.com/jn/article-abstract/147/12/2356/4727994 by guest on 28 September 2019
19.3
in the field. These findings may encourage other researchers to
741
apply the single-sample method to estimate relative bioefficacy.
761
Further, once bioefficacy has been determined, VA equivalence
65.3
can be calculated using Equations 4 and 5, and thus both
8.03
parameters could be estimated based on the analysis of a single
blood sample obtained at an appropriate time after dosing
(e.g., 21 d).
In addition to our results related to the VA equivalence of
carotenoids from MO leaves, we used composite data for plasma
[13C10]retinyl acetate–derived retinol and our compartmental
model to estimate a population value for VA TBS (823 mmol).
This value is similar to results reported for older Zambian
children (mean age: 6 y) who were free of severe infection
(;700 mmol) (42) and comparable to the mean value estimated
for healthy older Americans (892 mmol) (35). These results
suggest that children in the current experiment have adequate
VA status, likely due to adequate VA intake and the high-dose
VA supplement given 3 mo before the start of our study [60 mg
(210 mmol) retinyl palmitate]. We also calculated TBS using a
retinol isotope dilution equation (Equation 6) applied to data
for individuals (n = 7) sampled between 3 and 6 d, and we found
excellent agreement with the model-predicted value (mean of
832 mmol for the 7 children in the current study), further
supporting the emerging idea of applying isotope dilution earlier
than has been traditionally done (21 d after dosing in adults). If
we apply Equation 6 to data for the 2 children sampled at 14 d, a
time used by other investigators studying children (20, 42), and
use the current model-predicted value for Fa 3 S at 14 d (0.73),
TBS predictions were 696 and 624 mmol. If we calculate TBS at
14 d for these same 2 children using a modified version of the
original isotope dilution equation (43)—assuming a VA half-life
TABLE 4 Equation coefficients and predictions of individual
vitamin A TBS by retinol isotope dilution1
Time, d Fa S SAp Calculated TBS
3 0.69 4.3 0.00390 761
0.00499 594
5 0.75 1.7 0.00141 904
0.00185 689
6 0.76 1.3 0.00112 882
0.00087 1141
0.00116 852
1
Shown here are model-predicted, time-variant values for the coefficients in the
retinol isotope dilution equation (TBS = Fa 3 S 3 1/SAp) and individual values for SAp
and vitamin A TBS in the 7 children from the ‘‘super-child’’ study who were sampled 3,
5, and 6 d after administration of labeled doses. Fa, fraction of the dose absorbed and
found in stores; S, retinol specific activity in plasma versus that in stores; SAp, plasma
retinol specific activity; TBS, total body stores.
Super-child model for vitamin A value and kinetics 2361
of 513 d, based on the FCRs (0.135%/d) for the children in this
study—we obtain values of 306 and 274 mmol, respectively.
Because the same specific activities were used in the 2 equations,
the difference in these estimates is due to the differences in the
values used for the coefficients; this is important when we
compare our population estimate of TBS (823 mmol) with those
reported for Nicaraguan schoolchildren (;330 mmol) (44) and
Peruvian children (97 mmol) (20), for which a modified ver-
sion of the original retinol isotope dilution equation (43) was
applied.
When we plotted current plasma [13C10]retinol data versus
time (Figure 2), we noted several differences when the curve was
compared with those obtained for adult subjects (35, 36).
Specifically, isotope in plasma initially appeared and subse-
quently turned over more quickly than in adults. Parallel
differences have been reported when models for VA metabolism
in neonatal rats were compared with those for adult animals
(45). The sharpness of the peak, apparently due to the rapid
absorption and mobilization of retinol from hepatocytes after
chylomicron remnant clearance, was accommodated using a
delay component as the first step in the model (Figure 3), along
with a high value for L(3,2) (Table 2). In addition, data fitting
required 2 extravascular compartments: a storage compartment
that turned over slowly, similar to models developed for adults
(35, 36), and a second compartment that turned over more
quickly, as has been found in rat models (46). The recycling
number, or the mean amount of time a molecule of retinol
recycles from tissues to plasma before being irreversibly
utilized, was 65 in current subjects, substantially higher than
the number (;5) reported for adults (35, 36). Compared with
adults, the children in this study have a much higher fractional
turnover of VA from plasma to extravascular tissues (85
compared with ;7 d21 in adults) and a much faster transit time
for retinol in plasma (0.28 compared with ;4 h). In addition,
the estimated FCRs (0.135%/d) was much lower than values
reported in the literature for Peruvian children of a similar age
[2.2%/d (20)] and closer to the 0.5%/d determined for older
Zambian children (42). A mixing time of 8 d was determined
for these children, the same as the value reported for Peruvian
children (20).
