点击显示 收起
1 From the Jean Mayer US Department of Agriculture Human Nutrition Research Center on Aging at Tufts University, Boston.
2 Any opinions, findings, conclusions, or recommendations expressed in this publication are those of the authors and do not necessarily reflect the views of the US Department of Agriculture. 3 Supported by the US Department of Agriculture, under agreement no. 581950-9-001. BASF provided the labeled ß-carotene. 4 Address reprint requests to G Tang, Jean Mayer USDA Human Nutrition Research Center on Aging at Tufts University, 711 Washington Street, Boston, MA 02111. E-mail: guangwen.tang{at}tufts.edu.
ABSTRACT
Background: Quantitative information on the conversion of ß-carotene to vitamin A in humans is limited.
Objective: We determined the short- and long-term conversion of labeled ß-carotene to vitamin A by using a stable-isotope reference method.
Design: [2H8]ß-Carotene (11 011 nmol, or 6 mg) in oil was given with a liquid diet (25% of energy from fat) to 22 adult volunteers (10 men, 12 women). Three days after the [2H8]ß-carotene dose, the volunteers each took a dose of [2H8]retinyl acetate (8915 nmol, or 3 mg) in oil with the same liquid diet. Blood samples were collected over 56 d.
Results: The 53-d area under the serum [2H4]retinol response curve (from the [2H8]ß-carotene dose) was 569 ± 385 nmol · d, and the 53-d area under the serum [2H8]retinol response curve (from the [2H8]retinyl acetate dose) was 1798 ± 1139 nmol · d. With the use of [2H8]retinyl acetate as the vitamin A reference, the [2H4]retinol formed from [2H8]ß-carotene (11 011 nmol) was calculated to be equivalent to 3413.9 ± 2298.4 nmol retinol. The conversion factor of ß-carotene to retinol varied from 2.4 to 20.2, and the average conversion factor was 9.1 to 1 by wt or 4.8 to 1 by mol. This conversion factor was positively correlated with body mass index (r = 0.57, P = 0.006). The postabsorption conversion of ß-carotene was estimated as 7.8%, 13.6%, 16.4%, and 19.0% of the total converted retinol at 6, 14, 21, or 53 d after the [2H8]ß-carotene dose, respectively.
Conclusion: The quantitative determination of the conversion of ß-carotene to vitamin A in humans can be accomplished by using a stable-isotope reference method. This approach provides in vivo metabolic information after a physiologic dose of ß-carotene.
Key Words: [2H8]ß-Carotene • [2H8]retinyl acetate • humans • retinol equivalence • stable isotope • mass spectrometry
INTRODUCTION
ß-Carotene is a major and safe vitamin A source for a vast population throughout the world. It is well known that vitamin A deficiency or excessive intakes can cause serious health problems, such as night blindness or birth defects, respectively (1,2). Investigations into the conversion of ß-carotene to vitamin A with the use of physiologic doses in vitamin A–sufficient populations have been hampered by the fact that blood concentrations of ß-carotene and of the various forms of vitamin A do not change significantly in response to a physiologic dose. Novel isotope techniques using isotope-labeled ß-carotene in well-nourished adults (3–7) and in children (8) have recently been developed. These studies have greatly advanced our knowledge of ß-carotene bioconversion to vitamin A. Variation in the bioconversion of synthetic ß-carotene to vitamin A has been observed; that is, 1 mol ß-carotene might provide 0–0.27 mol vitamin A (6–8). Further investigation is needed to understand potential factors affecting this conversion, such as dose size (pharmacologic or physiologic), dose form (eg, crystalline, solublized in oil, or beadlet), dose stability (isomerization and degradation during storage), sex, and body fat composition.
The absorption of intact, isotope-labeled ß-carotene and the conversion of labeled ß-carotene to labeled vitamin A can be followed by mass spectrometry, even though circulating retinol is homeostatically controlled in populations with adequate vitamin A nutrition. To determine the quantity of vitamin A that can be provided in vivo from a given labeled ß-carotene dose (vitamin A equivalence of ß-carotene), a known amount of labeled vitamin A must be given as a reference dose (9). We developed an atmospheric pressure chemical ionization mass spectrometry (APCI-MS) technique to measure labeled ß-carotene in human serum and a gas chromatography–electron capture negative chemical ionization mass spectrometry (GC–ECNCI-MS) technique to measure labeled retinol in human serum after a labeled ß-carotene dose (10,11). With the use of a known amount of labeled vitamin A as a reference dose and by following the retinol formed from the ß-carotene dose and the retinol formed from the labeled vitamin A dose, a quantitative determination of in vivo ß-carotene conversion to retinol is possible.
