Comparative bioavailability to humans of ascorbic acids alone or in Citrus extract

Abstract

This study was performed to determine whether synthetic ascorbic acid (AA) alone or in a natural citrus extract containing bioflavonoids, proteins, and carbohydrates was more bioavailable to human subjects. The effect of a single 500-mg ascorbate dose of the two forms and a placebo citrus extract on plasma ascorbate was examined in eight fasting subjects. A comparison of the areas under the plasma concentration-time curves showed that the citrus extract was 35% more absorbed than AA (p less than 0.001) and was more slowly absorbed than AA (p less than 0.001). In six ascorbate-saturated male subjects the ascorbate in the citrus extract produced a greater ascorbate excretion than AA alone in 24-h post-dose urine (p less than 0.05). Citrus extract ascorbate was less excreted than AA (p less than 0.05) in 12 nonsaturated subjects. Ascorbate in the citrus extract was found to be more bioavailable than AA alone in human subjects.

Full text

Am J Gun Nuir l988;48:60l-4. Printed in USA. © 1988 American Society for Clinical Nutrition 601
Comparative bioavailability to humans of ascorbic acid alone
or in a citrus ,2
Joe A Vinson, PhD, and Pratima Bose, PhD
ABSTRACT This study was performed to determine whether synthetic ascorbic acid (AA)
alone or in a natural citrus extract containing bioflavonoids, proteins, and carbohydrates was
more bioavailable to human subjects. The effect of a single 500-mg ascorbate dose ofthe two
forms and a placebo citrus extract on plasma ascorbate was examined in eight fasting subjects.
A comparison ofthe areas under the plasma concentration-time curves showed that the citrus
extract was 35% more absorbed than AA (p <0.001) and was more slowly absorbed than AA (p
<0.001). In six ascorbate-saturated male subjects the ascorbate in the citrus extract produced a
greater ascorbate excretion than AA alone in 24-h post-dose urine (p <0.05). Citrus extract
ascorbate was less excreted than AA (p <0.05) in 12 nonsaturated subjects. Ascorbate in the
citrus extract was found to be more bioavailable than AA alone in human subjects. Am J
Clin Nutr l988;48:601-4.
KEY WORDS Ascorbic acid, citrus extract, bioflavonoids, bioavailability, plasma ascor-
bate, urinary ascorbate
Introduction
There is now considerable general interest in vitamin
C (ascorbic acid) supplementation because of recent
popular books on the treatment ofthe common cold and
cancer by synthetic ascorbic acid (abbreviated here as
AA). Also, certain populations such as smokers and el-
derly people are known to have low tissue levels of ascor-
bate (1, 2) and might benefit from supplementation. Al-
though natural and synthetic ascorbic acids are chemi-
cally identical, citrus fruits contain bioflavonoids such
as naringin and hesperidin as well as carbohydrates and
proteins that might affect the bioavailability ofthe ascor-
bate. However, individuals who drink frozen or reconsti-
tuted fruit juices are ingesting <4 mg of total bioflavo-
noids/L offruitjuice (3), which are removed in process-
ing due to their bitter taste.
The view that bioflavonoids might act as compounds
that reduce the need for ascorbic acid and thus affect its
tissue concentration was first postulated by a French
group > 35 y ago (4). They found that the amount of
ascorbate in the organs of guinea pigs was increased by
the addition of bioflavonoids to the diet. Our group de-
termined that a citrus extract containing ascorbic acid
was more bioavailable to guinea pigs than AA alone (5).
The effect of bioflavonoids on tissue ascorbate was the
subject ofan extensive review (6).
Human studies, on the other hand, have been equivo-
cal. A Canadian group recently found that the bioavail-
ability of pure AA was slightly superior to ascorbate in
orange juice (7). A study using interluminal perfusion
of human small intestine showed no difference between
ascorbate in reconstituted orange juice and AA (8). The
objective ofthe present study was to determine the rela-
tive bioavailability ofa citrus extract to that ofAA alone
in humans.