In conclusion, our results indicate that the super-child
approach can be successfully used to provide population es-
timates of the relative bioefficacy of carotenoids and the VA
value of provitamin A food sources; when combined with
model-based compartmental analysis, one can also obtain
estimates of VA stores and kinetic parameters describing retinol
metabolism. Our data show that some interesting differences
may exist in retinol kinetics in children compared with adults,
and that MO leaves may be a good source of available VA for
ameliorating VAD.
Acknowledgments
We thank Karla Marı́a Olivas Matas, Bertha Isabel Pacheco
Moreno, Orlando Tortoledo Ortiz, Mónica Guadalupe Soto
Bernal, Brianda Isamar Monreal Barraza, and Lucı́a Guadalupe
Duarte Ramı́rez for technical assistance and Joanne Balmer
Green (Penn State University) for assistance in revising the
manuscript. The authorsÕ responsibilities were as follows—
VL-T, HA-G, and MHG: designed the study; MP-C: did the
field work; GT and MAG: provided essential materials; VL-T,
MHG, JLF, and TM: analyzed the data; VL-T, LQ-C,
JLF, MHG, and HA-G: wrote the manuscript; VL-T and HA-G:
took primary responsibility for the final content; and all authors:
read and approved the final manuscript.
2362 Lopez-Teros et al.
References
1. WHO. Global prevalence of vitamin A deficiency in populations at risk
1995–2005. WHO global database on vitamin A deficiency. Geneva
(Switzerland): WHO; 2009.
2. Villalpando S, de la Cruz V, Shamah T, Rebollar R, Contreras-
Manzano A. Nutritional status of iron, vitamin B12, folate, retinol and
anemia in children 1 to 11 years old. Results of the Ensanut 2012. Salud
Publica Mex 2015;57:372–84.
3. Lopez-Teros V, Quihui-Cota L, Méndez-Estrada RO, Grijalva-Haro MI,
Esparza-Romero J, Valencia ME, Green MH, Tang G, Pacheco-
Moreno BI, Tortoledo-Ortiz O, et al. Vitamin A-fortified milk increases
total body vitamin A stores in Mexican preschoolers. J Nutr
2013;143:221–6.
4. Tanumihardjo SA, Russell RM, Stephensen CB, Gannon BM, Craft NE,
Haskell MJ, Lietz G, Schulze K, Raiten DJ. Biomarkers of nutrition for
Downloaded from https://academic.oup.com/jn/article-abstract/147/12/2356/4727994 by guest on 28 September 2019
development (BOND) - vitamin A review. J Nutr 2016;146:1816S–48S.
5. Rothman KJ, Moore LL, Singer MR, Nguyen U-SDT, Mannino S,
Milunsky A. Teratogenicity of high vitamin A intake. N Engl J Med
1995;333:1369–73.
6. Penniston KL, Tanumihardjo SA. The acute and chronic toxic effects of
vitamin A. Am J Clin Nutr 2006;83:191–201.
7. Underwood BA, Arthur P. The contribution of vitamin A to public
health. FASEB J 1996;10:1040–8.
8. West CE, Eilander A, Van Lieshout M. Consequences of revised esti-
mates of carotenoid bioefficacy for dietary control of vitamin A defi-
ciency in developing countries. J Nutr 2002;132(9 Suppl):2920S–6S.
9. Tang G, Qin J, Dolnikowski GG, Russell RM, Grusak MG. Spinach or
carrot can supply significant amounts of vitamin A as assessed by
feeding with intrinsically deuterium-labeled vegetables. Am J Clin Nutr
2005;82:821–8.
10. Wang J, Wang Y, Wang Z, Li L, Qin J, Lai W, Fu Y, Suter PM,
Russell RM, Grusak MA, et al. Vitamin A equivalence of spirulina
b-carotene in Chinese adults as assessed by using a stable-isotope
reference method. Am J Clin Nutr 2008;87:1730–7.
11. Tang G, Qin J, Dolnikowski GG, Russell RM, Grusak MA. Golden rice
is an effective source of vitamin A. Am J Clin Nutr 2009;89:1776–83.
12. Kurilich AC, Britz SJ, Clevidence BA, Novotny JA. Isotopic labeling and
LC-APCI-MS quantification for investigating absorption of carotenoids
and phylloquinone from kale (Brassica oleracea). J Agric Food Chem
2003;51:4877–83.
