点击显示 收起
1 From the Nestlé Research Center, Nestec Ltd, Lausanne, Switzerland (MR, ME, CH, IT, J-PG, AB, SM, SQ, CP-W, LS, HG, and LBF) and Nestlé Product Technology, Centre Konolfingen, Konolfingen, Switzerland (MG)
2 Address reprint requests to M Richelle, Nestec Ltd, Nestlé Research Center, PO Box 44, CH-1000 Lausanne 26, Switzerland. E-mail: myriam.richelle{at}rdls.nestle.com.
See corresponding editorial on page 3.
ABSTRACT
Background: Plant sterols reduce cholesterol absorption, which leads to a decrease in plasma and LDL-cholesterol concentrations. Plant sterols also lower plasma concentrations of carotenoids and -tocopherol, but the mechanism of action is not yet understood.
Objectives: The aims of this clinical study were to determine whether plant sterols affect the bioavailability of ß-carotene and -tocopherol in normocholesterolemic men and to compare the effects of plant sterol esters and plant free sterols on cholesterol absorption.
Design: Twenty-six normocholesterolemic men completed the double-blind, randomized, crossover study. Subjects consumed daily, for 1 wk, each of the following 3 supplements: a low-fat milk-based beverage alone (control) or the same beverage supplemented with 2.2 g plant sterol equivalents provided as either free sterols or sterol esters. During this 1-wk supplementation period, subjects consumed a standardized diet.
Results: Both of the milks enriched with plant sterols induced a similar (60%) decrease in cholesterol absorption. Plant free sterols and plant sterol esters reduced the bioavailability of ß-carotene by 50% and that of -tocopherol by 20%. The reduction in ß-carotene bioavailability was significantly less with plant free sterols than with plant sterol esters. At the limit of significance (P = 0.054) in the area under the curve, the reduction in -tocopherol bioavailability was also less with plant free sterols than with plant sterol esters.
Conclusions: Both plant sterols reduced ß-carotene and -tocopherol bioavailability and cholesterol absorption in normocholesterolemic men. However, plant sterol esters reduced the bioavailability of ß-carotene and -tocopherol more than did plant free sterols.
Key Words: Plant sterol • cholesterol • absorption • vitamin E • tocopherol • carotenoids
INTRODUCTION
Plant sterols reduce the absorption of cholesterol in the gut, possibly by competing with cholesterol when they are incorporated into the mixed micelles, by displacing the cholesterol from bile, or by decreasing the hydrolysis of cholesterol esters in the small intestine (1–3). However, the exact cholesterol-lowering mechanism of plant sterols is not yet fully known. Many clinical studies performed in humans show that the administration of plant sterols reduces cholesterol absorption and that, when plant sterols are given over a long period (>3 wk), plasma cholesterol and LDL-cholesterol concentrations are reduced by 5%–15% without major side effects (4–6). The reduction in LDL cholesterol is dose dependent: a measurable reduction of 6% with an intake of 0.9 g plant sterol/d and nearly maximum at 9.6% with an intake of 2 g/d (3, 7–9). Concomitant with cholesterol reduction, plant sterols decrease plasma concentrations of ß-carotene by 25%, -carotene by 10%, and vitamin E by 8%, but plasma vitamin A (retinol) and vitamins D and K are not significantly affected (7–13). Because sterols and stanols reduce the amount of LDL cholesterol and because lipophilic carotenoids and tocopherols are known to be associated with LDL particles, it may be appropriate to adjust the plasma concentrations of these carotenoids and vitamins to reduce LDL cholesterol. With such an adjustment, stanols and sterols do not significantly lower blood concentrations of vitamin E, but concentrations of ß-carotene were reduced by 8%–19% (8). The reason for this decrease in blood concentrations of ß-carotene is not known, but it could be the reduction in its absorption. A reduction in the bioavailability of ß-carotene is of particular concern for persons whose need for vitamin A is greater, such as pregnant and lactating women and young children.