Subjects and methods
Test solutions
The materials compared in this study were synthetic L-AA
from Fisher Chemical Company (Pittsburgh, PA) and rena-
tured vitamin C in citrus extract (CE), Citrus Fruit Media#{174},
manufactured by Grow Company (Hackensack, NJ). The pla-
cebo CE was a light brown water-soluble powder containing
0.2% AA, 19.5% citrus bioflavonoids (19.0% naringin, 0.34%
naringenin, and 0. 17% hesperidin)and a minimum of3O% car-
bohydrates and 15% protein. The CE was identical to the pla-
cebo CE except that AA was added to a final concentration of
25.1 g/100 g as measured by a high-pressure liquid chromatog-
raphy (HPLC) procedure after protein precipitation by meta-
phosphoric acid (9). The bioflavonoids were identified and
IFrom the Department of Chemistry, University of Scranton,
Scranton,PA 18510.
2Address reprint requests to JA Vinson, Associate Professor of
Chemistry, Department ofChemistry, University ofScranton, Scran-
ton,PA 18510.
Received October 6, 1986.
Accepted for publication October 1 3, 1987.
by guest on July 10, 2011www.ajcn.orgDownloaded from

602 VINSON AND BOSE
quantified by HPLC (10) and were the typical fiavanones of
citrus fruit.
Subjects and sampling
This study was approved by the Human Subjects Committee
of the University of Scranton. All subjects gave their written
informed consent before participation. Eight healthy, non-
smoking individuals(five males and three females) volunteered
for the plasma study. The subjects ranged in age from 18-41 y
with an average age of22 ± 6 y. None ofthe subjects was taking
vitamin supplements and none consumed foods high in ascor-
bic acid on the day ofthe study. Each subject reported between
0800 and 0830 after an overnight fast for a base-line blood
drawing by finger-prick sampling using an Autolet#{174}device
(Owen Mumford Ltd. Oxford, UK). This technique was shown
to give a sample ascorbate concentration equivalent to the
more common venipuncture method (1 1) at much less risk and
discomfort to the subj#{231}ct.The blood was collected in EDTA
tubes and the plasma was separated immediately after sam-
pling. In a random crossover experimental design, each subject
drank either 500 mg AA, 2 g placebo CE, or 2 g CE dissolved
in 50 mL of a solution of 278 mmol glucose/L. The placebo
CE provided 4 mg ascorbate and the CE, 502 mg. Finger-prick
samples were taken 1, 2, 3, 4, 6, and 8 h after dosing. Subjects
were not allowed to eat until after the last sampling but were
allowed to drink water as desired. A 0.1-mL aliquot of blood
plasma was immediately mixed with an equal volume of 1.56
mol metaphosphoric acid/L and frozen at -20 #{176}C.The next
day the samples were analyzed by a fluorometric method (12)
that measures unoxidized ascorbate. Bioflavonoids did not give
a fluorescence with this procedure and therefore did not inter-
fere. At weekly intervals the subjects ingested the other formu-
lations and the sampling procedure was repeated.
Six malesaged 21-41 y with an average age of26 ± 8 y vol-
unteered for the first urine study. The subjects were saturated
with AA in a protocol similar to that ofRobinson (13). After 2
wk of consuming 1 g AA to saturate the body stores, the sub-
jects avoided ascorbic acid for 2 d before the experimental day.
During the second day of this 2-d period, a 24-h urine collec-
tion was made for use as a blank. After fasting overnight the
subjects ingested in the morning 500 mg ofascorbate in one of
the two forms, AA or CE (dissolved in 50 mL of a solution
of 278 mmol glucose/L), in a random crossover experimental
design. The subjects were then allowed to eat lunch and dinner
while still avoiding foods high in ascorbic acid. A 24-h urine
collection was made. The following day the subjects resumed
taking 1 g AA for 5 d and then stopped for 2 d before ingesting
the other form. Twenty-four-hour blank and postdose urine
was collected as before. Urine was stored in plastic specimen
bottles over 50 mL of 3. 12 mol metaphosporic acid/L, kept
refrigerated between collections, and frozen at -20 #{176}Cuntil
analysis within 1 wk. The 24-h volume was recorded and an
aliquot was taken for fluorimetric analysis of total ascorbate
after reduction ofdehydroascorbic acid with mercaptoethanol.