13. Food and Nutrition Board, Institute of Medicine. Dietary reference in-
takes for vitamin A, vitamin K, arsenic, boron, chromium, copper, io-
dine, iron, manganese, molybdenum, nickel, silicon, vanadium and zinc.
Washington (DC): National Academies Press; 2001.
14. Furr HC, Clark RM. Intestinal absorption and tissue distribution of
carotenoids. J Nutr Biochem 1997;8:364–77.
15. de Pee S, West CE, Permaesih D, Martuti S, Muhilal, Hautvast JH.
Orange fruit is more effective than are dark-green, leafy vegetables in
increasing serum concentrations of retinol and beta-carotene in school-
children in Indonesia. Am J Clin Nutr 1998;68:1058–67.
16. González B. La increı́ble Moringa. Expreso.com.mx [Internet]. 2006
Oct 2 [cited 2008 May 5]. Available from: http://edicionimpresa.expreso.
com.mx/edicion_impresa/20061002/2/pagina2.jpg?r=0.07279418324956544.
17. Luqman S, Srivastava S, Kumar R. Maurya AK, Chanda D. Experi-
mental assessment of Moringa oleifera leaf and fruit for its antistress,
antioxidant, and scavenging potential using in vitro and in vivo assays.
Evid Based Complement Alternat Med 2012;2012:519084.
18. Olson ME, Sankaran RP, Fahey JW, Grusak MA, Odee D, Nouman W.
Leaf protein and mineral concentrations across the ‘‘miracle tree’’ genus
Moringa. PLoS One 2016;11:e0159782.
19. Tang G, Qin J, Dolnikowski GG, Russell RM. Short-term (intestinal)
and long-term (postintestinal) conversion of beta-carotene to retinol in
adults as assessed by a stable-isotope reference method. Am J Clin Nutr
2003;78:259–66.
20. Haskell MJ, Lembcke JL, Salazar M, Green MH, Peerson JM,
Brown KH. Population-based plasma kinetics of an oral dose of [2H4]
retinyl acetate among preschool-aged, Peruvian children. Am J Clin
Nutr 2003;77:681–6.
21. Cifelli CJ, Green MH, Green JB. Use of model-based compartmental
analysis to study vitamin A kinetics and metabolism. Vitam Horm
2007;75:161–95.
22. Lienau A, Glaser T, Tang G, Dolnikowski GG, Grusak MA, Albert K.
Bioavailability of lutein in humans from intrinsically labeled vegetables
determined by LC-APCI-MS. J Nutr Biochem 2003;14:663–70.
23. Riso P, Porrini M. Determination of carotenoids in vegetable foods and
plasma. Int J Vitam Nutr Res 1997;67:47–54.
24. Yeum KJ, Booth S, Sadowski J, Liu C, Tang G, Krinski N, Russell R.
Human plasma carotenoid response to the ingestion of controlled diets
high in fruits and vegetables. Am J Clin Nutr 1996;64:594–602.
25. Tietz NW. Clinical guide to laboratory tests. 3rd ed. Philadelphia: W.B.
Saunders; 1995.
26. WHO/CDC . Worldwide prevalence of anaemia 1993–2005: WHO
global database on anaemia. Geneva (Switzerland); WHO; 2008.
27. Astiazaran-Garcı́a H, Lopez-Teros V, Valencia M, Vázquez-Ortiz F,
Sotelo-Cruz N, Quihui-Cota L. Giardia lamblia infection and its im-
plications for vitamin A liver stores in school children. Ann Nutr Metab
2010;57:228–33.
28. Tanumihardjo SA, Permaesih D, Muherdiyantiningsih, Rustan E,
Rusmil K, Fatah AC, Wilbur S, Muhilal, Karyadi D, Olson JA. Vitamin
A status of Indonesian children infected with Ascaris lumbricoides after
dosing with vitamin A supplements and albendazole. J Nutr 1996;126:
451–7.
29. Ortega M, Quizán P, Morales G. [Dietary intake calculation and ade-
quacy percentages from 24 hour recall and food frequency question-
naire. Food intake estimation]. Hermosillo, Sonora (Mexico): Centro de
investigación en alimentación y desarrollo; 1999.
30. Tang G, Qin J, Dolnikowski GG. Deuterium enrichment of retinol in
humans determined by gas chromatography electron capture negative
chemical ionization mass spectrometry. J Nutr Biochem 1998;9:408–14.