Plant free sterols have been incorporated into a low-fat milk-based beverage. In a previous clinical trial, we showed that midly hypercholesterolemic subjects consuming this low-fat milk-based beverage enriched with 1.8 g plant free sterols/d had a 40% reduction in cholesterol absorption (14). In a longer-term clinical trial performed in midly hypercholesterolemic Danish men and women, daily consumption of 1.2 and 1.6 g plant free sterols in this low-fat milk-based beverage over 4 wk reduced LDL cholesterol by 7.1% and 9.6%, respectively. In addition to cholesterol, both doses of plant sterols decreased the percentage change in plasma - and ß-carotene, lutein, and vitamin E (15).
The objective of the present clinical study was to investigate whether daily consumption of 2.2 g plant sterols (as free equivalent) either as free or ester form would affect the bioavailability of ß-carotene and -tocopherol in humans compared with the same test performed without sterol ingestion (control treatment). In addition, the effect of plant free sterols was compared with that of plant sterol esters. To our knowledge, the effects of nonesters and esters on the bioavailability of these fat-soluble vitamins have not been evaluated in the same study. Esters are presumed to have a greater effect on fat-soluble vitamins because they partition into the oil phase of the intestine, whereas free sterol partitions into the micellar phase (16).
SUBJECTS AND METHODS
Subjects
Thirty-three healthy men were enrolled in the study. The inclusion criteria were that the subjects be nonvegetarians and nonsmokers and that they not have metabolic disorders such as diabetes; hypertension; renal, hepatic, or pancreatic disease; or ulcers. Subjects were normolipidemic: ie, they had plasma cholesterol concentrations <5.2 mmol/L (or 200 mg/dL), a ratio of plasma cholesterol to HDL cholesterol <5.0 mmol/L, and plasma triacylglycerol concentrations < 2.0 mmol/L. Because of the large amount of blood (827.5 mL) that was drawn during the study, subjects were required to have a blood hemoglobin concentration of >13 g/dL. Subjects were excluded from the study if they used cholesterol-altering medication or vitamin and mineral supplements from 3 mo before the start of the study until the completion of the study; were milk intolerant or had had major gastrointestinal surgery; exercised intensively, such as running marathons; and consumed daily >2 glasses of wine (3 dL), >2 beers (3 dL), or >1 glass (shot glass) of hard liquor.
The protocol was approved by the ethics committee of Nestlé (Lausanne, Switzerland). Subjects received information on the background and design of the study and gave written informed consent before participation. They were free to withdraw from the study at any time.
Study design
This was a placebo-controlled, double-blind, randomized, 3-period, 3-treatment crossover clinical trial. The protocol is presented in Figure 1.
FIGURE 1.. Scheme of the experimental design of the clinical study.
Subjects consumed a standard diet over 6 d that was designed to provide a constant daily intake of 250–260 mg cholesterol. The subjects consumed this standard diet during the three 1-wk treatment periods. Breakfast was a traditional European-style breakfast: ie, bread, butter, and jam. Lunches and dinners consisted of ready-to-eat meals and bread slices. The caloric intake was 2600 kcal. In addition to this standard diet, subjects drank daily, as an intervention, 600 mL of low-fat milk-based beverage supplement—ie, a 300-mL dose at breakfast and a second 300-mL dose at lunch.
On the morning of day 4 after an overnight fast, subjects arrived at the Metabolic Unit of Nestlé Research Center (Vers-chez-les-Blanc, Switzerland) and received a simultaneous intravenous infusion of 45 mg [13C2]cholesterol [incorporated into 15 mL of parenteral emulsion (Interlipid 10%; Pharmacia-Upjohn, Stockholm)] and an oral dose of 15 mg of [2H5]cholesterol (diluted in 1 mL sunflower oil). Both markers were prepared freshly and were administered within 2 h.