In the present study of well-nourished adults, we used stable-isotope-labeled [2H8]ß-carotene in corn oil, which was converted in vivo to [2H4]vitamin A, to evaluate ß-carotene absorption and the conversion of ß-carotene to retinol in vivo. [2H8]Vitamin A in corn oil was used as the reference dose, so that the mass spectrometer could distinguish the vitamin A reference dose from the vitamin A produced from ß-carotene.
SUBJECTS AND METHODS
Labeled compounds
Crystalline all-trans-[2H8]ß-carotene (11, 11', 19, 19, 19, 19', 19', 19'-2H8-ß-carotene; 82.0% in the all-trans form, 8.0% in the 13-cis form, 4.2% in the 9-cis form, and 3.4% in the 15-cis form) in a sealed amber ampoule was provided by BASF (Ludwigshafen, Germany). The purity of the [2H8]ß-carotene was checked by HPLC. The [2H8]ß-carotene was 97.5% spectroscopically pure, but it contained [2H7]ß-carotene (15.7%), [1H]ß-carotene (2.9%), and [2H6]ß-carotene (0.3%) as measured by APCI-MS. After 6 y of storage at -70 °C in a sealed ampoule, we found that the chemical purity of the crystalline ß-carotene was 90.6% and that both the stereoisomer and isotopic profiles showed no significant changes. The ß-carotene doses (in capsules with corn oil) were made in batches of 10 according to the following procedure: 6.000 mg ß-carotene was weighed (Mettler M3; Mettler-Toledo, Columbus, OH) from the stock ß-carotene crystals, dissolved in corn oil, and put into gelatin capsules, which were kept at 4 °C until given to each volunteer within a few months. For 8 subjects who took the doses that had been stored at -70 °C for 3–5 y, the data were corrected by a factor of 7% (97.5% - 90.6%); no correction was made for 14 subjects who took the ß-carotene dose that was stored at -70 °C for < 3 y.
[2H8]Retinyl acetate (10, 14, 19, 19, 19, 20, 20, 20-2H8-retinyl acetate) was purchased from Cambridge Isotope Laboratory (Andover, MA). The purity of the [2H8]retinyl acetate was > 98%. No significant degradation was detected after 6 y of storage, and no correction was made in our data analysis. The deuterated fraction was 99.95% (0.05% was unlabeled retinyl acetate), with 67.2% as [2H8]retinyl acetate, 14.1% as [2H7]retinyl acetate, and 2.3% as [2H6]retinyl acetate.
Subjects
Twelve postmenopausal women (aged 40–71 y) and 10 age-matched men who were healthy nonsmoking adults who had not taken vitamin A or ß-carotene supplements within the past month were recruited from the general public of the greater Boston area (identified through advertising) to participate in the study. These volunteers had body weights within 20% of the standard for their height and a body mass index (BMI; in kg/m2) 30 (12). A screening blood sample was taken to ensure that carotenoid and retinol concentrations were within the normal range. Volunteers attended a screening session during which they were instructed how to follow a diet containing low amounts of carotenoid-containing foods. They were provided with lists of foods to select and to avoid while living at home. They were also given menu suggestions. No alcohol was allowed during the study. The following situations excluded potential volunteers from the study: severe and symptomatic cardiac disease or hypertension; history of bleeding disorders; chronic history of gastric, intestinal, liver, pancreatic, or renal disease; any portion of the stomach or the intestine removed (aside from the appendix); history of intestinal obstruction or malabsorption; active smoking (smoking was stopped =" BORDER="0"> 1 mo before the beginning of the study and during the study); history of chronic alcoholism; a convulsive disorder; or an abnormality in screening blood or urine samples. Informed written consent was obtained from all volunteers under the guidelines established by the Human Investigation Review Committee of Tufts University and the Tufts–New England Medical Center.
