Clinical Science (1997) 93,585-591 (Printed in Great Britain)
585
Absorption of all-trans and 9-cis j-carotene in human ileostomy
volunteers
Richard M. FAULKS, David J. HART, Peter D. G. WILSON, K. John SCOT and Susan SOUTHON
Institute of Food Research, Nomich Research Park, Colney. Nomich NR4 7UA. U.K.
(Received 19 May/9 July 1997; accepted 24 JuC 1997)
1. Mass balance studies were carried out in fasted
ileostomy subjects (n = 5 ) given an oral physiological dose (10 mg) of /?-carotene [all-trans: g-cis,
8416 (w/w)] dispersed in vegetable oil. Blood and
ileal effluent samples were collected and analysed
for /?-carotene.
2. Results showed that 90% (range 97.0-74.3%) of
the total /?-carotene was absorbed without measurable perturbation of plasma total /?-carotene concentration, or change in the all-trans: 9-cis
/?-carotene ratio. Peak loss of /?-carotene in ileal
effluent occurred at 4.9 h (range 2.9-8.4 h) postingestion, and no further loss was detected after
5.4-12.4 h, depending upon the individual. Comparison of the ratio of all trans-/?-carotene to 9 cis/?-carotene in the test meal and effluent indicated
that isomerization did not occur during passage
through the gastrointestinal tract and that both isomers were similarly absorbed. However, the alltruns:9-cis /?-carotene ratio of the plasma did not
change. Reasoned assumptions allowed the construction of a mathematical model of plasma /?-carotene disposal.
3. It is concluded that physiological doses of isolated all-trans and 9-cis /?-carotene are well absorbed without necessarily causing detectable
excursions in plasma /?-carotene concentrations, o r
altering the ratio of all-trans to 9-cis /?-carotene. Isomerization of /?-carotene does not occur during passage through the gastrointestinal tract. Absorbed
/?-carotene is rapidly cleared from the plasma to a n
unobservable pool at a rate similar to that of chylomicron triacylglycerol.
INTRODUCTION
There is little doubt that the consumption of
increased amounts of vegetables and fruits has a
positive health benefit [11, and issues relating to the
potentially ‘bioactive’ components of these foods are
being debated and explored extensively. The carotenoids are present in relatively large amounts in
these foods and it has been postulated that their
antioxidant capability [2] and ability to enhance the
expression of immunologically important cell surface
molecules [3] may provide mechanisms for the protective role of vegetables and fruits against the initiation and progression of chronic disorders such as
cancer and vascular disease. However, there is
controversy as to the potential adverse effects of
high-dose, long-term consumption of synthetic
preparations of carotenoids in general, and /?-carotene in particular [4].The indications, at present,
are that /?-carotene is most protective when consumed as carotenoid-rich foods. Reasons for this
could include the fact that the consumption of these
components in the form of foods limits the possibility of consuming potentially adverse amounts; the
total diet is more likely to provide balanced intakes
of a range of biologically active components, and the
total dietary mix of such components may potentiate
the action of individual compounds.
In recognition of the strong association between
vegetable and fruit intake and reduced risk of
chronic disease, advice aimed at increasing consumption to at least five portions per day (400 g/day,
excluding potatoes) is being disseminated widely.
However, the absorption and transport processes of
many of the potentially bioactive components of
these foods, including /?-carotene, are complex and
poorly understood, and appropriate experimental
protocols for quantifying bioavailability in humans
have not been determined.
It is known that the lipophilic carotenoids are
taken up into the enterocytes with dietary lipids
from mixed micelles formed during digestion, and
that they initially appear predominantly in the chylomicrons of the thoracic duct [5]. Studies with
/?-carotene, or foods containing /?-carotene, elicit
plasma excursions that are very variable in magnitude and duration, and range from ‘non-responders’
[6] to elevated plasma levels, which are still detectable 10-20 days post-dose [6, 71. The wide variation
in individual response is difficult to interpret in
terms of mechanisms of absorption and disposal.
It has been suggested that the carotenoids may be
taken up by the enterocyte and slowly released into
Key words: absorption, all trons-,%carotene,fi-carotene, 9-cis-/?-carotene,computer model, disposal, human, ileostomy, kinetics.