On the morning of day 7 after an overnight fast, each subject consumed a standard meal consisting of 15 mg [2H8]ß-carotene and 30 mg [2H6]-tocopherol incorporated in 35 g peanut oil that was mixed with 70 g wheat semolina (cooked with 200 mL tap water). In addition, they consumed 40 g bread, 60 g cooked egg whites, and 600 mL low-fat milk-based beverage supplement. This meal was consumed within 30 min. No other food was allowed over the subsequent 9 h, but subjects were allowed to drink bottled water (Vittel, Vittel, France).
To ensure subject compliance, breakfasts and lunches were consumed in the Metabolic Unit on the weekdays under the supervision of the unit's staff, and packed meals for dinners and the weekend were provided for home consumption. Subjects were repeatedly instructed not to consume any food or beverages other than those provided by the Metabolic Unit.
Milk supplement
Control milk refers to a low-fat, ultrahigh temperature-treated, milk-based beverage (600 mL) containing 0.63% butter oil, 0.56% rapeseed oil, and 0.39% corn oil. The fatty acid profile of the milk comprises 15% palmitic acid, 5% stearic acid, 37% oleic acid, 21% linoleic acid, and 3% -linolenic acid. This milk was packed in a 200-mL sealed container. Control milk (600 mL) enriched with 2.2 g soybean nonhydrogenated, nonesterified plant sterols and 2.2 g sorbitan tristearate (as emulsifier), packed in a 200-mL sealed container, is referred to as plant free sterol milk. Control milk (600 mL) enriched with soybean nonhydrogenated, esterified plant sterols (2.2 g free sterols equivalent), packed in a 200-mL sealed container, is referred to as plant sterol ester milk.
The composition of the 3 milks was adjusted to provide a similar fatty acid profile as well as similar ß-carotene and -tocopherol content. The sterol content in the milk supplement was ascertained after manufacture and was in agreement with the 2.2 g free plant sterol equivalent.
Isotopic markers
The 2, 2, 4, 4, 6-[2H5] cholesterol tracer (95 atom%) (Medical Isotopes, Pelham, NH) was diluted in sunflower oil (15 mg/g) and placed on bread. Ready-to-inject 15-mL syringes containing 45 mg 3-4-[13C2] cholesterol tracer (99 atom%; Medical Isotopes) dissolved in Intralipid 10% were prepared by the Centre Hospitalier Universitaire Vaudois (Lausanne, Switzerland) and checked for sterility.
We purchased 10,10',19,19,19,19',19',19'-[2H8]-ß-carotene (95 atom%) and 2R,4'R,8'R-[2H6]--tocopherol (98 atom%) from Orphachem (Clermont-Ferrand, France). A dose of 15 mg [2H8]-ß-carotene and 30 mg [2H6]--tocopherol was incorporated into 35 g peanut oil.
Collection of blood samples
A fasting blood sample was drawn from an anticubital vein by venipuncture into a potassium EDTA-containing evacuated tube that was immediately placed in an ice-water bath. The tube was protected from light and then centrifuged (10 min, 4 °C, 3000 x g) to separate the plasma, which was isolated and stored at –20 °C until it was analyzed. Samples were analyzed within 6 mo.
For measurement of cholesterol absorption, a fasting blood sample was collected on the morning of day 4 before the administration of the 2 cholesterol markers ([13C2]-cholesterol and [2H5]-cholesterol) and then on the morning of days 6 and 7. For measurement of ß-carotene and -tocopherol bioavailability, a fasting blood sample was collected on the morning of day 7 before the administration of the 2 markers—ie, [2H8]-ß-carotene and [2H6]--tocopherol—as well as at 2.5, 3, 4, 5, 6, 7, 8, and 9 h after the administration of these markers.