Study design
Participants were housed in the Metabolic Research Unit of the Jean Mayer US Department of Agriculture Human Nutrition Research Center on Aging at Tufts University for a 10-d resident stay, but were then free-living from days 11–57. For the 2 wk before the resident period, the subjects were instructed to consume their normal diet, but without vitamin supplements or foods containing large amounts of ß-carotene or vitamin A. Each volunteer was required to complete a 3-d food record for assigned days for each of the 2 wk. The food records were reviewed by the dietitian at the time of admission for compliance (the results showed that intakes of carotenoids were < 1 mg/d).
On day 1, after fasting overnight, each volunteer consumed 10 gelatin capsules (#1 gelatin capsules; Gallipot, St Paul) containing a total of 6 mg [2H8]ß-carotene (11 011 nmol) and 6 g corn oil with a formulated liquid breakfast. The liquid breakfast diet (containing no fiber) was formulated to contain fat (13.5 g, or 25% of total energy), protein (17 g), and carbohydrate (78 g) to provide a total energy content of 481 kcal. The percentages of energy from saturated, monounsaturated, and polyunsaturated fatty acids were 20%, 1.4%, and 1.6%, respectively. Five hours after the breakfast, the subjects consumed the same amount of liquid diet as a lunch. In the evening, 10 h after the breakfast, the volunteers received a dinner containing 31 g fat and 35 g protein with a total energy content of 880 kcal (containing 7 µg ß-carotene). Three days after the [2H8]ß-carotene dose, the volunteers consumed a capsule containing 3.0 mg [2H8]retinyl acetate (8915 nmol) in 170 mg corn oil with the same liquid diet as consumed with the ß-carotene dose on day 1. For the first 9 d of the study, the volunteers consumed a 2-d rotation diet containing 100 µg vitamin A and 25 µg ß-carotene per day at the Nutrition Center; from day 10 to day 56 the volunteers were free-living. At the end of the resident phase, the volunteers were discharged with instructions from the dietitian to avoid large intakes of foods containing ß-carotene and vitamin A. Subjects were given a list of fruit and vegetables to avoid and a list of fruit and vegetables that have low amounts of ß-carotene. In addition, the volunteers were counseled to abstain from multivitamins, minerals, nutritional supplements, fortified cereals, and fish liver oil.
The carotenoid and vitamin A amounts used to create all food instruction sheets were derived from the US Department of Agriculture and the Nutrition Coordinating Center Carotenoid Database for US Foods-1998. Nutrient calculations were performed by using NUTRITION DATA SYSTEM FOR RESEARCH software (NDS-R version 4.02; Nutrition Coordinating Center, University of Minnesota, Minneapolis). The research dietitian followed up weekly with each volunteer by phone to check for compliance.
Serum samples were collected at 0, 3, 5, 7, 9, 11, and 13 h of the first and the fourth days of the study. Fasting serum samples were collected daily for 9 d while the subjects stayed in the Nutrition Center and weekly while they were free-living at days 15, 22, 29, 36, 43, 50, and 57 after the subjects had fasted for 12 h overnight. Blood samples were kept at room temperature for 0.5 h after drawing and were then centrifuged with Sure-sep II (Organon Teknika Corp, Durham, NC) at 4 °C and 800 x g for 15 min. Serum was stored at -70 °C until processed.
HPLC analysis of serum samples
Three milliliters of chloroform:methanol (2:1, by vol) was added to a 100-µL serum sample. The mixture was mixed by vortexing and then centrifuged for 10 min at 4 °C and at 800 x g. The chloroform layer was collected. Hexane (2 mL) was added to the aqueous layer to reextract the fat-soluble nutrients. The hexane layer was combined with the chloroform layer and evaporated under nitrogen gas on an N-EVAP (Organomation Associates Inc, South Berlin, MA). The residue was dissolved in 100 µL ethanol, and 50 µL was injected into an HPLC apparatus. Concentrations of carotenoids and retinol in a 100-µL aliquot of serum were measured with an HPLC apparatus equipped with a YMC C30 column (Waters, Milford, MA) and a Waters 994 programmable photodiode array detector with the wavelength set at 450 nm for carotenoids and 340 nm for retinoids (13). The concentrations of ß-carotene and retinol, together with the percentage isotopic enrichment of ß-carotene and retinol, determined by liquid chromatography (LC)-MS and GC-MS were used to calculate the molar enrichments of [2H8]ß-carotene, [2H4]retinol, and [2H8]retinol in the processed serum samples.