Abbreviations: CV,coefficient of variation.
Correspondence: Dr R M. Faulb.
586
R. M. Faulks et al.
the plasma [6], although there is no evidence in the
literature indicating temporary storage of carotenoids in the enterocyte, or the presence of the protein or lipid structures which would be necessary.
On the other hand it has been suggested that small
changes in plasma p-carotene may not be due to
poor absorption but to the sloughing of enterocytes
before the carotenoid has been transferred to the
serosal side [5].
Measurement of the absorption of p-carotene is
further complicated by the possible isomerization
between 9 4 s p-carotene and all-trans p-carotene
and at least partial conversion of both isomers into
retinol in the enterocyte [8]. Although conversion
into retinol during absorption will reduce the
amount of p-carotene entering the plasma pool, the
proportion converted is likely to be variable and the
plasma retinol concentration appears to remain
almost constant over a wide range of plasma p-carotene concentrations [9].
Simple approaches for the measurement of
absorption, disposal and distribution kinetics such as
the area under the curve in plasma, or plasma lipoprotein fractions, may only be useful for comparative purposes, unless there is some independent
measure of the total mass of carotenoid absorbed.
Disposal kinetics obtained from intravenous administration of all-trans p-carotene in lipid structures
must also be approached with care, since the carotenoid is not cleaved in the enterocyte and is not
contained in a natural chylomicron, the structure of
which influences the rate of hydrolysis of triacylglycerol and absorption in the extrahepatic capillary
bed [10,11].
Because the plasma response to oral /?-carotene in
individuals appears to be very variable, and there is
little information on the kinetics or mechanism of
clearance from the plasma pool, it is not possible to
determine how much is absorbed and how it may
alter plasma /?-carotene concentration.
The purpose of this study was to characterize the
absorption and disposal of isolated p-carotene using
ileostomy subjects eating a normal diet, thus avoiding any confounding influence of large bowel fermentation, or biased data due to the avoidance of
dietary carotenoids.
SUBJECTSAND METHODS
/?-Carotene
p-Carotene was isolated from Dunaliella salina.
For the purposes of characterizing p-carotene distribution and metabolism in vivo, the carotenoid was
randomly labelled with 13C, by growing Dunaliella
salinn in a 5 litre fermenter culture and dosing with
[13C]sodium bicarbonate [12]. The results described
here are limited to consideration of p-carotene
absorption and disposal.
The algae were harvested by vacuum filtration
through a filter bed (Celite 545) and the carotenoids
were extracted by thorough washing with acetone.
The carotenoid solution was evaporated to dryness
at 35°C under reduced pressure, made up in
dichloromethane (50 ml) and saponified for 1h at
room temperature by the addition of methanolic
KOH (50 ml, 100 g/l). Petroleum ether (100 ml) was
added to the saponified mixture, which was
thoroughly washed with water to remove all traces
of KOH before drying at 35°C under reduced pressure. The resulting p-carotene was a mixture of alltrans p-carotene and 9 4 s p-carotene in a molar
ratio of approximately 1:l. The mixed isomers were
redissolved in a minimum quantity of hexane and
cooled to -20°C for 2-3 h to crystallize the all-trans
/?-carotene. The crystals were removed by filtration
and washed with a small amount of ice-cold hexane
[13]. The dry yield of p-carotene was approximately
0.15 mmol and contained 84% all-trans p-carotene
and 16% 9-cis p-carotene by HPLC. Several batches
of p-carotene were produced, ranging from 4 0 4 0 %
atom percent 13C, as measured by low-resolution
electron ionization MS (MS890 with DS-90 data
system, Kratos, Manchester, U.K.). The chromatographic behaviour (HPLC) of the labelled carotenoids was identical with the natural abundance
standard.
Ethics
The study protocol was approved by the Norwich
District Ethics Committee and the experimental
work was carried out in accordance with the Declaration of Helsinki (1989).
Volunteers
Adult ileostomy volunteers (three male, two
female) gave informed written consent. All were
free of intestinal disease, took no antibiotics during
the study and had minimal loss of small intestine
following surgery for ulcerative colitis. Volunteers
were aged 49.9 (SD 18.8) years, weighed 74.18 (SD
19.8) kg and had a body mass index of 24.98 (SD
3.78) kg/m2.