Plasma cholesterol enrichment
Plasma lipids were extracted and separated by thin-layer chromatography as described by Pouteau et al (14). The free cholesterol layer of each lipid extract sample was scraped off into a tube and then extracted with 3 mL EtOH:CHCl3 (1:2, by vol). The mixture was mixed by vortex for 10 s and centrifuged for 2 min at 2000 x g at room temperature. Finally, the free cholesterol extract was split into 2 parts: one part of 0.3 mL for measurement of [13C]-cholesterol enrichment by using gas chromatography-combustion-isotope ratio mass spectrometry (GC/C/IRMS) analysis and another part of 2.7 mL for measurement of [2H]-cholesterol enrichment by high-temperature conversion-elemental analyzer-isotope ratio mass spectrometry (TC-EA/IRMS) on an isotope ratio mass spectrometer (Delta Plus XL; Thermo Finnigan MAT, Bremen, Germany).
The [2H/H] isotope ratio measurement of free cholesterol was performed by using TC-EA/IRMS. The free cholesterol fraction (2.7 mL) was evaporated under N2 at 30 °C. The dry residue was dissolved in 25 µL EtOH/CHCl3 (1:2, by vol), mixed by vortex for 5 s, and then transferred to a silver capsule. The solution was evaporated at 30 °C, and the dried capsule introduced to the solid autosampler of the TC-EA/IRMS system. Analytic conditions of the TC-EA/IRMS system were pyrolysis reactor temperature of 1450 °C, GC column temperature of 90 °C, and helium pressure of 1 bar. We measured the [2H/H] isotopic enrichment of cholesterol by monitoring ions' mass-to-charge ratios of 3 and 2, expressed it in against standard mean ocean water (an international standard for [2H]), and further converted it to molar percent excess.
The [13C:12C] isotope ratio measurement of free cholesterol was performed by using GC/C/IRMS (MAT 252, Thermo Finnigan Mat) according to Pouteau et al (14). The [13C] isotopic enrichment was expressed in against Pee Dee belemnite (an international standard for [13C]) and then converted to molar percent excess.
Triacylglycerol-rich lipoprotein isolation and lipid extraction
Plasma (4 mL) was thawed and introduced in an ultracentrifuge tube. A solution of 1.006 g sodium bromide/mL was deposited on top of the plasma solution without mixing the 2 solutions. The tube was filled with this 1.006 g/mL solution and ultracentrifuged at 100 000 x g and 15 °C for 30 min. The upper phase containing triacylglycerol-rich lipoproteins (TRLs)—mainly chylomicrons with low amounts of VLDL—was collected. TRLs (200–400 µL) were adjusted up to 1 mL with distilled water. Subsequently, 1 mL EtOH, 5 µL deferoxamine mesylate (10 mg/mL water), and 2 mL hexane containing 350 mg BHT/L were added. The tube was mixed by vortex for 20 s, then centrifuged at 2000 x g and room temperature for 10 min in a tabletop centrifuge. The organic phase was collected while the water phase was extracted again with an additional 2 mL hexane-BHT. The hexane layers were combined and evaporated to dryness under N2. The sample was dissolved in 200 µL hexane-BHT for tocopherol analysis and in 120 µL dioxane-ethanol-acentonitrile (1:1:2, by vol) for carotenoid analysis. A volume of 60 or 100 µL was injected onto the HPLC system for tocopherol and carotenoid analysis, respectively. The lipid extract of the biological sample was quite stable for 1 wk at 4 °C and then for 2 d at room temperature before being used for HPLC injection.
Measurement of TRL-[2H8]-ß-carotene
We isolated [2H8]-ß-carotene from ß-carotene by using an HPLC method. The [2H8]-ß-carotene was then quantified with the use of a diode array detector according to the method of Duecker et al (17).
Determination of TRL-D6--tocopherol
We isolated [2H6]--tocopherol from -tocopherol by using an HPLC method. The [2H6]--tocopherol was then quantified with the use of an ultraviolet detector according to the method of Richelle et al (18).