LC–APCI-MS analysis
To determine the percentage enrichment of labeled ß-carotene, an LC–APCI-MS method was used. One- to 2-mL serum samples (whole serum) were added to an NH2 column (3 mL, 500 mg; JT Baker, Phillipsburg, NJ) preconditioned with hexane. Hexane was used as the eluent. The ß-carotene in the hexane eluent was evaporated under nitrogen gas. The residue was resuspended in 70 mL methyl t-butyl ether:methanol (2:1, by vol) and injected onto an HPLC system equipped with a C30 column (SC-150; Bischoff Chromatography, Leonberg, Germany). The ß-carotene fraction from the HPLC separation was collected and dried under nitrogen gas. The purified and dried ß-carotene fraction was resuspended in ethanol and injected into an online LC–APCI-MS system (10) with a C18 Prizm column (Keystone Scientific, PA) with 95% methanol and 5% ethanol:methanol:tetrahydrafuran (75:20:5, by vol). The actual enrichment of labeled ß-carotene in human serum was determined by LC–APCI-MS at mass-to-charge ratios (m/z) (M+H)+ of 537 (1H), 538 (13C-1H), 539 (13C-13C-1H), 544 (2H7), 545 (2H8 + 13C-2H7), 546 (13C-2H8, 13C-13C-2H7), and 547 (13C-13C-2H8) with Data Analysis Esquire-LC MS Processing (version 1.6m; Bruker, Billerica, MA). The LC instrument was an Agilent 1100 (Andover, MA) and the mass spectrometer was a Bruker Daltonic Esquire LC.
GC–ECNCI-MS analysis
To determine the percentage enrichment of labeled retinol, a 200-µL (up to 400 µL for poor responders or later time points) serum sample was extracted by following the procedure described in the section "HPLC analysis of serum samples." The extract was injected into an HPLC apparatus equipped with a C18 column (Perkin-Elmer Inc, Norwalk, CT) (11,14). The retinol collected from the HPLC system was dried under nitrogen gas, and the residue was derivatized with N,O-bis(trimethylsilyl)trifluoroacetamide containing 10% trimethylchlorosilane (Pierce, Rockford, IL). The GC-MS instrument was an Agilent 5973N, and the GC column was coated with a DB-1 stationary phase from Agilent Technologies (Folsom, CA). The percentage enrichment of [2H4]retinol derived from [2H8]ß-carotene was calculated by integrating the peak areas under the reconstructed mass chromatograms of the negative ions at m/z 271 (2H3), 272 (2H4 + 13C-2H3), and 273 (13C-2H4) divided by the total area response of labeled and unlabeled retinol ions. The percentage enrichment of [2H8]retinol derived from [2H8]retinyl acetate was calculated by integrating the peak area under the reconstructed mass chromatogram of the negative ions at m/z 274 (2H6), 275 (2H7), 276 (2H8), 277 (13C-2H8), and 278 (13C2-2H8) divided by the total area response of labeled and unlabeled retinol fragment ions. The linearity of the GC-MS response and the detection limit of the GC–ECNCI-MS system were previously addressed (10,11). The analysis showed that the enrichment of all the samples analyzed until 56 d after the dose, including the samples from the lowest responder, was above the detection limit of 0.005% enrichment. The percentage enrichments measured by GC-MS and the concentration of retinol in serum were used to calculate the concentration of labeled retinol in the circulation.
Areas under the curve of labeled retinol or ß-carotene in serum
Total serum responses to the [2H8]ß-carotene doses were determined by multiplying the total serum volume (0.0435 L/kg body wt) by the concentration of [2H8]ß-carotene and [2H4]retinol, respectively, in the circulation. Similarly, total serum responses to the [2H8]retinyl acetate doses were determined by multiplying the total serum volume (0.0435 L/kg body wt) by the concentration of [2H8]retinol in the circulation. Areas under the serum labeled retinol or ß-carotene response curve (AUCs; in nmol · d) after the [2H8]ß-carotene or [2H8]retinyl acetate dose were calculated by using the curves of total serum responses versus time with INTEGRAL-CURVE of KALEIDAGRAPH (Synergy Software, Reading, PA).