Study protocol
Volunteers avoided gross consumption of food
items known to contain large amounts of carotenoids (a list was provided) but otherwise ate and
recorded their ‘habitual’ diet for the 3 days before
the test day and for the two following days. Volunteers fasted from 19.00 hours the day before the test
day but were free to drink water as needed. At 07.00
hours on the test day (day 0) they provided the first
blood sample (fasted; t = 0), emptied or changed the
stoma1 effluent collection bag and then consumed a
breakfast of milk shake, made by dispersing a known
Absorption of /?-carotene in human ileostomy
weight (10 mg) of 13C-labelled p-carotene in sunflower oil (10ml) and homogenizing in 200ml of
skimmed milk with a proprietary brand of milk
shake powder. Milk shakes contained a total of
9.9 pg of retinol (as esters) and 286pg all-transp-carotene. volunteers were provided with selfselected carotenoid-free midday (12.00-12.30 hours)
and evening meals (17.30-18.00 hours) and light
refreshments. Venous blood samples (25 ml) were
collected via antecubital cannula into 10 ml lithium
heparin tubes (Sarstedt, Leicester, U.K.) every 2 h
up to 12 h. Fasting samples were collected by venepuncture at 0, 24, 48 and 72 h. Plasma was separated
by centrifugation (5 min at 5000g) and frozen on
solid carbon dioxide. Stomal effluents were collected
into weighed bags every 2 h up to 12 h and then,
when the subjects returned home, at convenient
time points (recorded) up to 24 h. All effluent collections were frozen immediately on solid carbon
dioxide and weighed.
507
described by a model consisting of two pools, one
representing the plasma and a second unobservable
pool, as shown in Fig. l(a).
The steady-state solution of this linear model
shows that the ratio of the pool sizes is equal to the
ratio of the rate constants. For the subjects in this
study, the total mean plasma triacylglycerol, calculated from total plasma volume [17], was 3.0g,
and mean total body fat, calculated from anthropometric data [18], was 18.8 kg. For triacylglycerols,
the clearance rate constant, kl, must therefore be of
the order of 6300 times larger than the return rate
constant, k2. Under the assumption that p-carotene
kinetics follow those of triacylglycerols, and for a
small perturbation of the total body carotenoid
inventory, the system will effectively display firstorder kinetics for clearance of fi-carotene.
Analytical methods
Plasma samples (500 pl) were treated with SDS
(0.5 ml, 10 mmol/l) and ethanol (1 ml) to precipitate
plasma proteins. The carotenoids and retinol were
extracted twice by the addition of hexane (2 ml), and
the pooled hexane fraction was dried with a stream
of nitrogen gas. The dry residue was dissolved in
dichloromethane (100 pl) before adding 400 p1 of
acetonitrile/methanol (79:21). The carotenoids were
measured by HPLC [14]. Retinol was quantified by
the same procedure but with monitoring of the column effluent at 326 nm (retinol). The internal
standard, tocopherol acetate, was monitored at
297 nm.
Stomal effluent was subjected to solvent extraction [15], the extract was dried under reduced pressure at 35"C, made up in dichloromethane (25ml)
and saponified for 1h at room temperature by the
addition of methanolic KOH (25 ml). The unsaponified fraction was extracted by the addition of petroleum ether (50ml) and washed free of excess KOH
with water. The organic phase containing the carotenoids and unsaponified material (sterols) was dried as
before, made up to a known volume and analysed
for carotenoids by HPLC [14].
1
t, = 4 min
I
t, = 3 min
:t,
0
2
4
= 0.5jmin
6
I
8
1
0
Time (h)
Kinetic model
To investigate the expected appearance of p-carotene in the plasma, a mathematical model was constructed with the following assumptions: (1)
p-carotene is cleared from the plasma in parallel
with chylomicron triacylglycerol [161; (2) p-carotene
appears in the plasma pool at a constant rate over
95% of the period of excretion of the unabsorbed
oral dose; and (3) triacylglycerol (and thus p-carotene) kinetics can, as a first approximation, be
Fig. I.Model for absorption of /&carotene. (a) Simple two-compment
model of the absorption and disposal of oral 8-carotene. F is the flux from
the gut into the plasma pol, kl and kz are the rate constants describingthe
fluxes to and from the unobservable pool. (b) Rate of absorption of oral
/?-caroteneassuming the whole oral dose (10 mg) is absorbed linearly over
10 h, i.e. 1 mglh. (c) Cumulativeabsorption of oral dose of 8-carotene predicted from (b). (d) Predicted excursions of plasma /?-carotene(assuming no
conversion into retinol) calculated for a range of plasma pool half-lives (ti).