31. Linderkamp O, Versmold HT, Riegel KP, Betke K. Estimation and
prediction of blood volume in infants and children. Eur J Pediatr
1977;125:227–34.
32. Wastney ME, Patterson BH, Linares OA, Greif PC, Boston RC. Inves-
tigating biological systems using modeling. San Diego (CA): Academic
Press; 1999.
33. Green MH. Evaluation of the ‘‘Olson equation,’’ an isotope dilution
method for estimating vitamin A stores. Int J Vitam Nutr Res 2014;84
(Suppl 1):9–15.
34. Landaw EM, DiStefano JJ. Multiexponential, multicompartmental, and
noncompartmental modeling. II. Data analysis and statistical consid-
erations. Am J Physiol 1984;246:R665–77.
35. Cifelli CJ, Green JB, Wang Z, Yin S, Russell RM, Tang G, Green MH.
Kinetic analysis shows that vitamin A disposal rate in humans is posi-
tively correlated with vitamin A stores. J Nutr 2008;138:971–7.
36. Green MH, Ford JL, Oxley A, Green JB, Park H, Berry P, Boddy AV,
Lietz G. Plasma retinol kinetics and b-carotene bioefficacy are quanti-
fied by model-based compartmental analysis in healthy young adults
with low vitamin A stores. J Nutr 2016;146:2129–36.
37. Talsma EF, Brouwer ID, Verhoef H, Mbera GN, Mwangi AM,
Demir AY, Maziya-Dixon B, Boy E, Zimmermann MB, Melse-
Boonstra A. Biofortified yellow cassava and vitamin A status of Kenyan
children: a randomized controlled trial. Am J Clin Nutr 2016;103:
258–67.
38. Tang G, Gu X, Hu S, Xu Q, Qin J, Dolnikowski GG, Fjeld CR, Gao X,
Russell RM, Yin S. Green and yellow vegetables can maintain body stores
of vitamin A in Chinese children. Am J Clin Nutr 1999;70:1069–76.
39. Pedroza-Tobı́as A, Hernández-Barrera L, López-Olmedo N, Garcı́a-
Guerra A, Rodrı́guez-Ramı́rez S, Ramı́rez-Silva I, Villalpando S,
Carriquiry A, Rivera JA. Usual vitamin intakes by Mexican popula-
tions. J Nutr 2016;146:1866S–73S.
Downloaded from https://academic.oup.com/jn/article-abstract/147/12/2356/4727994 by guest on 28 September 2019
40. Muzhingi T, Gadaga TH, Siwela AH, Grusak MA, Russell RM, Tang G.
Yellow maize with high b-carotene is an effective source of vitamin A in
healthy Zimbabwean men. Am J Clin Nutr 2011;94:510–9.
41. Ford JL, Green JB, Lietz G, Oxley A, Green MH. A simple plasma
retinol isotope ratio method for estimating b-carotene relative bio-
efficacy in humans: validation with the use of model-based compart-
mental analysis. J Nutr 2017;147:1806–14.
42. Gannon B, Kaliwile C, Arscott SA, Schmaelsle S, Chileshe J,
Kalungwana N, Mosonda M, Pixley K, Masi C, Tanumihardjo SA.
Biofortified orange maize is as efficacious as a vitamin A supplement in
Zambian children even in the presence of high liver reserves of vitamin
A: a community-based, randomized placebo-controlled trial. Am J Clin
Nutr 2014;100:1541–50.
43. Furr HC, Amedee-Manesme O, Clifford AJ, Bergen H III, Jones AD,
Anderson DP, Olson JA. Vitamin A concentration in liver determined by
isotope dilution assay with tetradeuterated vitamin A and by biopsy in
generally healthy adult humans. Am J Clin Nutr 1989;49:713–6.
44. Ribaya-Mercado JD, Solomons NW, Medrano Y, Bulux J,
Dolnikowski GG, Russell RM, Wallace CB. Use of the deuterated-
retinol-dilution technique to monitor the vitamin A status of Nicar-
aguan schoolchildren 1 y after initiation of the Nicaraguan national
program of sugar fortification with vitamin A. Am J Clin Nutr
2004;80:1291–8.
45. Tan L, Green MH, Ross AC. Vitamin A kinetics in neonatal rats vs.
adult rats: comparisons from model-based compartmental analysis.
J Nutr 2015;145:403–10.
46. Green MH, Green JB, Lewis KC. Variation in retinol utilization rate
with vitamin A status in the rat. J Nutr 1987;117:694–703.
Super-child model for vitamin A value and kinetics 2363