Vitamin A equivalence calculations
The AUC of the serum [2H4]retinol response was compared with the AUC of the vitamin A reference dose (8915 nmol [2H8]retinyl acetate). The amount of [2H4]retinol was calculated as follows:
RESULTS
The age and BMI of the subjects and their serum concentration of retinoids, carotenoids, and tocopherols at the beginning of the study are presented in Table 1. There were no significant differences between the men (n = 10) and the women (n = 12) in age, BMI, or serum concentrations of retinol, carotenoids (ß-carotene, cryptoxanthin, and lycopene), and tocopherols (- and -tocopherol). The serum concentration of lutein, however, was higher in the men than in the women (P = 0.05). Transformed serum concentrations of lutein showed a sex difference at P = 0.07. Serum concentrations of retinoids, carotenoids, and tocopherols decreased at all time points during the first 10 d because of the low-carotenoid diet, increased during the free-living phase, and then leveled off by the end of the study (data not shown). The serum concentrations and the changes in serum concentrations of retinoids, carotenoids, and tocopherols at each sampling time point over the whole study period showed no correlation either with the blood response of labeled retinol or with the conversion efficiency of labeled ß-carotene.
View this table:
TABLE 1 . Subject characteristics and fasting serum concentrations of retinol, carotenoids, and tocopherols at the beginning of the study1
Conversion factor 53 d after the dose
The serum response curves after the [2H8]ß-carotene dose and the [2H8]retinyl acetate reference dose of 3 subjects are presented in Figure 1. Three subjects were chosen to show the interindividual variation in the conversion of ß-carotene to retinol by wt among an average converter (5.6:1; top panel), the best converter (2.4:1; middle panel), and the poorest converter (20.2:1; bottom panel). The enrichment measurements on all the data points were above the detection limit of our GC-MS method.
FIGURE 1. . Serum response curves of [2H4]retinol (derived from ß-carotene; ), [2H8]retinol (derived from retinyl acetate; ), and [2H8]ß-carotene (•) after doses of 6 mg [2H8]ß-carotene (day 0) and [2H8]retinyl acetate (day 3). The top panel is from an average converter (a man aged 47 y) with a conversion factor by wt of 5.6, the middle panel is from the best converter of the 22 subjects (a woman aged 64 y) with a conversion factor by wt of 2.4, and the bottom panel is from the poorest converter of the 22 subjects (a man aged 71 y) with a conversion factor by wt of 20.2.
The mean AUC serum responses of [2H4]retinol formed from the [2H8]ß-carotene dose for all subjects and for the men and women separately are presented in Table 2. The difference between the men and women was not significant (P = 0.2). The AUC serum response of [2H8]retinol formed from the [2H8]retinyl acetate dose for each subject is also presented in Table 2. No significant difference was observed (P = 0.4) between the sexes.
View this table:
TABLE 2 . Mean 53-d serum [2H4]retinol and [2H8]retinol responses after doses of [2H8]ß-carotene (ß-C) and [2H8]retinyl acetate (RAc), with calculated vitamin A equivalences and calculated conversion factors of [2H8]ß-C to vitamin A1
The retinol equivalence of the ß-carotene dose is presented in Table 2. There was no significant difference between the men and women (P = 0.3).
The average conversion factor of ß-carotene to retinol after a 6.0-mg ß-carotene dose was 4.8 (range: 1.3–10.8) on a molar basis or 9.1 (range: 2.4–20.2) on a weight basis (Table 2). No sex difference in the conversion factor was observed (P = 0.8). As shown in Figure 2, conversion factors in the individual subjects were significantly (positively) correlated with BMI (r = 0.57, P = 0.006). However, the correlation between the conversion factor and BMI was r = 0.41 (P = 0.24) in the men and r = 0.72 (P = 0.01) in the women.