The horizontal broken line indicatesthe excursion needed to achieve a 95%
confidence level that the analysed plasma values are significantly different to
the fasting concentration.
1
2
R. M. Faulks et al.
5aa
Figure 1 shows the solution to this model for a
constant absorption rate of p-carotene, F , of 1 mg/h
over a 10 h period (Fig. lb). This is of the same
order as expected in this study. The cumulative
p-carotene absorbed is shown in Fig. l(c), and the
predicted plasma p-carotene excursion, P , above
basal levels is shown in Fig. l(d) for a range of
values for the kinetic constant, around the values
observed for chylomicron triacylglycerol turnover
[19, 201. As the half-life of the pool increases, the
expected plasma /I-carotene excursion also increases,
reaching a steady state within a few minutes. This
steady state plasma p-carotene excursion, AP, is
related to the absorption rate, F, and the clearance
constant kl, by: AP = F/kl.
Thus, for a given plasma p-carotene excursion and
absorption rate, the half-life of the absorbed p-carotene can be estimated. For the absorption of p-carotene to be observable in plasma samples against a
background of analytical error, the plasma excursion, AP, needs to be greater than twice the coefficient of variation (CV) of the measurement error,
represented by the broken horizontal line at 0.05 mg
in Fig. l(d). We may thus calculate the maximum
half-life [t;= In (2)lklI of plasma p-carotene for its
plasma excursion during the absorptive phase to be
within the measurement error, and thus experimentally unobservable. In the example shown therefore,
an unobservable plasma response would indicate a
half-life of less than 2 min.
This maximum half-life is given by:
t; =
In (2) x 2 x CV x absorption period (h)
mass absorbed as p-carotene
The half-life for each subject was calculated using
the assumption that the absorption time can be
approximated by the time for 95% of the unabsorbed p-carotene to appear in the ileostomy effluent
and for a range of conversion of p-carotene into
retinol and retinol esters. The absorption times are
determined for each subject by fitting a cumulative
Gaussian curve to p-carotene appearance in the
ileostomy effluent using a commercial curve-fitting
package (Tablecurve, Jandel Scientific, Ekrath,
Germany). The CV of the measurement is taken to
be 4.1% of the mean [14]. Group data (n = 5 ) are
presented as the mean (SD) units.
RESULTS
Mean stomal effluent production 29.5 (9.3) g/h
for all subjects (n = 5 ) (Table 1) was plotted as the
percentage cumulative mass over the 24 h total
collection period against time (h) (Fig. 2). The
regression line (R2 = 0.955) indicated that there was
a uniform output of effluent with time. Total loss of
both isomers of p-carotene in the effluent for each
volunteer was calculated from the mass of effluent
collected at each time point and the respective
p-carotene content. The percentage absorption of
p-carotene 90.0 (9.2)% (Table 1) was found by difference.
Normalized (24 h) cumulative loss of p-carotene
in stomal effluent was plotted against time for all
subjects and a Gaussian cumulative curve was fitted
using commercial curve-fitting software (TableCurve) to characterize the time course of excretion
for each subject (Table 1) and for the subjects as a
group (Fig. 3). Excretion of the all-trans p-carotene
peaked at 4.9 (2.15) h and was 95% complete at 8.34
(3.12) h, ranging between 5.4 and 12.4 h depending
upon the individual. There was no evidence of a
prolonged tailing loss of p-carotene with time.
The ratio of all-trans p-carotene to 9-cis p-carotene in the effluent for all volunteers (n = 5 ) was
compared by plotting the mass of the two isomers in
each effluent sample. The percentage of 9-cis p-carotene in the effluent samples was also plotted
against time to determine if residence time in vivo
affected cis-trans isomerization. The ratio of isomers in the oral dose and in all the effluent samples
was found to be constant and there was no effect of
time.