Twelve subjects (six males and six females) volunteered for
the second urine study. They ranged in age from 18-41 y with
an average age of 22 ± 6 y. A blank 24-h urine collection was
made the day before dosing. After an overnight fast subjects
consumed either AA, placebo CE, or CE as in the plasma study.
Twenty-four-hour urine was collected as described previously.
One week later another 24-h urine blank was collected. The
other forms were given and the procedure was repeated. The
subjects avoided foods high in ascorbic acid in their diet 1 wk
before the experiment and during the week ofthe experiment.
TABLE 1
Areas under the plasma ascorbate time-concentration curves after oral
administration ofSOO mg ascorbate in the form ofsynthetic ascorbic
acid (AA) alone or in a citrus extract (arbitrary units)
Subject Age Sex
Plasma ascorbate area
AA Citrus extract
y
1
2
3
4
5
6
7
8
18
18
19
41
21
22
21
18
F
F
F
M
M
M
M
M
571
748
389
672
470
565
656
651
824
842
629
816
873
772
879
748
±SD 590±117 797±82
5Significantly different by a paired ttest(p <0.00 1).
Data analysis
Areas under the plasma ascorbate concentration-time
curves were measured for each individual by manual planime-
try with a reproducibility of 2.5%. The points for determining
the time for maximal ascorbate concentration were plotted by
a cubic spline curve-fitting program on an IBM PC(model XT)
computer (Armonk, NY).
Statistical comparisons ofthe results were made by a paired
or two-sample Student’s ttest.
Results
The initial mean ±SD fasting plasma ascorbate was
35.2 ± 13.7 imol/L (range, 14.8-69.8 mol/L). These
are typical values ofa normal population (14). There was
no significant difference between male and female sub-
jects, with levels of 35. 1 ± 14.2 and 35.2 ± 1 1 .1 imol/L,
respectively. The extent of absorption of ascorbate was
determined by comparing the areas under the plasma
ascorbate concentration-time curves after administra-
tion of each formulation. Because absorption of AA is
dose-dependent (15), comparison of AA and CE was
made at the same dose of ascorbic acid. The areas were
measured for each subject and the data are shown in
Table 1.There was no significant difference between
males and females for either AA or CE; therefore, the
results were combined. All subjects showed a greater area
under the curve for CE than for AA. The ascorbate in the
CE was found to be significantly more absorbed than the
AA (p <0.001).
The average plasma ascorbate curves for the three for-
mulations are shown in Figure 1. The maximal ascorbate
concentrations were 197 ± 20.2 mol/L for AA, 209
± 32.9 tmol/L for CE, and 56.8 ± 1 1 .9 mol/L for pia-
cebo CE. The time for maximal plasma ascorbate was
2.9±0.3hforAAand4.l±0.ShforCE(p<0.001).A
statistical evaluation of the AA and CE curves in Figure
1 indicates that there was no significant difference be-
by guest on July 10, 2011www.ajcn.orgDownloaded from

TIME (hours)
COMPARATIVE ASCORBATE BIOAVAILABILITY 603
5Significantly different by a paired Itest (p <0.05).
I
I
FIG 1.Plasma time-concentration curves for eight fasting subjects
ingesting 500 mg ofascorbate as synthetic ascorbic acid (AA) alone or
in 2 g of a citrus extract, or 2 g placebo citrus extract, or 2 g placebo
citrus extract ( ± SD). (To convert mg/dL to mol/L, multiply by
56.78.)
tween the two dose forms except at 4 h after ingestion
when the CE produced a 68% greater plasma ascorbate
than AA (p <0.005). The placebo CE produced a small
but significant increase in plasma ascorbate after 3 h
compared with the 0-h value (p <0.05).
The results ofthe first urine excretion study after a pre-
vious saturation ofbody stores with ascorbate are shown
in Table 2. All six subjects showed a greater excretion
ofascorbate after ingestion ofCE compared with AA (p
<0.05).
Results ofthe second urine study ofsubjects who were
not saturated with ascorbate are displayed in Table 3.
The placebo CE produced no net excretion of ascorbate.