FIGURE 2. . Correlation of ß-carotene–to–vitamin A conversion factor (by wt) with BMI in individual subjects. For women (), conversion factor = 0.17 x BMI - 35.1 (r = 0.71, P = 0.01); for men (•), conversion factor = 0.7 x BMI - 9.5 (r = 0.41, P = 0.24); and for all subjects, conversion factor = 1.2 x BMI - 22.1 (r = 0.57, P = 0.006).
ß-Carotene postabsorption and conversion
The AUCs of total [2H8]retinol from the [2H8]retinyl acetate dose and of [2H4]retinol from the [2H8]ß-carotene dose at various time points (1, 6, 14, 21, and 53 d after each dose) are provided in Table 3. Also shown in Table 3 is the calculated AUC of [2H4]retinol from the [2H8]ß-carotene dose if only ß-carotene conversion to vitamin A during absorption by the intestine is considered, ie, if no postabsorption conversion took place. The increase in the AUC of [2H4]retinol from the ß-carotene dose over time was greater than the increase in the AUC of [2H8]retinol from retinyl acetate over time, indicating that additional [2H4]retinol from ß-carotene appeared in the circulation with time. This additional retinol was estimated to be 7.8%, 13.6%, 16.4%, and 19.0 % of the total converted retinol at 6, 14, 21, and 53 d after the ß-carotene dose, respectively.
View this table:
TABLE 3 . Serum retinol responses from 22 volunteers 1, 6, 14, 21, and 53 d after the [2H8]ß-carotene (ß-C) and [2H8]retinyl acetate (RAc) doses1
DISCUSSION
We investigated the absorption and conversion of a physiologic dose (6 mg) of [2H8]ß-carotene in oil and a physiologic reference dose of [2H8]retinyl acetate in oil in a well-nourished population with normal vitamin A and vitamin E status (Table 1). The serum response showed that the absorption of ß-carotene and its metabolite, retinol, peaked 1 d after the administration of the ß-carotene dose. The serum response to the retinyl acetate dose also peaked 1 d after the administration of the retinyl acetate dose, as shown in Figure 1 (representative response curves from one volunteer). Whereas the serum responses of individual subjects to either [2H8]ß-carotene or [2H8]retinyl acetate varied 8–10-fold, all of our subjects responded to the physiologic doses of ß-carotene and vitamin A.
With the use of the isotope reference method, the reference tracer (labeled retinyl acetate) should not be able to compete with the absorption of the main tracer (labeled ß-carotene) in the intestine. We have observed a dose-dependent conversion efficiency of ß-carotene to vitamin A (9), which implies an intestinal limitation on the absorption and conversion of ß-carotene to vitamin A. Because we lack detailed knowledge of the capacity of the intestine to absorb and convert ß-carotene, we gave the 3-mg reference dose ([2H8]retinyl acetate) to the volunteers 3 d after the oral 6-mg ß-carotene dose to avoid the potential absorption competition. In other studies, 2 different labeled doses were either given 7 d or 2 y apart (6,7,9) or given simultaneously as an oral dose of 80 µg [13C10]ß-carotene and 80 µg [13C10]retinyl palmitate (8). Because the [2H8]retinyl acetate reference dose was given 3 d after the [2H8]ß-carotene dose and the serum samples were collected for 56 d after the [2H8]ß-carotene dose, the AUCs were calculated up to 53 d.
The isotope-labeled ß-carotene can be administered either as a single dose (acute dose method) or as multiple doses (chronic dose method). The advantages and disadvantages of the 2 approaches depend on the enrichment detectability of the mass spectrometric technology. On the basis of the current detection limit of mass spectrometric technology, a method using a single dose is limited to the study of foods with a ß-carotene content > 1 mg in the meal (3). Therefore, foods or diets with low ß-carotene content (< 1 mg ß-carotene per meal) can only be studied by using multiple doses to reach detectable enrichment concentrations in the circulation. The advantage of a method using multiple doses is that it is more closely related to usual dietary practice; also, fewer blood collections are required to determine conversion factors, which facilitates studies in young children. When using multiple doses of ß-carotene in oil (low doses), however, the stability of the ß-carotene is of concern because it can be easily isomerized, as van Lieshout et al (8) have reported. The absorption and clearance kinetics of a single dose of ß-carotene in oil (collecting many blood samples) can provide valuable information on the absorption and conversion of ß-carotene to vitamin A, such as the results that are provided in the present study.