There was no significant excursion in plasma
p-carotene in any of the volunteers at any time
point, although one volunteer (Code No. 716) did
show a transient increase in plasma retinol at 8 h
post-oral dose.
The total plasma volume [17] and body-fat pool
size [18] were calculated for each volunteer using
Table I. Ileal effluent production rate, !-carotene absorption and excretion of oral dose and plasma 8-carotene pool size
Effluent
production
/?-carotene
excretion
/&carotene
excretion
/?-carotene
absorption
Subject No.
Wh)
[t(h) at SO%]*
[t(h) at 95%]*
(% oral dose)
Total plasma
/?-carotene
(Pol)
296
33 I
373
716
561
34
43.6
25.4
21.4
21.9
5.4
2.9
4.I
3.7
8.4
10.8
7.4
5.7
5.4
12.4
95.8
97.0
93.7
74.3
89.6
1.154
0.155
0.548
I.076
1.334
29.5 (9.3)
4.9 (2.15)
8.34(3.12)
90.0(9.2)
0.853(0.488)
~~
Mean (SD) . .
.
*t(h) at 50% and 95% are the times at which 50% and 95% of the unabsorbed oral dose has been excreted in ileal effluent (see under analytical methods).
589
Absorption of fl-carotene in human ileostomy
age, body weight and sex. The total plasma b-carotene (Table 1) was calculated from total plasma
volume and plasma b-carotene concentration at
t = 0. Calculation of the half-life for absorbed p-carotene in the plasma pool, based on reasoned
assumptions (see under methods) for a range of
conversion into retinol for each volunteer, is given
in Table 2. The data suggest tentatively that absorbed b-carotene has a mean half-life of
1.70-6.81 min, depending on the percentage conversion into retinol in these volunteers.
DISCUSSION
The linear production of ileal effluent (Fig. 2) is
consistent with gastric emptying also being linear
[21], and confirms that transit is commonly a continuous linear process with constant meal size, as was
given in this study. The mean effluent production
rate for volunteers was 29.5 (9.3) g/h. Ileal losses of
p-carotene are therefore consistently timed and not
unduly retarded or accelerated. This is important,
since the kinetics of absorption will be related to the
rate and linearity of gastric emptying if ileal absorption does not become saturated.
Greater absorption of carotenoids is observed if
fats or oils are simultaneously ingested [22-241,
there being no known active transport of carotenoids [5]. Maximum absorption would therefore
be expected if isolated carotenoids were dissolved/
dispersed in vegetable oil before ingestion.
Absorption of the oral dose (10mg) was 90
(9.2)%, range 74-97% (Table 1). In all cases there
was no indication of lipid malabsorption, as judged
by the amount of lipid present in the solvent extract
of the effluent. These amounts are generally greater
than some published values [23] for /?-carotene in
oil, where mass balance was taken as the difference
between intake and faecal excretion over a nonspecified period. A more recent study, using an
alimentary tract lavage technique, demonstrated
/3-carotene absorption in the range 13-18% without
...... ................................. ......
._..
0
..'.,....'
0
0 0
.-C0
g
,,,__
,,
d
.
100 r
0
c
0
0
0
X
0)
0)
5
x
$
100
80
80
C
3
4
-
en
60
0
=
0
I-
5
20
0
6
0
z
12
18
T i m e (h)
0
0
4
12
8
16
20
24
Time (h)
Fig. 2. Cumulative ileal effluent production expressed as a percentage of 24 h production for all subjects. r1 = 0.955.
Fig. 3. Cumulative excretion of !-carotene in ileal effluent collections
normalized to total 24 h excretion for each subject (points) together
with the fitted cumulative Gaussian model (solid line) and 95% confidence limits on the fit (broken lines)
Table 2. Half-life of plasma !-carotene in ileostomy subjects. Values refer to the half-life of plasma j3-carotene, assuming an
absorption period over which 95% of the unabsorbed oral dose is excreted in ileal effluent (see under Analytical methods).
Half-life of plasma /?-carotene
Conversion (%)*...
Subiect
295
33 I
373
716
562
Mean (SD)
...
*Assumed Dercentane conversion into retinol.