Assessment of bioavailability
Cholesterol
Two cholesterol markers were used: one marker, injected intravenously, gave information on the in vivo metabolism of cholesterol, and the other, administered orally, accounted for intestinal cholesterol absorption and its in vivo metabolism. This method allowed the measurement of the absolute bioavailability of cholesterol in each subject, which was calculated by multiplication of the ratio of the plasma enrichment of cholesterol administered orally and plasma enrichment of cholesterol injected intravenously by the ratio of the intravenous dose of cholesterol and the oral dose administered to the subject, and this total was then multiplied by 100.
ß-Carotene and -tocopherol
The ß-carotene and -tocopherol molecules were administered orally only, and therefore relative bioavailability was ascertained. This assessment was performed by using the pharmacokinetic measurements of the appearance of these labeled molecules in the blood circulation of the subjects, such as area under the curve (AUC), maximum plasma concentration (Cmax), and the time to reach the Cmax (tmax).
Statistical analysis
From time 0 to 9 h, 3 pharmacokinetic parameters were calculated: area under the plasma concentration versus time curve [AUC(0–9 h)], (Cmax), and tmax. The AUC(0–9 h) over baseline was calculated with the use of the trapezoidal rule. These 3 pharmacokinetic values were calculated by using NCSS software (version 2000; NCSS Statistical Software, Kaysville, UT).
For cholesterol, ß-carotene, and -tocopherol, statistical analyses were based on noninferiority tests with a lower equivalence limit of 80% (19). Differences between treatments were analyzed by using a linear mixed-effect model with treatment and period as fixed effects and by using subject as a random effect (SAS software, version 8.2; SAS Corp Inc, Cary, NC). The rejection level in statistical tests was 5%. Data are presented as means ± SEMs.
RESULTS
Among the 33 subjects enrolled in the study, 7 did not complete the study: 3 for personal reason, 2 because of noncompliance with the treatment, and 2 for medical reasons. Twenty-six subjects, aged 29 ± 1 y, completed the study. Their mean starting body weight was 78 ± 2 kg, and their mean starting body mass index (BMI; in kg/m2) was 24.1 ± 0.4. The values were analyzed only for these 26 subjects who completed the 3 treatment periods. The body weight of the subjects did not change during the course of the study.
Cholesterol absorption
Our normocholesterolemic subjects consuming control milk had cholesterol absorption of 47 ± 2% (Figure 2), which is in agreement with data reported in the literature (20) and our previous study (21). As expected, the administration of plant free sterols and plant ester sterols reduced cholesterol absorption to 23 ± 2% and 20 ± 2%, respectively, of baseline. This reduction in cholesterol absorption by 60% is in agreement with data reported by Lees et al (6) and Ostlund et al (3).
FIGURE 2.. Mean (±SEM) percentage of cholesterol absorption in subjects who daily consumed a low-fat milk-based beverage (control) or the same beverage supplemented with either 2.2 g plant free sterols or plant sterol esters in quantities equivalent to 2.2 g plant free sterols.
TRL production
Dietary lipophilic compounds have to be incorporated into chylomicrons, which are their vehicles in the bloodstream. Chylomicron production is characterized by the triacylglycerol content of the TRLs. Consumption of the standard meal containing 35 g peanut oil led to a marked production of chylomicron particles (Figure 3; Table 1). The pharmacokinetics of the TRL-triacylglycerol concentration consisted of a rapid increase, a Cmax, and a prompt decline. Nine hours after the consumption of the treatments, chylomicrons were totally cleared, which led to TRL-triacylglycerol concentrations that did not differ significantly from the baseline concentration. The AUC, Cmax, and tmax were similar for the 3 treatments, which showed the equivalence of the triacylglycerol absorption (Table 1).
FIGURE 3.. Mean (±SEM) triacylglycerol-rich lipoprotein (TRL)-triacylglycerol concentrations after the consumption of a standard diet containing 35 g peanut oil and milk. : Control, low-fat milk-based beverage; : free sterols, the same beverage enriched with a daily dose of 2.2 g plant free sterols; : sterol esters, the same beverage enriched with a daily dose of plant sterol esters (2.2 g free sterol equivalent).