Vitamin A in the human circulation includes mainly retinyl esters and retinol. In theory, any one of these compounds could be monitored by using the isotope reference method to calculate the vitamin A equivalence of ß-carotene. However, retinol is the predominant form of vitamin A in the circulation under steady state conditions, and it can be studied in well-nourished populations only with the use of labeled compounds. We chose not to consider postprandial labeled retinyl esters because 1) the postabsorption retinyl esters’ response peaks shortly (< 12 h) after the dose (3,6,15), 2) the absorbed retinyl esters begin to be converted to retinol right after the dose, and 3) the continuous conversion of retinyl esters to retinol ends at 24 h after the dose. Thus, for this long-term isotope reference study, retinyl esters were not monitored.
As shown in Table 2, the AUC for the total serum response of [2H4]retinol to the [2H8]ß-carotene dose showed a trend of sex difference (P = 0.2), even though our female subjects were =" BORDER="0"> 1 y postmenopausal. Because of the large SD within the same sex, power calculation suggests that =" BORDER="0"> 37 subjects for each sex would be needed to reach statistical significance.
The mean conversion factor of 6.0 mg ß-carotene in oil to retinol determined by using this approach was 9.1 (by wt), with a variation from 2.4 to 20.2. This is different from the conversion factors determined previously by using a depletion-repletion method for the ß-carotene equivalence (conversion factor) to vitamin A. Hume and Krebs (16) reported a depletion study conducted in 16 healthy subjects aged 19–34 y (7 additional subjects served as positive control subjects). After 12 mo of depletion, only 3 of the subjects were vitamin A deficient. Of the 3 subjects with "unmistakable" signs of vitamin A deficiency, 2 were given ß-carotene and 1 was given preformed vitamin A. The results showed that daily doses of 1500 µg ß-carotene or 390 µg retinol were sufficient to treat vitamin A deficiency in these subjects. Therefore, from this human study, the conversion ratio of ß-carotene to vitamin A was 3.8 to 1 by wt. In 1974, Sauberlich et al (17) reported another extensive and well-controlled vitamin A depletion-repletion study in humans. They recruited 8 healthy men aged 31–43 y. These volunteers were depleted in vitamin A within 359–771 d. The investigators repleted 5 subjects with vitamin A and 3 subjects with ß-carotene and found that daily doses of 600 µg retinol or 1200 µg ß-carotene were required to cure vitamin A deficiency. Thus, in that study, the conversion factor of ß-carotene to vitamin A was 2 by wt. In our study results, only 4 of 22 subjects (18%) had a conversion factor < 3.1 (2.4, 2.7, 3.0, and 3.1) by wt. Our subjects had normal vitamin A status, whereas in the earlier studies, all subjects had been vitamin A depleted. Hence, one explanation for the different results might be that the conversion of ß-carotene to vitamin A is less efficient in a population with normal vitamin A status than in subjects who are vitamin A deficient. In addition, we observed in a previous study that the conversion efficiency of ß-carotene to vitamin A decreases as the ß-carotene dose increases (14). We speculate that if a lower dose of ß-carotene were given (such as 1 to 2 mg compared with the 6-mg dose used in the present study), it would more efficiently converted than 9:1 and that the variation of the conversion factor (2.4–20.2) would be smaller as well.
All of our subjects responded to the 6-mg dose of ß-carotene in oil (0.1%); that is, every subject absorbed ß-carotene and converted it to retinol. The lowest converter of this group was a 71-y-old man who absorbed ß-carotene normally. The best converter was a 64-y-old woman, as shown in the middle panel of Figure 1.
The cis isomers of ß-carotene in our doses were not a major component of the dose. The contribution of the cis isomers of ß-carotene to the conversion factor was not evaluated.
There was a significant correlation between the ß-carotene–to–vitamin A conversion factor and the BMI of the subjects (Figure 2). This implies that body composition may have some bearing on the efficiency of ß-carotene conversion. Considering that the correlation between BMI and percentage body fat was 0.58–0.86 for men and 0.53–0.85 for women (18) and that ß-carotene is mainly stored in the fat, it is reasonable to speculate that the correlation between BMI and the conversion factor of ß-carotene to retinol is linked to the fat content of the body. This suggests that subjects with more body fat have a lowered capability to convert ß-carotene to vitamin A, and that fat serves as a repository or sink for ß-carotene, which will be poorly metabolized.