0
25
50
75
2.53
0.23
0.65
1.52
3.58
3.37
0.3 I
0.86
2.03
4.78
5.06
0.46
I.29
3.04
7.17
10.1 I
0.92
2.59
6.08
14.34
1.70(1.37)
2.27(1.83)
3.40(2.75)
6.81 (5.49)
24
590
R. M. Faulks et al.
a meal and 40-65% with 4184 kJ, 40% fat diet [25].
The lower values found in previous mass balance
studies [23] may still be an overestimation of absorption, since destruction of carotenoid by the colonic
microflora was not taken into account. On the other
hand, any carotenoid associated with sloughed
enterocytes may have led to an underestimation of
absorption efficiency. These opposing factors might
provide an explanation of the discrepancy between
results obtained in previous studies [23] and those
presented in this paper.
The mean time of peak ileal loss of p-carotene
from individual fitted cumulative Gaussian curves
was 4.9 (2.15) h, and is consistent with the known
transit of digesta through the stomach and ileum.
No further significant ileal loss of /?-carotene was
found after 5.4-12.4 h. This time-scale of transit is
consistent with the appearance of p-carotene in
chylomicrons from the thoracic duct [26,27], where
the peak concentration was found at 4-6 h and
clearance was complete by 12 h. If p-carotene was
excreted in sloughed enterocytes it would have been
found in the ileal effluent as a tail following the
excretion of any unabsorbed oral dose. No such losses were seen in any of our volunteers. Furthermore,
the solubility of p-carotene in lipid is low at around
0.1% w/w [28], and the amount of lipid in the gut
wall is probably not capable of dissolving and retaining the amount of p-carotene absorbed. Thus
absorption, as determined by mass balance in an
ileostomy model, is not confounded by absorption
and retention in enterocytes or sloughing of enterocytes.
Chylomicron triacylglycerol has a short half-life,
ranging from < 5 min in normal subjects to
> 20 min in hypertriglyceridemic subjects, and this
difference is reflected in the size of the plasma lipid
pool [19, 201. The rapid clearance of the carotenoidcarrying chylomicrons, and the absence (as in this
study) or relatively small perturbation of the plasma
carotenoid concentration sometimes seen in volunteers given large oral doses of p-carotene [6, 291,
could suggest that: (1) it is not absorbed; or (2) it is
rapidly cleared from the plasma along with the triaglycerols in the extrahepatic capillary bed; or (3)
that the carotenoids remain with the chylomicron
remnants and are cleared by the liver and not
immediately re-released in other lipoproteins. The
results presented here demonstrate that a lack of
perturbation of plasma carotenoid concentration per
se does not necessarily imply poor absorption. If the
assumptions that both the period of absorption and
plasma p-carotene excursion are within the precision
of the analytical method are correct, the small
plasma pool of p-carotene (Table l), and the efficient absorption of the relatively large oral dose,
without significant plasma concentration perturbation, indicates that the half-life of the p-carotene is
brief and is similar to that seen for chylomicron triaglycerol. The evidence therefore supports the hypothesis that absorbed p-carotene is cleared from the
plasma concurrently with chylomicron triacylglycerol. Furthermore, the model used (although it
gives a large value for ki; flux from plasma) also
explains why the plasma p-carotene concentration
does not fall rapidly to zero when absorption from
the gut stops, because of buffering by the large
unobservable pool. This model predicts therefore
that relatively large cumulative doses would be
needed to induce a significant change in fasting
plasma p-carotene concentration. The half-life of
the body store of p-carotene, as deduced from individuals supplemented with p-carotene for 42 days
and exhibiting carotenoderma [30], can be estimated
at z l 5 days. There would not appear to be a
requirement to consume carotenoid containing
foods daily or even weekly, provided the overall
intake is maintained.