View this table:
TABLE 1. Pharmacokinetic measurements of lipids of the triacylglycerol-rich lipoproteins after the consumption of the standard meal alone (control) or of the meal enriched with 2.2 g plant free sterols or 2.2 g plant sterol esters (plant free sterol equivalent)1
TRL-[2H8]-ß-carotene
A labeled deuterated ß-carotene has been used to distinguish the ß-carotene consumed from endogenous ß-carotene—ie, the ß-carotene already present in the body. Therefore, it is not surprising that, at time 0, there was no [2H8]-ß-carotene in the TRLs (Figure 4). After the consumption of the control milk, TRL-[2H8]-ß-carotene concentrations rose to a Cmax and then declined slowly; detectable concentrations remained 9 h after absorption (Figure 4; Table 1). The 50% reductions in the AUC and the Cmax confirm that both plant sterols decreased the bioavailability of [2H8]-ß-carotene. This reduction was significantly higher with the plant sterol esters (57%) than with the plant free sterols (48%).
FIGURE 4.. Mean (±SEM) triacylglycerol-rich lipoprotein (TRL)-[2H8]-ß-carotene concentrations after the consumption of a standard diet containing 15 mg [2H8]-ß-carotene and milk. : Control, low-fat milk-based beverage; : free sterols, the same beverage enriched with a daily dose of 2.2 g plant free sterols; : sterol esters, the same beverage enriched with a daily dose of plant sterol esters (2.2 g free plant sterol equivalent).
TRL-retinyl palmitate
Within the enterocytes, some of ß-carotene molecules are cleaved by the 15,15' dioxygenase, which leads to retinol that is partly transformed into retinyl palmitate. The TRL-retinyl palmitate pharmacokinetics were quite similar to those described for [2H8]-ß-carotene: a prompt increase to Cmax and a slow decline, leaving concentrations still elevated 9 h after absorption (Figure 5; Table 1). Both plant sterols significantly decreased the concentration of TRL-retinyl palmitate as characterized by reductions in the AUC(0–9 h) and Cmax. Similar to the reduction in [2H8]-ß-carotene, the reduction in TRL-retinyl palmitate was significantly higher with the plant sterol esters (48%) than with the plant free sterols (32%).
FIGURE 5.. Mean (±SEM) triacylglycerol-rich lipoprotein (TRL)-retinyl palmitate concentrations after the consumption of a standard diet containing 15 mg [2H8]-ß-carotene and milk. : Control, low-fat milk-based beverage; : free sterols, the same beverage enriched with a daily dose of 2.2 g plant free sterols; : sterol esters, the same beverate enriched with a daily dose of plant sterol esters (2.2 g free sterol equivalent).
TRL-D6--tocopherol
The pharmacokinetics of [2H6]--tocopherol had a slower increase and decrease than did the pharmakinetics of [2H8]-ß-carotene. Nine hours after the consumption of both treatments, the concentrations of [2H6]--tocopherol remained quite high (Figure 6; Table 1). Plant free sterols had no effect on the bioavailability of [2H6]--tocopherol, whereas plant sterol esters reduced it 27% in comparison to the control treatment. Of the 2 plant sterol treatments, plant sterol esters tended to decrease the bioavailability of [2H6]--tocopherol more than did plant free sterols, as characterized by a significant difference in Cmax whereas the difference in the AUCs was at the limit of significance (P = 0.054).
FIGURE 6.. Mean (±SEM) triacylglycerol-rich lipoprotein (TRL)-[2H6]--tocopherol concentrations after the consumption of a standard diet containing 30 mg [2H6]--tocopherol and milk. : Control, low-fat milk-based beverage; : free sterols, the same beverage enriched with a daily dose of 2.2 g free plant sterols; : sterol esters, the same beverage enriched with a daily dose of plant sterol esters (2.2 g free plant sterol equivalent).