Although no reports exist concerning the association between ß-carotene conversion to vitamin A and BMI, a negative association was seen between the serum ß-carotene concentration and BMI in both men (n = 253) and women (n = 276), with the association being statistically significant in women but not in men (19). In another report, the baseline plasma ß-carotene concentration was significantly and negatively associated with BMI in all subjects, whereas the increases in plasma ß-carotene concentration after the consumption of a controlled high-carotenoid diet for 15 d were significantly and inversely correlated with BMI in older women (20). We did not observe an association between baseline serum concentrations of ß-carotene and BMI in our subjects (10 men, 12 women). Furthermore, there was no significant difference between the men and the women in either ß-carotene–to–vitamin A conversion factors or BMIs. However, the correlation between ß-carotene conversion to vitamin A and BMI was observed in the women (r = 0.71, P = 0.01), whereas no correlation was observed in the men (r = 0.41, P = 0.2). These observations warrant further investigation regarding the effect of body fat on ß-carotene and vitamin A absorption, conversion, and tissue deposition. Nevertheless, our data support the idea that if a person has a high BMI, a larger portion of the ß-carotene becomes stored (probably in fat) and that the stored ß-carotene is less available for conversion to vitamin A.
The intestine is the most important site for the conversion of ß-carotene to vitamin A. However, in vitro studies have shown that liver, fat, lung, and kidney are all capable of converting ß-carotene to retinoids (21). A recent article reported that human liver possesses substantial dioxygenase activity as compared with small intestine (22). In addition, mathematical modeling showed that to fit a physiologic compartmental model prediction to the experimental data for in vivo ß-carotene conversion to retinol, the intestine and liver should both be considered (23). We used 1 d as the cutoff for the intestinal conversion of ß-carotene to retinol and calculated the AUC of [2H4]retinol. With the assumption that the retinol from the ß-carotene dose follows the same vitamin A kinetics as the [2H8]retinol from the [2H8]retinyl acetate dose, the AUCs of [2H4]retinol from labeled ß-carotene at different time points should be the same as what is calculated from the vitamin A kinetics (presented in Table 3). However, we found that [2H4]retinol (from the [2H8]ß-carotene dose) did not follow the same serum response kinetics as [2H8]retinol (from the [2H8]retinyl acetate dose). The difference between the measured AUC of [2H4]retinol (column 2 of Table 3) and the calculated AUC of [2H4]retinol (column 3 of Table 3) is probably due to continuous conversion of already absorbed [2H8]ß-carotene to retinol in vivo in tissues such as the liver.
Our results showed that the total postabsorption conversion over 53 d reached 19.0% (last column of Table 3). Therefore, it appears that, on average in a well-nourished population, 81% of total vitamin A formed from a ß-carotene dose is from intestinal cleavage and 19% is from extraintestinal cleavage (postabsorptive or whole-body). Further investigation on the postintestinal conversion of ß-carotene is warranted.
In our subjects, there were no sex differences in serum concentrations of ß-carotene, cryptoxanthin, tocopherols, and vitamin A. However, there was a marginal sex difference in the serum concentration of lutein.
In summary, the quantitative determination of the conversion of ß-carotene to vitamin A in humans can be accomplished by using this isotope reference method. In well-nourished adults after a 6-mg dose, 9.1 µg ß-carotene on average is nutritionally equivalent to 1 µg retinol. This conversion factor of ß-carotene to retinol is positively correlated with BMI in women. The postintestinal absorption conversion is estimated to be 19% of the total converted retinol.
ACKNOWLEDGMENTS
We thank the volunteers for participating in the study, Youngsim Park for HPLC data processing, the nurses and staff of the Metabolic Research Unit at the Human Nutrition Research Center on Aging, and BASF for providing the labeled ß-carotene.
Author contributions are as follows: study design, GT and RMR; data collection, GT, JQ, and GGD; data analysis, GT and JQ; and writing of the manuscript, GT, RMR, and GGD. None of the authors had a personal interest or advisory board affiliation with any of the supporters of this research.
REFERENCES