Finally, it has been demonstrated in previous
studies that the ratio of 9-cis p-carotene to all-trans
a-carotene in plasma is not the same as in an oral
dose, suggesting that there is discrimination against
the absorption of 9 4 s p-carotene [31, 321, or that
the 9-cis /?-carotene is converted into the all-trans
isomer at some stage during absorption [S]. The oral
p-carotene dose given in the present study contained
16% of 9-cis p-carotene. If isomerization or selective
absorption had occurred in vivo the ratios of the cis/
trans isomers in the ileal effluent would be expected
to change, with the longest retained samples showing the greatest deviation. The ratio of 9-cis p-carotene in all the effluent samples and in the oral dose
were compared and was found to be the same in
both cases, confirming that both isomers were
equally well absorbed and that no selective losses
occurred with residence time in vivo. This would
support the contention that 9-cis p-carotene is not
discriminated against but is converted into the all
trans isomer during absorption [S].
From this study it is concluded that isolated 9-cis
p-carotene are well absorbed (90%) in the presence
of dietary fat, and there is no evidence of isomerization of the p-carotene during passage through the
stomach and ileum. No evidence was found to
support the view that the absence of plasma
response is indicative of lack of absorption, temporary storage in enterocytes or of excretion in
sloughed enterocytes. A reasoned estimate of the
half-life of plasma p-carotene would suggest that it
is cleared from the plasma in parallel with chylomicron triacylglycerols.
ACKNOWLEDGMENTS
The authors gratefully acknowledge the support
of The Ministry of Agriculture, Fisheries and Food,
Dr I. W. Fellows, DM MRCP (U.K.), Dr H. J. Kennedy, MD FRCP and Mrs J. McCulloch of the Norfolk & Nonvich Health Care NHS Trust, for their
help in recruiting volunteers, and Tracy Newman for
her nursing skills.
Absorption
of /I-carotene in human ileostomy
REFERENCES
I. Block G, Paterson B, Subar A. Fruit, vegetables and cancer prevention: a review
of the epidemiological evidence. Nutr Cancer 1992; 18: 1-29.
2. Krinsky NI. Antioxidant functions of carotenoids. Free Radicals Biol Med 1989; 7:
617-35.
3. Hughes DA, Wright AJA Finglas PM, Peerless ACJ, Bailey AL, Artley SB, Pinder
AC, Southon S. The effect of beta-carotenesupplementation on the immune
function of blood monocytes from healthy male non-smokers. J Lab Clin Med
I997; I 2 9 309- 17.
4. Heinonen OP, Albanes D. The effect of vitamin E and beta-carotene on the
incidence of lung cancer and other cancers in male smokers. N Engl J Med 1994;
3 3 0 1029-35.
5. ErdmanJW,Bierer TL, Gugger ET. Absorption and transport of carotenoids. In:
Canfield LM, Krinsky NI, Olson ]A, eds. Carotenoids in human health. Ann NY
Acad Sci 1993; 691: 76-85.
6. JohnsonEJ, Russell RM. Distribution of orally administered&carotene among
lipoproteins in healthy men. Am J Clin Nutr 1992; 5 6 128-35.
7. Kostic D, White WS, Olson ]A. Intestinalabsorption, serum clearance, and
interactions between lutein and j-carotene when administeredto human adults
in separate and combined oral doses. Am J Clin Nutr 1995; 6 2 604-10.
8. You CS, Parker RS, Goodman KJ,Swanson JE, Corso TN. Evidence of cis-trans
isomerisation of 94s-beta-caroteneduring absorption in humans. Am J Clin Nutr
1996; 6 4 177-83.
9. ConstantinoJP, Kuller LH, Begg L, Redmond CK, Bates MW. Serum level
changes after the administration of a pharmacologic dose of j-carotene. Am J
Clin Nutr 1988; 4 8 1277-83.
10. Redgrave TG, Vassiliou GG, Callow MJ. Cholesterol is necessary for
triacylglycerol-phospholipid emulsions to mimic the metabolism of lipoproteins.
Biochim Biophys Acta 1987; 921: 154-7.
I I. Mortimer B-C, Tso P, Phan CT,Beveridge DJ,Wen J Redgrave TG. Features of
cholesterol strucNre that regulate the clearance of chylomicron-like lipid
emulsions. J Lipid Res 1995; 3 6 2038-53.
12. Wilson PDG, Hilton MG, Waspe CR, Steer DC, Wilson DR Production of ''C
labelled fi-carotene from Dunaliello solino. Biotechnol Lett 1997; 1 9 401-5.
13. Mokady S, Avron M, Ben-Amotz A. Accumulation in chick livers of 9 4 s versus
all-trans-j-carotene. J Nutr 1990; I 2 0 889-92.