There was no significant difference between the male and
female excretion data for AA or CE although males
tended to excrete more ascorbate after AA ingestion than
did females. Five of the six males and four of the six fe-
males excreted more ascorbate after ingestion ofAA than
TABLE 2
Net 24-h urinary excretion ofascorbate after oral administration of
500 mg synthetic ascorbic acid (AA) alone or in a citrus extract to AA-
saturated subjects(mg)
Subject
Urinary ascorbate
AA Citrus extract
1
2
3
4
5
6
8.0
73.4
49.6
102.1
36.8
108.4
26.4
157.7
189.9
172.1
43.2
262.1
i±SD 63±38.9 141.9±90.5w
TABLE 3
Net 24-h urinary excretion ofascorbate after oral administration of a
placebo citrus extract or 500 mg synthetic ascorbic acid (AA) alone or
in a citrus extract (mg)5
Subjects
Urinary ascorbate
pjAA Placebo
citrus extract Citrus
extract
Males(n=6)
Females(n =6)
Ailsubjects
181±89
93 ± 32
135±79
-2.3±11.4
-26.9 ± 38.7
-10.5±23.2
72±35
7 1 ± 29
72±31
<0.05
NS
<0.05
1±SD.
t Significance of difference between AA and citrus extract by a
paired I test.
after ingestion ofCE and the difference between the two
forms was significant (p <0.05).
Discussion
A comparison ofthe areas under the plasma ascorbate
concentration-time curves (Table 1) allows the determi-
nation of the relative bioavailability of the two forms.
The ascorbate in the CE was found to be 35% more bio-
available than AA alone (p <0.001). This result is sim-
ilar to the 48% greater bioavailability ofCE found in our
guinea pig study (8). These findings seem to contradict
an earlier human study by Pelletier and Keim (6) who
found that AA plus the bioflavonoid rutin produced a
slightly lower serum ascorbate increase than did AA
alone. However, that study measured ascorbate 2 h post-
dose, at which time our data showed no significant
difference between AA alone and the CE that contained
bioflavonoids (Fig 1). The area under the placebo CE
curve was 9.4% ofthat under the CE curve. This percent-
age is more than would be expected from the ascorbate
in the placebo CE, which is only 0.8% of the CE ascor-
bate. This result from a human study corroborates previ-
ous data reviewed by Hughes and Wilson (5) that showed
elevated tissue ascorbate in guinea pigs after administra-
tion of bioflavonoids.
As seen in Figure 1, the CE produced a peak plasma
ascorbate concentration at a longer time interval after
ingestion than did AA. This was not a dosage-form
difference because both CE and AA were given in solu-
tion. The result indicates that the CE was more slowly
absorbed than the AA. The time for maximum concen-
tration for AA was 2.9 h which is nearly identical to the
2.8 h found by Zeitler et al (15). The time for the CE was
4. 1 h, similar to the 4.-S h found by other workers after
ingestion oftimed-release formulations ofAA (13, 16).
To use urinary-excretion data to compare the absorp-
tion of the two forms of ascorbic acid, it is necessary to
saturate the body stores because the retention of ascor-
bate in the body is influenced by the nutritional status of
the individual (1 3). The fact that the subjects were satu-
rated is corroborated by the significantly greater blank
by guest on July 10, 2011www.ajcn.orgDownloaded from

604 VINSON AND BOSE
24-h urinary ascorbate in the saturated group than in the
unsaturated group in the second urine study (158.0
± 59.5 mg and 77.5 ± 35.2 mg, respectively [p <0.05 by
a two-sample ttest]). The excretion from the saturated
group is similar to that reported by Jones (16), 237
± 25.4 mg. The data on presaturated individuals in Ta-
ble 2confirm the plasma results that indicated that the
CE was absorbed more than the AA alone. The urine
results are similar to a previous study (6) which showed
that AA plus rutin or fresh orange juice alone produced
a greater ascorbate excretion than AA alone in presatu-
rated subjects. Another report that confirmed these re-
sults (16) showed that black currantjuice, high in biofla-
vonoids, produced a greater urinary excretion of ascor-
bate than did AA alone.
The second urine study (using nonsaturated subjects)
produced different results from the first study. The re-
sults in Table 3 demonstrate that the CE produces sig-
nificantly less ascorbate excretion than AA alone in non-
saturated subjects. In light of the plasma study and the
first urine study, which showed increased absorption for
the CE compared with AA alone, the second urine study
results may be interpreted to indicate that the ascorbate
in the CE is sequestered to a greater extent than AA alone
in the tissues ofnonsaturated subjects.