DISCUSSION
Daily consumption of plant sterols for several weeks has been proven to reduce LDL cholesterol in both normocholesterolemic (11, 20) and hypercholesterolemic (3, 11, 21) persons. Inhibition of the intestinal absorption of exogenous (dietary) and endogenous (biliary) cholesterol was described as the major mechanism of the cholesterol-lowering action of plant sterols, although the exact mechanism of action remains to be investigated in detail. Long-term daily consumption of plant sterols induces a reduction in plasma cholesterol and LDL-cholesterol concentrations. These depletions are related to the dose of plant sterol consumed, the form of the plant sterol, the vehicle the plant sterols are solubilized in, the background diet, genetic factors, and, in some cases, the starting cholesterol concentration (3). The vehicle in which the plant sterols are solubilized can clearly affect the efficacy of plant sterols in lowering LDL cholesterol. Although we have shown in several studies that plant free sterols properly solubilized in low-fat partly vegetable oil-filled milks can lower cholesterol absorption (14) and LDL cholesterol (15), a recent study found that plant sterol did not effectively lower LDL cholesterol in a low-fat liquid matrix (22).
Esters of plant sterols or stanols similar to those of the free forms have been shown to induce a similar effect when provided at the same free sterol equivalent dose (11, 23, 24), but the issue is not completely resolved. In the present study, normocholesterolemic men who daily consumed 2.2 g plant sterols in either free or ester form had similar 60% reductions in cholesterol absorption, which is in agreement with data reported by Lees et al (6) and Ostlund et al (3). Several studies showed that the consumption of stanyl and steryl esters reduces plasma concentrations of fat-soluble antioxidants such as ß-carotene, lycopene, and -tocopherol (7, 9, 11, 25–27). It is hypothesized that, during the absorption processes in the intestine, plant sterols could displace not only cholesterol but also other lipophilic molecules and replace them in incorporation into mixed micelles. It was shown that esters are more likely to partition into the oil phase of the intestine, whereas free sterols are more likely to partition into the mixed micellar phase (16). As a consequence, the absorption of these lipophilic molecules is reduced, and there is a more marked effect with plant sterol esters. Although this assumption is based on lower plasma concentrations when subjects are consuming plant sterols, such reduced absorption has not yet been shown.
Therefore, the purpose of the present study was to test the null hypothesis that consumption of 2.2 g plant sterols either as free or ester form (expressed in free sterol equivalent) would not affect the absorption of ß-carotene and -tocopherol in normocholesterolemic men. The clinical trial was designed as a placebo-controlled, randomized, double-blind, 3-treatment, 3-period study. Subjects consumed a controlled diet that provided similar daily intakes of cholesterol, ß-carotene, and -tocopherol. In addition, they were supplemented daily for 7 consecutive days with a low-fat milk-based beverage enriched with 2.2 g plant sterols (intervention groups) or not enriched (control group). The 1-wk duration of plant sterol administration ensured a sufficient (60%) reduction in cholesterol absorption. In the present study, ß-carotene and -tocopherol bioavailability was assessed at the end of the 1-wk treatment by using a postprandial test. Administered ß-carotene and -tocopherol were both labeled with deuterium to differentiate the dietary intake from the concentration already present in the subject's body. The principle of the postprandial test consists of the administration of a standard meal that allows efficient production of chylomicrons, which will carry dietary lipophilic molecules into the bloodstream of the subject. The pharmacokinetics of TRL-triacylglycerol concentration as measured against time is a measure of bioavailability of lipophilic molecule based on the assumption that the TRL fraction mainly contains intestinally derived lipoproteins (chylomicrons and their remnants) and only some liver-derived lipoproteins (VLDL). To compare the treatments, a reproducible measure of chylomicron production—ie, an intraindividual variability that is relatively low compared with interindividual variability—is required. Van Vliet et al (28) reported an unexpectedly large intraindividual variability (62%) in their population. To minimize the intraindividual variation, they recommended the standardization of ß-carotene bioavailability with chylomicron production, ie, triacylgylcerol responses. In the present study, the 3 treatments led to similar amounts of chylomicron production, as characterized by an equivalent AUC, Cmax, and tmax of TRL-triacylglycerol pharmacokinetics. Thus, ß-carotene and -tocopherol bioavailabilities have not been standardized with the bioavailability of triacylglycerol.