14. Hart DJ, Scott KJ. Development and evaluation of an HPLC method for the
analysis of carotenoids in foods, and the measurement of carotenoid content of
vegetables and fruits commonly consumed in the UK. Food Chem 1995; 5 4
101-11.
15. Bligh EG, Dyer WJ.A rapid method of total lipid extraction and purification.
Can J Biochem Biophys 1959; 37: 91 I-I 7.
16. van Vliet T, Schreurs WHP, van den Berg H. Intestinal &carotene absorption and
cleavage in men: response of 8-carotene and retinyl esters in the triglyceride-rich
59 I
lipoprotein fraction after a single oral dose of !-carotene. Am J Clin Nutr 1995;
62: 110-1 I.
17. Geigy Scientific Tables. Physical chemistry, composition of blood, hematology,
and somatometric data. Vol3. Lentner C, ed. Basle: Ciba-Geigy Ltd, 1984.
18. Watson EP, Watson ID, Batt RD. Total body water volumes for adult males and
females estimated from simple anthropometric measurements. Am J Clin Nutr
1980; 33: 27-39.
19. Grundy SM, Mok HY.Chylomicron clearance in normal and hyperlipidaemic
man. Metabolism 1976; 2 5 1225-39.
20. Cohen JC. Noakes TD, Benade AJ. Postprandial lipemia and chylomicron
clearance in athletes and in sedentary men. Am J Clin Nutr 1983; 4 9 443-7.
21. Stubbs DF. Models of gastric emptying. Gut 1977; 18: 202-7.
22. Jayarajan P, Reddy V, Mohanram M. Effect of dietary fat on absorption of
/?-carotenefrom green leafy vegetables in children. Indian J Med Res 1980; 71:
53-6.
23. Klaui H, BauernfeindJC. Carotenoids as food colors. In: Klaui H, BauemfeindJC,
eds. Carotenoids as colorants and vitamin A precursors. New Yotk Academic
Press, I98 I;47-3 I7.
24. Dimitrov VN, Meyer C. Ullrey DE, Chenoweth W, Michelakis A, Malone W,
Boone C, Fink G. Bioavailability of j-carotene in humans. Am J Clin Nutr 1988;
48: 298-304.
25. Bowen PE. Mobarhan S, Smith JC. Carotenoid absorption in humans. Methods
Enzymol 1993; 2 1 4 3-17.
26. Goodman DS, Blomstrand R, Wemer B, Huang HS, Shiratori T. The intestinal
absorption and metabolismof vitamin A and &carotene in man. J Clin Invest
1966; 45: 1615-23.
27. Blomstrand R, Werner B. Studies on the intestinal absorption of radioactive
/?-carotene and vitamin A in man. Sand J Clin Lab Invest 1967; 1 9 339-45.
28. Bore1 P, Grolier P, h a n d M. Partier A, Lafont H, Lairon D, his-Braesco V.
Carotenoids in biological emulsions: solubility, surface-to-core distribution, and
release from lipid droplets. J Lipid Res 1996; 37: 250-61.
29. Brown E, Micoai NE, Craft NE, Bieri JG, Beecher G. Edwards BK, Rose A,
Taylor PR, Smith JC. Plasma carotenoids in normal men after a single ingestion of
vegetables or purified j-carotene. Am J Clin Nutr 1989; 4 9 1258-65.
30. Micoai MS, Brown ED, Taylor PR, Wolfe E. Carotenoderma in men with
elevated carotenoid intake from foods and j-carotene supplements. Am J Clin
Nutr 1988; 48: 1061-4.
3 I. Stahl W, Schwartz W, von Larr J, Sies H. All-trans j-carotene preferentially
accumulates in human chylomicronsand very low density lipoproteins compared
with the 9 4 s geometrical isomer. J Nutr 1995; 125 2128-33.
32. Gaziano JM, Johnson EJ, Russell RM, Manson JE, Stampfer MI. Ridker PM. Frei B,
Hennekens CH, Krinsky NI. Discrimination in absorption or transport of
j-carotene isomers after oral supplementation with either all-trans or 9 cisj-carotene. Am J Clin Nutr 1995; 61: 1248-52.