The finding that the ascorbate in CE is more bioavail-
able and more slowly absorbed than AA alone, ie, that it
acts as a timed-release formulation, is advantageous for
supplementation. The water-soluble CE does not suffer
from the disadvantages of previous timed-release forms
which were found to be less bioavailable than conven-
tional forms ofAA (17). Although the mechanism of the
greater bioavailabiity of the ascorbate in CE is not
known, previous studies with bioflavonoids produced
two possible mechanisms: enhanced absorption and sta-
bilization ofascorbate (5).
In conclusion, the ascorbate in the CE was found to be
more slowly absorbed and more bioavailable than AA
alone and is thus the preferred form ofascorbate for sup-
plementation. 13
We gratefully acknowledge the assistance ofO Pelletier in the prepa-
ration ofthis manuscript. The citrus extract was kindly supplied by R
Koctzner ofGrow Company, mc, Hackensack, NJ 07601.
References
1.Pelletier 0. Vitamin C status of smokers and non-smokers. Am J
am Nutr 1970; 23:520-4.
2. Burr ML, Elwood PC, Hole DJ, Hurley RJ, Hughes RE. Plasma
and leucocyte ascorbic acid levels in the elderly. Am J Can Nutr
1974;27: 144-51.
3. Veldhuis MK, Swift LI, Scott WC. Fully-methoxylated flavones in
florida orangejuices. J Agric Food Chem 1970; 18:590-2.
4. Cotereau H, Gabe M, Gero E, Parrot JL. Influence of vitamin C
upon the amount ofascorbic acid in the organs ofthe guinea pigs.
Nature 1948; 16 1:557-8.
5. Hughes RE, Wilson HK. Flavonoids: some physiological and nu-
tritional considerations. In: Ellis GP, West GB, eds. Progress in
medicinal chemistry. Vol 14. Elsevier/North Holland: Amster-
dam, 1977:285-300.
6. Pelletier 0, Keith MO. Bioavailability of synthetic and natural
ascorbic acid.J Am Diet Assoc 1974;64:271-5.
7. Nelson EW, Streiff RR, Cerda JJ. Comparative bioavailability of
folate and vitamin C from a synthetic and a natural source. Am J
Can Nutr 1975;28:1014-9.
8. Vinson JA, Bose P. Comparative bioavailability of synthetic
and natural vitamin C in guinea pigs. Nutr Rep Int l983;27:
875-9.
9. Ziegler SJ, Meier B, Sticher 0. Fast and selective assay of 1-ascor-
bic acid in rose hips by RP-HPLC coupled with electrochemical
and/or spectrophotometric detection. Planta Med l986;5:
383-7.
10. Park GL, Avery SM, Byers JL, Nelson DB. Identification of bio-
flavonoids from citrus. Food Technol 1983;98-l05.
1 1. Bradley DW, MaynardJE, EmergyG. Comparison ofascorbic acid
concentrations in whole blood obtained by venipuncture or fing-
erprick. Clin Chem 1972; 18:968-70.
12. Hubmann B, Monnier D, Roth M. Rapid and accurate determina-
tion of ascorbic acid in blood plasma. Gin Chim Acta l969;25:
16 1-7.
13. Yung 5, Mayersohn M, Robinson JR. Ascorbic acid absorption in
humans: a comparison among several dosage forms. JPharm Sci
1982;7 1:282-5.
14. Anonymous. Nutrition Canada Interpretive Standards. National
survey. Ottawa: Ottawa Information Canada, 1973.
15. ZeitlerG, SeidelG, Siegers CP, Iven H. Pharmacokinetics of
bic acid in man. Eur J Gin Pharmacol 1976; 10:273-82.
16. Jones E, Hughes RE. The influence ofbioflavonoidson the absorp-
tion ofvitamin C. IRCS Med Sci 1984; 12:320.
17. Leopold G, Breur J, Loebich F. Different formulations of vitamin
Cand absorption in man. IRCS J Med 5th 1975;3:74.
by guest on July 10, 2011www.ajcn.orgDownloaded from