Supplementation with plant sterols, in either free or ester form, reduced the bioavailability of ß-carotene by 50% and that of -tocopherol by 20%. In the case of ß-carotene, the reduction in bioavailability was significantly less with plant free sterols than with plant sterol esters. The reduction in -tocopherol bioavailability was also less with plant free sterols than with plant sterol esters characterized by a lower Cmax and a reduction in AUC at the limit of significance (P = 0.054). These results indicate for the first time that plant sterols reduce not only the absorption of cholesterol but also the bioavailability of ß-carotene and -tocopherol. The results also show that sterol esters provided at the same equivalent dose and in the same vehicle as plant free sterols had a greater effect on ß-carotene and -tocopherol bioavailability than did plant free sterols. It is reasonable to speculate that the absorption of other lipophilic molecules that partition into the intestinal oil phase will be compromised by plant free sterols. The present results contrast with those of Relas et al (29), who showed that a single 1-g dose of dietary stanyl esters incorporated in margarine and combined with fat-soluble vitamins did not detectably interfere with serum concentrations of cholesterol, triacylglycerol, -tocopherol, ß-carotene, retinol, or retinyl palmitate. However, in their study, there was no direct evidence that cholesterol absorption was reduced; they did find a decrease in plasma concentrations of campesterol:cholesterol, which is a marker of cholesterol absorption.
The chronic consumption of plant sterols would be expected to induce a progressive decrease in plasma carotenoid concentrations, although the biological significance of this expected effect is not clear. One way to counterbalance this reduction in absorption is to increase dietary carotenoid intake. Noakes et al (30) showed that daily consumption of 5 servings fruit and vegetables with a minimum of 1 carotenoid-rich serving/d allows the maintenance of lipid-standardized plasma carotenoid concentrations in subjects consuming either 2.5 g plant sterol ester-2 or 2.3 g plant stanol ester-1. An alternative consists of an increase in the intake of carotenoid-rich food such as apricots, cantaloupe, broccoli, and spinach (26) or the consumption of a food supplement (27). In conclusion, plant sterols reduce not only the absorption of cholesterol but also the bioavailability of ß-carotene and -tocopherol.
ACKNOWLEDGMENTS
We thank Olivier Ballèvre for discussions during the development of this study and Etienne Pouteau for advice on marker selection and dose determination. We greatly thank Patricia Dibling, Sylviane Oguey-Araymon, Annie Blondel-Lubrano, Bernard Decarli, Micheline Chabloz, and Jean-Claude Maire of the Metabolic Unit for their collaboration during this clinical trial. We also thank Isabelle Ré for her participation in the HPLC analyses and Charles Schindler of the Centre Hospitalier Universitaire Vaudois de Lausanne for preparing the emulsions. Finally, we thank the volunteers who participated in this clinical trial.
The study was conceived and designed by MR, ME, CH, AB, and LBF. The design and production of formulations of the phytosterol milk were achieved by MG, AB, and LS. MR coordinated the trial and supervised the analytic aspects. IT, J-PG, SM, SQ, and CP-W contributed to the development and implementation of isotopic analytic methods and to isotope analysis and data collection. ME and CH performed all statistical testing. MR wrote the manuscript, and all authors were involved in interpreting the results and in critical revision of the paper. No authors had any advisory board affiliations.
REFERENCES
Related articles in AJCN: